WO2007123692A2

WO2007123692A2 - Detecting and minimizing effects of optical network faults

Info

Publication number: WO2007123692A2
Application number: PCT/US2007/007928
Authority: WO
Inventors: David A. Delew; Ryan D. Houlgate; Nathan D. Knuth; Manie C. Steyn; Weijun Li; Joseph D. Miguel; E. Russell Smith
Original assignee: Tellabs Petaluma, Inc.
Priority date: 2006-04-05
Filing date: 2007-03-30
Publication date: 2007-11-01
Also published as: EP2002576A2; WO2007123692A3; CA2639905A1

Abstract

A method of identifying a passive optical network failure comprising: identifying a control optical network terminal (ONT) from among multiple ONTs in a passive optical network, the control ONT functioning normally with a normal, non-data, output signal level; identifying a test ONT from among the multiple ONTs, the test ONT potentially malfunctioning with an above normal, non-data, output signal level; and determining the test ONT is actually malfunctioning by attempting to range the control ONT and the test ONT and observing both ONTs fail to range. Apparatus, network and computer-readable medium associated.

Description

DETECTING AND MINIMIZING EFFECTS OF OPTICAL NETWORK FAULTS

RELATED APPLICATIONS

This application is: i) a Continuation of U.S. Application No. 1 1/651,329 entitled "Method and Apparatus for Rogue Tolerant Ranging and Detection," filed January 8, 2007, which: (i.l) claims the benefit of U.S. Provisional Application No. 60/848,955, entitled "Method and Apparatus for Rogue Tolerant Ranging and Detection," filed on October 3, 2006; and i.2) which is a Continuation-In-Part of U.S. Application No. 1 1/515,504 entitled "Method and Apparatus for Identifying a Passive Optical Network Failure," filed on September 1 , 2006 which: i.2. a) claims the benefit of U.S. Provisional Application No. 60/793,748, filed on April 21, 2006; and i.2.b) 11/515,504 also is a Continuation-In-Part of U.S. Application No. 1 1/514,461 entitled "Method and Apparatus for Diagnosing Problems on a Time Division Multiple Access (TDMA) Optical Distribution Network (ODN)," filed on August 31, 2006, which claims the benefit of U.S. Provisional Application No. 60/789,357, filed on April 5, 2006; and this application is also: ii) a Continuation of U.S. Application No. 1 1/514,421 entitled "Method and Apparatus for ONT Ranging with Improved Noise Immunity," filed on September 1, 2006, which is: ii.l) a Continuation-In-Part of U.S. Application No. 11/432,292 entitled "Method and Apparatus for ONT Ranging with Improved Noise Immunity," filed on May 10, 2006; and which is ii.2) a Continuation-In-Part of U.S. Application No. 1 1/514,461 entitled "Method and Apparatus for Diagnosing Problems on a Time Division Multiple Access (TDMA) Optical Distribution Network (ODN)," filed on August 31, 2006; 1 1/432,292 and 11/514,461 claim the benefit of U.S. Provisional Application No. 60/789,357, filed on April 5, 2006. The entire teachings of the above applications are incorporated herein by reference.

BACKGROUND OF THE INVENTION

A passive optical network (PON) can contain multiple Optical Line Terminals (OLTs), each connected by a shared optical fiber to a respective Optical Distribution Network (ODN) with multiple Optical Network Terminals (ONTs) on individual optical fibers. ONTs can malfunction and interfere with communications between the ONTs and the OLT on a shared optical fiber. Such malfunctions are generally the result of power outages or typical communication systems errors or failures. Other disruptions in communications can be caused by optical fibers being cut, such as by a backhoe. If ONTs are malfunctioning for any other reason, identifying the issue requires a technician to inspect each ONT, possibly causing costly interruptions to service.

SUMMARY OF THE INVENTION

A method or corresponding apparatus for quickly determining a particular optical network terminal (ONT) is malfunctioning in a passive optical network

(PON) in accordance with an embodiment of the present invention is provided. An example embodiment includes: identifying a control ONT from among multiple ONTs in a passive optical network, the control ONT functioning normally with a normal, non-data, output signal level; identifying a test ONT from among the multiple ONTs, the test ONT potentially malfunctioning with an above normal, non- data, output signal level; and determining the test ONT is actually malfunctioning, as opposed to being a different network fault, such as a line cut or power outage, by attempting to range the control ONT and the test ONT and observing both ONTs fail to range. A method for diagnosing problems on a time division multiple access

(TDMA) optical distribution network (ODN) is provided. A method according to an example embodiment of the invention includes: (i) measuring a no-input signal power level on a communications path configured to carry upstream communications between multiple optical network terminals (ONTs) and an optical line terminal (OLT) in a passive optical network (PON) at a time no upstream communications are on the communications path from the ONTs to the OLT; (ii) comparing the measured no-input signal power level to a threshold; and (iii) generating a notification in an event the threshold is exceeded.

A method for ranging an optical network terminal (ONT) in a passive optical network (PON) is provided. The method according to an example embodiment of the invention includes: (i) transmitting a ranging request from an Optical Line Terminal (OLT) to an ONT in connection with a transport layer ranging window; (ii) monitoring for a ranging response from the ONT during at least one physical layer ranging window within the transport layer ranging window, the transport layer ranging window having a duration longer than the physical layer ranging window; and (iii) determining at least one metric associated with the ranging response for use in connection with upstream communications between the ONT and the OLT. The metric(s), used in connection with upstream communications, are accurately determined, and communications faults during normal operations are thus reduced.

A method or corresponding apparatus for ranging an optical network terminal (ONT) which is tolerant to a fault condition is provided in accordance with an embodiment of the present invention. An example embodiment includes resetting a receiver of an optical line terminal (OLT) at about a time a ranging response from an ONT is expected to be received to tolerate a fault condition otherwise affecting ranging of the ONT.

BRIEF DESCRIPTION OF THE DRAWINGS The foregoing will be apparent from the following more particular description of example embodiments of the invention, as illustrated in the accompanying drawings in which like reference characters refer to the same parts throughout the different views. The drawings are not necessarily to scale, emphasis instead being placed upon illustrating embodiments of the present invention. FIG. IA is a block diagram of an example passive optical network (PON) employing embodiments of the present invention;

FIG. 1 B is a network diagram illustrating an example technique of determining a control optical network terminal (ONT) and a test ONT in a network employing an embodiment of the present invention; FIG. 1C is a network diagram illustrating an example technique of verifying a test ONT is malfunctioning with an above normal, non-data, output signal;

FIG. 2 is a flow diagram representing the example techniques of FIGS. IB and 1C;

FIGS. 3A-3D are network diagrams illustrating a method for identifying control ONTs and test ONTs;

FIG. 4 is a flow diagram illustrating a method for attempting to range multiple ONTs together and identifying the ONTs that fail to range; FIG. 5 is a flow diagram illustrating a method for identifying control ONTs;

FIG. 6 is a flow diagram illustrating a method for verifying a control ONT;

FIG. 7 is a flow diagram illustrating a method for identifying a test ONT;

FIGS. 8 A — 8 J are network diagrams illustrating a method for identifying a test ONT;

FIGS. 9 A - 9C are flow diagrams illustrating a method for identifying a test ONT;

FIG. 10 is a block diagram illustrating an apparatus for identifying a passive optical network (PON) fault; FIG. 11 is a block diagram illustrating a control ONT identification module;

FIG. 12 is a block diagram illustrating a test ONT identification module;

FIG. 13 is a block diagram illustrating a verification module;

FIG. 14 is a block diagram illustrating an optical line terminal (OLT) containing a notification generator; FIG. 15 is a flow diagram illustrating a method for identifying a PON failure and notifying an operator that an ONT is malfunctioning;

FIG. 16 is block diagram illustrating a PON capable of identifying that a test ONT is malfunctioning;

FIG. 17 is a block diagram illustrating a computer-readable medium containing a sequence of instructions which enable a processor to identify a PON failure;

FIG. 18 is a network diagram of an example PON;

FIG. 19 is a power level diagram illustrating power levels associated with an input signal and a no-input signal in accordance with example embodiments of the invention;

FIG. 20A is block diagram illustrating layer 2 communications established between an OLT and ONTs in accordance with example embodiments of the invention;

FIG. 2OB is a network block diagram illustrating measuring a no-input signal power level on an upstream communications path prior to establishing layer 2 communications between an OLT and an ONT in accordance with example embodiments of the invention; FIG. 2OC is a network block diagram illustrating measuring a no-input signal power level on an upstream communications path after establishing layer 2 communications between an OLT and ONTs in accordance with example embodiments of the invention; FIGS. 21A-21C are upstream communications frames illustrating example embodiments of measurements of a no-input signal power level on an upstream communications path being measured during a time there are no upstream communications;

FIG. 22 is a power level diagram illustrating an extinction ratio and no-input extinction ratio in accordance with example embodiments of the invention;

FIG. 23A is a power level diagram illustrating an integrated no-input signal power level ramping over time;

FIG. 23B is a timing diagram illustrating an integrated no-input signal power level ramping over a ranging window; FIG. 24A is a block diagram of an example OLT;

FIG. 24B is a block diagram of an example processor supporting example embodiments of the invention;

FIG. 25A is a flow diagram of an example process performed in accordance with an example embodiment of the invention; FIG. 25B is a flow diagram of an example process performed in accordance with an example embodiment of the invention.

FIG. 26 is a message diagram illustrating a procedure of ranging an ONT;

FIGS. 27 A and 27B are message diagrams illustrating communications from communicating ONTs halted, and a ranging request and a ranging response exchanged, during a transport layer ranging window in accordance with an example embodiment of the invention;

FIG. 28 is a diagram illustrating lengths of a transport layer ranging window, physical layer ranging window, and ranging response in accordance with an example embodiment of the invention; FIGS. 29A and 29B are a timing diagram illustrating an integrated no-input signal power level ramping over a physical layer ranging window in accordance with example embodiments of the invention; FlG. 30 is a series of timing diagrams illustrating dynamically adjusting a physical layer ranging window in an iterative manner in accordance with an example embodiment of the invention;

FIG. 31 is a series of timing diagrams illustrating shifting a physical layer ranging window within a transport layer ranging window in accordance with an example embodiment of the invention;

FIGS. 32A-C are a series of timing diagrams illustrating shifting a physical layer ranging window incrementally across the transport layer ranging window in accordance with an example embodiment of the invention; FIG. 33 is a timing diagram illustrating shifting a physical layer ranging window by an amount expected to result in receiving a ranging response in full in accordance with an example embodiment of the invention;

FIG. 34 is a timing diagram illustrating lengthening the duration of the physical layer ranging window in accordance with an example embodiment of the invention;

FIGS. 35 is a diagram illustrating monitoring for ranging response during a series of physical layer ranging windows in accordance with an example of embodiment of the invention;

FIGS. 36A and 36B are diagrams illustrating shifting a series of physical layer ranging windows by an amount expected to result in receiving a ranging response in full in accordance with an example embodiment of the invention;

FIG. 37 is a diagram illustrating reducing a physical layer ranging window if a measured no-input signal power level exceeds a threshold in accordance with an example embodiment of the invention; FIG. 38 is a block diagram of an example optical line terminal (OLT) supporting example embodiments of the invention;

FIG. 39 is a block diagram of an example monitor unit supporting examples embodiments of the invention;

FIG. 40 is a flow diagram of an example process performed in accordance with an example embodiment of the invention;

FIG. 41 is a flow diagram of an example process performed in accordance with an example embodiment of the invention; FIG. 42 is a flow diagram of another example process performed in accordance with an example embodiment of the invention.

FIG. 43 is a block diagram of an example system to tolerate a fault condition otherwise affecting ranging of an ONT in accordance with an embodiment of the present invention;

FIG.44 is a timing diagram illustrating an integrated no-input signal power level ramping over a ranging window;

FIG. 45 is a timing diagram illustrating resetting a receiver of an optical line terminal (OLT) at about a time a ranging response from an optical network terminal (ONT) is expected to be received in accordance with an embodiment of the present invention;

FIGS. 46A-46B are timing diagrams illustrating changing a time to reset a receiver of an OLT by adding and subtracting a delay in accordance with embodiments of the present invention; FIG. 47 is a timing diagram illustrating changing a time to reset a receiver of an OLT by delaying for one or more delay increments in accordance with an embodiment of the present invention;

FIGS. 48A-48B are timing diagrams illustrating incrementing a time to reset a receiver of an OLT with each successive ranging attempt in accordance to an embodiment of the present invention;

FIG. 49 is a timing diagram illustrating incrementing a time to reset a receiver of an OLT through a range of delay increments in accordance with an embodiment of the present invention;

FIG. 50 is a flow chart of an example process ranging an ONT in accordance with an embodiment of the present invention; and

FIG. 51 is a flow chart of an example process identifying a fault condition in accordance with an embodiment of the present invention.

DETAILED DESCRIPTION OF THE INVENTION

A description of example embodiments of the invention follows. FIG. IA is an example passive optical network (PON) 10 illustrating an optical line terminal (OLT) 15 in communication with n number of optical network terminals (ONTs) 20a, 20b, 20c...20«, via optical communication paths 25and 27a, 27b, 27c...27« and an optical splitter/combiner 29. In this example, a first ONT 20a is undergoing a ranging procedure, which includes receiving a ranging request or grant 30 and responding with a ranging response 35. A second ONT 20b is not communicating with the OLT 15 during the ranging procedure between the OLT 15 and the first ONT 20a. However, in this example, rather than sending upstream communications and communicating with the OLT 15, the second ONT 20b is observed sending a no-input signal 40 — a form of non-communication, which is an example of a fault caused by a faulty optical transmitter outputting optical power when none should be. As also illustrated, a third ONT through and an nth ONT 20c...20« are communicating and sending upstream communications 45.

A fault in a PON or a PON fault may be detected by comparing a result of ranging a test ONT, such as the first ONT 20a, with a result of ranging a control ONT, in accordance with example embodiments described in reference to FIGS. 1 B- 17. A PON fault may also be detected by measuring a power level associated with a no-input signal, such as the no-input signal 40, on an upstream communications path, in accordance with example embodiments described in reference to FIGS. 18- 25b.

Effects of a PON fault may be minimized by ranging an ONT during a physical layer ranging window that may be shorter in duration than a transport layer ranging window, in accordance with example embodiments described in reference to FIGS. 26-42. Effects of a PON fault may also be minimized by resetting a receiver or transceiver of an OLT, such as the OLT 15, at about a time a ranging response (e.g., the ranging response 35) is expected to be received, in accordance with example embodiment described in reference to FIGS. 43-51. FIGS. IB-17 illustrate example embodiments for detecting a passive Optical network (PON) fault by comparing or otherwise verifying a result of ranging a test optical network terminal (ONT) or otherwise a suspected rogue ONT with a result of ranging a control ONT. As used herein, a control ONT is an ONT functioning normally with a normal, non-data, output signal level. In contrast, a test ONT is an ONT that is potentially malfunctioning with an above normal, non-data, output signal level. A rogue ONT is an ONT that has an optical transmitter that outputs an above normal output signal level when not transmitting data. A non-data signal level refers to a signal level output by a transmitter in an ONT during a time period in which it is not transmitting data (i.e., 1 's or O's) in the upstream direction, as illustrated in the example network herein.

Normal, non-data, signal levels are less than -4OdBm, such as between - 6OdBm and -8OdBm. Logical "zero" data signal levels are typically about -5dBm, and logical "one" data signal levels are typically between about IdBm and 3dBm. An above-normal, non-data signal level has been observed to be between -35dBm and -25dBm, but higher levels are also possible. Above-normal, non-data signal levels are caused by a failure in an optical transmitter and can lead to upstream communications errors due to measurements made during a ranging process or as a result of the above-normal, non-data levels adversely affecting an optical receiver during normal communications. In the ranging process scenario; the measurement errors may disrupt upstream communications for some or all ONTs communicating with an optical line terminal (OLT). When a rogue ONT is present in a passive optical network (PON) it may not initially appear as a failure depending on the sensitivity of the corresponding PON card to detect non-data signals. Additionally, it may not initially affect the communication of other ONTs with the OLT. The rogue ONT typically causes a failure in communications when the OLT requests the ONTs in the same optical distribution network (ODN) as the rogue ONT to range. The above normal, non- data, output signal coming from the rogue ONT causes the ONTs on the shared optical fiber to'fail to range, adversely affecting it own or multiple ONTs' communications with the OLT. Other times a PON is typically affected by a rogue ONT is when a new ONT is added to an ODN and the ONT is a rogue ONT or when an ONT loses ranging on an ODN containing a rogue ONT.

FIGS. IB and 1C are network diagrams illustrating an example method of identifying a control ONT and a test ONT and verifying that the test ONT is actually malfunctioning (i.e.. a rogue ONT) by having an above normal, non-data, output signal. This example method is referred to herein as a rogue ONT detection method. In FIG. IB, an OLT 105 is shown containing a control ONT identification module 1 10 and a test ONT identification module 1 15. Each ONT 135a-l 35e sends non- data signals 145a-145e and communication signals (not shown) in an upstream direction up individual optical fibers 140a - 14Oe. The signals are combined at a splitter/combiner 130, and the combined output 150 is sent to the OLT 105. In operation, the OLT 105 performs the rogue ONT detection method by first using the combined output 150 to determine if the network is rogue affected. If the network is rogue affected, then the combined output is used to determine if at least one control ONT can be identified using the control ONT identification module 1 10. If at least one control ONT is identified, the combined output 150 is used to identify a test ONT using the test ONT identification module 1 15. The control ONT identification module 110 isolates a control ONT, here illustrated as ONT 135a. The test ONT identification module 1 15 isolates a test ONT that is potentially malfunctioning, here illustrated as ONT 135c.

The output indicators 145a — 145e represent the output signal levels of the respective ONTs 135a - 135e. An ONT with an output indicator of "normal output" is an ONT that is functioning normally with a normal, non-data, output signal level and can be defined as a control ONT as is ONT 135a. An ONT with an output indicator of "above normal output," illustrated in this example as ONT 135c, is potentially malfunctioning with an above normal, non-data, output signal level.

Referring to FIG. 1C, a verification module 120 in the OLT 105 distinguishes the type of malfunction ONT 135c, the test ONT, is experiencing by attempting to range the test ONT 135c with ONT 135a, the control ONT. Ranging requests 155a and 155c are sent down optical fibers 140a and 140c to range ONT 135a with ONT 135c. The control ONT 135a and the test ONT 135c responsively send ranging responses 160a and 160c up the optical fibers 140a and 140c to the verification module 120. If ONT 135a, the control ONT, is unable to range with ONT 135c, the test ONT, the verification module 120 confirms the test ONT 135c is malfunctioning because of an above normal, non-data, output signal level rather than, for example, a power outage, typical communications system errors or failures, or a broken optical fiber.

FIG. 2 illustrates a method of identifying a passive optical network (PON) failure. A control ONT and a test ONT are identified (205, 210) from among the multiple ONTs. The test ONT is verified (215) as malfunctioning with an above normal, non-data, output signal level by attempting to range the control ONT with the test ONT and observing both ONTs fail to range.

Before describing details of the generalized description of FIGS IB, 1 C and 2 above, an enumerated listing illustrating an embodiment that may be used to identify an ONT transmitting an above-normal, non-data signal level is presented. For purposes of simplifying the enumerated listing, an ONT transmitting an above- normal, non data signal level is referred to as a "rogue" ONT. The term E-STOP refers to an emergency stop state that effectively shuts off an ONT transmitter, thereby preventing it from sending signals to the OLT. 1. Determine if a PON is affected by a rogue ONT: a. create a list of existing ONTs in the PON; b. force all of the ONTs of the PON to un-range then to range; c. create a list of ONTs that fail to range, if all ONTs range, the PON is not affected by a rogue ONT; d. E-STOP all except a first "un-ranged ONT;" e. attempt to range the first ONT on the list to determine if a rogue ONT was preventing it from ranging previously in step Ic above; f. if the first un-ranged ONT can now range, label the ONT as a "control ONT;" g. since it is possible that the first un-ranged ONT was powered down and coincidentally was powering up during the ranging request, check the next un-ranged ONT on the list by E-STOP all except the second un-ranged ONT. Then attempt to range the second ONT; h. if the second ONT can now range, label the ONT as a second control ONT; i. the process of identifying control ONTs can either abort after the first control ONT is identified or continue to identify multiple control ONTs. 2. Isolate the rogue ONT by one of two methods or a blend of the methods: a. Multi-Rogue Algorithm: i. sequence through all the ONTs on the list and attempt to range each one individually while all other ONTs on the list are E-STOPed, labeling all ONTs that fail to range as "test ONTs." b. Single-Rogue Algorithm: i. divide the existing ONTs in half, E-STOP one half and attempt to range the other half, if the other half ranges the rogue ONT is one of the E-STOPed ONTs; ii. sequence through dividing the group of ONTs known to contain the rogue ONT in half and determining which half contains the rogue ONT. When the size of each half is one ONT, label the ONT that fails to range as the "test ONT."

3. Verify a test ONT is a rogue ONT: a. sequence through the list of test ONTs, attempting to range all, or at least a subset of, control ONTs with each test ONT, while all other ONTs are E-STOPed. Those test ONTs that prevent all (or at least the subset of) control ONTs from ranging are further verified in the next step. Those that do not prevent the control ONTs from ranging are removed from the test ONT list; b. to further verify the test ONTs, E-STOP all existing ONTs except the control ONTs. Wait for the control ONTs to range. Check if all (or at least the subset of) the control ONTs are ranged. If the control ONTs range with the test ONTs in E-STOP, the test ONTs are rogue ONTs, and the and the verification process has eliminated broken optical fibers, power outages, and typical communications systems errors or failures as the cause of the malfunction in the PON.

4. Present a list of verified rogue ONTs to an operator. FIGS. 3A - 3D are network diagrams illustrating identifying a method for identifying control ONTs and test ONTs. In FIG 3 A, an OLT 340 sends ranging requests 310a-310c down shared optical fibers 315a - 315c to splitter/combiners 320a — 320c. The splitter/combiners 320a — 320c send the ranging requests down the individual communications paths 325a - 325o to ONTs 305a- 305o. The ONTs 305a - 305o send ranging responses 330a — 330c back to the OLT 340. In this illustration, the ONTs 305f- 305j are identified as failing to range. Referring to FIG. 3B, the OLT (not shown) sends a signal 31 1b, such as an

E-Stop ON or E-Stop OFF signal, to disable or enable the outputs of the ONTs 305f - 305j down the shared optical fiber 315b to the splitter/combiner 320b, which, in turn, directs the signal 31 Ib to the ONTs 305f- 305j. The indicators 335f- 335j above respective communications paths 325f — 325j illustrate that the output of ONTs 305f- 305j are disabled.

In FIG. 3C, the OLT (not shown) sends another ranging request signal 31 Ob to each ONT 305f — 305j individually and receives back a ranging response signal 33Ob indicating whether the ONTs 305f - 305j are able to range individually. Between each ranging request signal 31 Ob, the OLT sends a signal 31 1b (not shown) enabling and disabling the outputs of the ONTs 3O5f - 305j in turn, such that only the output of the ONT to be ranged is enabled. The indicators 335f- 335j illustrate the status of the outputs of ONTs 305f — 305j for each ranging request.

Referring to FIG. 3D, the ONTs 305f, 305g_; 305i, and 305j are illustrated as having ranged and may be defined as control ONTs. The ONT 305h in this example is illustrated as having failed to range and is defined as a test ONT.

FIG. 4 is a flow diagram 400 illustrating a method for attempting to range the multiple ONTs of the PON together and determining which ONTs fail to range. After the flow diagram starts (405), an attempt is made to range the multiple ONTs of the PON (410). Cycling through each ONT in the PON (415), the ONT is checked to determine if it ranges (420). If the ONT fails to range, it is added to a list of ONTs that fail to range (425). If the ONT ranges, it is not a control ONT or a test ONT, and the ONT is ignored. If the ONT being checked is the last ONT in the PON (430), the flow diagram 400 exits to the methods shown in FIGS. 5-7.

FIG. 5 is a flow diagram 500 illustrating a method to determine a control ONT. After a list has been made of the ONTs that fail to range by the method shown in FIG. 4, the outputs of the ONTs on the list are disabled (505). Starting with the first ONT on the list (510), the output of the ONT is enabled (515), and an attempt is made to range the ONT individually (520). If the ONT ranges (525), the ONT is a control ONT (530). Optionally, the cycle can exit after the first control ONT is determined (535). If the ONT does not range or more then one control ONT is needed, a check is made if, optionally, the ONT is the last ONT on the list (540) or if a condition is met (540). Such a condition includes at least one of the following: a time limit, a specified number of control ONTs have been identified, a percentage of the multiple ONTs are determined to be control ONTs, a percentage of the ONTs that failed to range are determined to be control ONTs, and a stop command from an operator is received. If the ONT is not the last ONT on the list or, optionally, the condition is not met, the cycle repeats from 505 through 540. If the ONT is the last on the list or the condition is met, the cycle is complete and flow diagram 500 exits (545).

FIG. 6 is a flow diagram 600 illustrating a method to verify an ONT is properly labelled as a control ONT. It is possible that the ONT identified as the control ONT is actually a test ONT, has a broken optical fiber, or was powered down and coincidentally powered up during the ranging request. Therefore, after a list has been made of the ONTs that fail to range by the method shown in FIG. 4, the outputs of the ONTs on the list are disabled (605). Where "z" represents an ONT on the list, starting with the first ONT on the list (610), the output of the ONT is enabled (615) and an attempt is made to range the ONT individually (620). If the

ONT ranges (625), a check is made to see if the previous ONT on the list was able to range individually (630). If yes, the ONT is verified as a control ONT (635) and the flow diagram 600 exits (640). If the ONT either fails to range individually (625) or the previous ONT on the list failed to range, a check is made if the current ONT is the last ONT on the list (645). If yes, the cycle is complete and flow diagram 600 exits (650). If no, the cycle is repeats from 605 through 645.

In another embodiment, after the outputs of the ONTs on the list are disabled (605), verifying an ONT is properly labelled as a control ONT optionally includes cycling through the ONTs of the multiple ONTs. Where "z" represents an ONT of the multiple ONTs, starting with the first ONT of the multiple ONTs (610), the output of the ONT is enabled (615) and an attempt is made to range the ONT individually (620). If the ONT ranges (625), a check is made to see if the ONT is on the list of ONTs that failed to range and the ONT is at least the second ONT of the multiple ONTs (630). If yes, the ONT is verified as a control ONT (635) and the flow diagram 600 exits (640). If the ONT either fails to range individually (625) or the ONT is not on the list of ONTs that failed to range and/or is not at least the second ONT of the multiple ONTs (630), a check is made if the current ONT is the last ONT of the multiple ONTs (645). If yes, the cycle is complete and flow diagram 600 exits (650). If no, the cycle is repeats from 605 through 645.

FIG. 7 is a flow diagram 700 illustrating a method for identifying a test ONT. After a list has been made of the ONTs that fail to range by the method shown in FIG. 4, the outputs of the ONTs on the list are disabled (705). Starting with the first ONT on the list (710), the output of the ONT is enabled (715) and an attempt is made to range the ONT individually (720). If the ONT fails to range (725), the ONT is a test ONT (730). If the ONT ranges or after it has been identified as a test ONT, the ONT is checked to determine if it is the last ONT on the list (730). If yes, all test ONTs have been identified and flow diagram 700 exits (740). If no, the cycle repeats from 705 through 735.

FIGS. 8A - 8J are network diagrams illustrating another method for identifying a test ONT when only one test ONT exists. Referring to FIG. 8A, the multiple ONTs of a PON are divided into a group 1 (805), illustrated as ONTs 815a and 815b, and a group 2 (810), illustrated as ONTs 815c-815e. An OLT (not shown) sends a signal 820 to disable the outputs of the ONTs down a shared optical fiber 821, through a splitter/combiner 825, and down the individual communication paths 83Oa - 83Oe to the ONTs 815a - 815e. The indicators 835a-835e above the respective communication paths 83Oa - 83Oe illustrate the outputs of ONTs 815a - 815e are disabled.

In FIG. 8B, the OLT (not shown) sends a signal 822 to enable the outputs of the ONTs of group 1 (805). The indicators 835a and 835b illustrate the outputs of ONTs 815a and 815b are enabled. Referring to FlG. 8C, the OLT (not shown) sends a ranging request signal 823 to group 1 (805). The ONTs, 815a and 815b, of group 1 (805) send ranging response signals 84Oa and 840b back confirming whether they range. In this illustration, all of the ONTs in group 1 (805) successfully range, indicating the test ONT is in group 2 (810). In FIG. 8D, group 2 (810) of FIG. 8C, known to contain the test ONT, is divided into two new groups, group 1 ' (806), illustrated as being ONT 815c, and group 2' (81 1 ), illustrated as being ONTs 815d and 815e. The OLT (not shown) sends a signal 820 to disable the outputs of all the ONTs. The indicators 835a - 835e illustrate the outputs of ONTs 815a - 815e are disabled. Referring to FIG. 8E, the OLT (not shown) sends a signal 822 to enable the output of the ONT of group 1 ' (806). The indicator 835c illustrates the output of ONT 815c is enabled. In FIG. 8F, the OLT (not shown) sends a ranging request signal 823 to the ONT of group 1 ' (806). ONT 815c sends back ranging response signal 840c confirming whether it ranges. In this illustration, group 1 ' (806) fails to range and, therefore, contains a test ONT. To verify that there is not a test ONT in group 2' (81 1) as well; group 2' (81 1) is also ranged.

In FIG. 8G, the OLT (not shown) sends a signal 820 to disable the outputs of the ONTs. The indicators 835a - 835e illustrate the outputs of ONTs 815a - 815e are disabled. Referring to FIG. 8H, the OLT (not shown) sends a signal 822 to enable the outputs of group T (81 1). The indicators 835d and 835e illustrate the outputs of the ONTs of group 2' (81 1) are enabled. In FIG. 81, the OLT sends ranging request signal 823 to the ONTs of group 2' (811). The ONTs 815d and 815c of group 2' (81 1) send ranging response signals 84Od and 84Oe back confirming whether they range. In this illustration, group 2' (81 1) successfully ranges indicating that group 1 ' (806) contains the test ONT. As shown in FIG. 8J, group 1 ' (806) contains only one ONT, ONT 815c. Therefore, ONT 815c is the test ONT.

FIGS. 9A - 9C are flow diagrams illustrating a method for identifying a test ONT as outlined in network diagrams FIGS. 8A — 8J. Group 1 and group 2 are defined from the multiple ONTs of the PON (902). The outputs of all the ONTs are disabled (903). Starting with the first group (904), the output of the group is enabled and an attempt is made to range the ONTs in the group (905). If the ONTS of the group successfully range (906). the other group contains the test ONT (907). If the number of ONTs in the group containing the test ONT is one (908), the ONT of that group is the test ONT (909), and the cycle is completed (910). If the group containing the test ONT has more then one ONT (908), that group is divided into a new group 1 and group 2 (91 1). The cycle repeats from 903 through 906.

If the ONTs of the group fail to range (906), a check is made if the group is group 2 (912). If the group is not, the cycle repeats from 903 through 906. If the group is group 2, then multiple test ONTs exist (913) and the method illustrated in FIG. 9B is used to identify the test ONTs. Referring to FIG. 9B, an attempt is made to range the multiple ONTs of the PON (914). Cycling through each ONT in the PON (915), the ONT is checked to determine if it ranges (916). If the ONT fails to range, it is added to a list of ONTs that fail to range (917). If the ONT does range, it is not a test ONT and is ignored. If the ONT being checked is the last ONT in the PON (918), the process exits to the method shown in FIG. 9C.

In FIG. 9C, the outputs of the ONTs on the list are disabled (919). Starting with the first ONT on the list (920), the output of the ONT is enabled (921), and an attempt is made to range the ONT individually (922). If the ONT fails to range (923), the ONT is defined as a test ONT (924). If the ONT ranges or after it has been identified as a test ONT, the ONT is checked to determine if it is the last ONT on the list (925). If yes, all test ONTs are identified and the cycle is complete (926). If no, the cycle repeats from 919 through 925.

FIG. 10 is a block diagram illustrating an apparatus for identifying a PON fault. An optical line terminal (OLT) 1005 includes a control ONT identification module 1010, a test ONT identification module 1015, and a verification module 1020. Reference number 1, 2, and 3 show a first, second, and third communication made with ONTs 1030a- 1030«. The control ONT identification module 1010, the test ONT identification module 1015, and the verification module 1020 in turn send a signal 1021 which includes a ranging request to the splitter/combiner 1025 and on to the individual ONTs 1030a-1030«. The ONTs 1030a-1030« send a ranging response signal 1022 back to the OLT 1005 indicating their ranging response. The control ONT identification module 1010 monitors the multiple ONTs and identifies control ONTs. Similarly, the test ONT identification module 1015 monitors the multiple ONTs and identifies test ONTs. The verification module 1020 is configured to determine that the test ONT is actually malfunctioning due to having an above normal, non-data, output signal by ranging the control ONT with the test ONT and observing both ONTs fail to range when the test ONT has its output enabled, and also observing the control ONT successfully ranges when that same test ONT has its output disabled.

FIG. 11 is a block diagram illustrating a control ONT identification module 1105. The control ONT identification module 1105 includes a ranging unit 1 110, an enabling/disabling unit 1 1 15, and a logic unit 1120. The ranging unit 11 10 and the logic unit 1 120 are in communication with one another. Optionally, the enabling/disabling unit 1 1 15 is in communication the ranging unit 1 1 10 and/or the logic unit 1 120. The enabling/disabling unit 1 1 15 sends signals to the ONTs (not shown) to either enable or disable their outputs, while the ranging unit 1 110 sends signals to attempt to range to the ONTs. The logic unit 1 120 identifies ONTs that successfully range individually as control ONTs.

In addition, the control identification module 1 105 can optionally include a verification unit 1 125 and/or limiting unit 1130, both in communication with the logic unit 1 120. The verification unit 1 125 verifies a control ONT identified by the logic unit 1 120 is not actually a test ONT, does not have a broken optical fiber, and was not powered down and coincidentally powering up at the time it was identified as a control ONT. The limiting unit 1 130 stops the logic unit 1 120 from identifying control ONTs when a specified condition has been met. The condition includes at least one of the following: a time limit, a specified number of control ONTs are determined, a percentage of the multiple ONTs are determined to be control ONTs, a percentage of the ONTs that failed to range are determined to be control ONTs. and a stop command from an operator is received.

FIG. 12 is a block diagram illustrating a test ONT identification module 1205. The test ONT identification module 1205 includes a ranging unit 1215, enabling/disabling unit 1220, and a logic unit 1225. The ranging unit 1215 and the logic unit 1225 are in communication with one another. Optionally, the enabling/disabling unit 1220 is in communication with the ranging unit 1215 and/or the logic unit 1225. The enabling/disabling unit 1220 sends signals to the ONTs (not shown) to either enable or disable their outputs, while the ranging unit 1215 sends signals to attempt to range to the ONTs. The logic unit 1225 identifies ONTs that fail to range individually as test ONTs or, optionally, identifies a group of ONTs that fail to range as containing a test ONT.

In addition, the test ONT identification module 1205 can optionally include a dividing unit 1210. a test ONT unit 1235, a verification unit 1230, and a switch unit 1240. The test ONT unit 1235 is in communication with the logic unit 1225 and the dividing unit 1210. The verification unit 1230 is in communication with the logic unit 1225 and switch unit 1240. The dividing unit 1210 defines two groups of ONTs. The test ONT unit 1235 communicates with the dividing unit 1210 to divide a group identified as containing a test ONT by the logic unit 1225 into two new groups and has the logic unit 1225 identify which of the new groups of ONTs fail to range. The verification unit 1230 checks whether only one group contains a test ONT. If the verification unit 1230 determines that both groups contain a test ONT, the verification unit 1230 notifies the switch unit 1240. The switch unit 1240 then sends a signal to the test ONT unit 1235 to attempt to range the ONTs individually and to identify ONTs that fail to range as test ONTs.

FIG. 13 is a block diagram illustrating a verification module 1305. The verification module 1305 includes a ranging unit 1310, an enabling/disabling unit 1315, a logic unit 1320, and a verification unit 1325. The ranging unit 1310 and the logic unit 1320 are in communication with one another. Optionally, the enabling/disabling unit 1315 is in communication with the ranging unit 1310 and/or the logic unit 1320. The verification unit 1325 is in communication with the logic unit 1320.

Once the control ONT identification module (not shown) identifies a control ONT and the test ONT identification module (not shown) identifies a test ONT, the enabling/disabling unit 1315 sends signals to the ONTs (not shown) either to enable or disable their outputs. The ranging unit 1310 then sends a signal to attempt to range the control ONT with the test ONT. The logic unit 1320 identifies whether the test ONT and control ONT range. If not, the verification unit 1325 confirms that the test ONT is malfunctioning by sending an above normal, non-data, output signal level rather than from a power outage, broken optical fiber, or typical communications systems errors or failures. FIG. 14 is a block diagram illustrating an OLT 1405 including a control ONT identification module 1410, a test ONT identification module 1415, a verification module 1420, and an optional notification generator 1425 in communication with the verification module 1420. Reference number 1 , 2, and 3 show a first, second, and third communication made with the ONTs 1435a- 1435«. The control ONT identification module 1410, the test ONT identification module 1415, and the verification module 1420 in turn send a ranging request signal 1426 to the splitter/combiner 1430 and on to the individual ONTs 1435a- 1435«. The ONTs 1435a-1435« send a ranging response signal 1427 back to the OLT 1405 indicating their ranging response. The control ONT identification module 1410 monitors the ONTs 1435a - 1435« and identifies control ONTs. Similarly, the test ONT identification module 1415 monitors the ONTs 1435a — 1435« and identifies test ONTs. The verification module 1420 ranges a control ONT with a test ONT and, if both fail to range, confirms the test ONT is malfunctioning by outputting an above normal, non-data, output signal. The notification generator 1425 generates a notification that an ONT is malfunctioning.

FIG. 15 is a flow diagram illustrating a method identifying a PON failure and notifying an operator that a test ONT is malfunctioning. A control ONT and a test ONT are identified (1505 and 1510) from among multiple ONTs in a passive optical network. The test ONT is verified (1515) as malfunctioning with an above normal, non-data, output signal by attempting to range the control ONT identified in 1505 with the test ONT identified in 1510 and observing both ONTs fail to range. Lastly, an operator is notified that a test ONT is malfunctioning (1520).

FIG. 16 is block diagram illustrating a PON 1640 capable of identifying that a test ONT is malfunctioning. Each OLT 1605 includes a control ONT identification module 1610, a test ONT identification module 1615, a verification module 1620, and an optional notification generator 1635 in communication with the verification module 1620. For each OLT 1605, reference numbers 1, 2, and 3 show a first, second, and third communication made with the ONTs 163Oa-1630«. The control ONT identification module 1610, the test ONT identification module

1615, and the verification module 1620 in turn send a ranging request signal 1621 to the splitter/combiner 1625 and on to the individual ONTs 163Oa-1630«. The ONTs 1630a- 1630« send the ranging response signal 1622 back to the OLT 1605 indicating their ranging response.

The control ONT identification module 1610 monitors the ONTs 1630a - 1630« and identifies control ONTs. Similarly, the test ONT identification module 1615 monitors the ONTs 163Oa - 1630« and identifies test ONTs. The verification module 1620 determines the test ONT is malfunctioning with an above normal, non- data, signal level by ranging a control ONT with a test ONT and observing both ONTs fail to range. The notification generator 1635 generates a notification that an ONT is malfunctioning. Optionally, a malfunctioning ONT signal 1645, indicating an ONT is malfunctioning with an above normal, non-data, signal level, is sent from a notification generator in a PON 1640 to a network management server 1650. The network management server 1650 is in communication with a service provider 1655 and can send an alert 1651 to a service provider 1655. Alternatively, a service provider 1655 can send a query 1652 to the network management server 1650 to determine if a malfunctioning ONT signal 1645 has been received from the PON 1640. Optionally, a malfunctioning ONT signal 1660 can be sent to an ONT where it will be received by, for example, a service operator, a client, and/or a communication device such as a local area network or a computer. FIG. 17 is a block diagram illustrating a computer-readable medium 1720 containing a sequence of instructions which identify a PON failure. The instructions include identifying a control ONT (1705) and identifying a test ONT (1710) from among multiple ONTs in a passive optical network. Lastly, an instruction verifies the test ONT (1715) as actually malfunctioning with an above normal, non-data, output signal by attempting to range the control ONT identified in 1705 with the test ONT identified in 1710 and observing both ONTs fail to range.

«/summary of 2090-00 l»The above description referring to FIGS. 1-17 describes determining a particular optical network terminal (ONT) in a passive optical network (PON) is malfunctioning by sending a continuous stream of light up a shared fiber, which results in adversely affecting communications between the ONT and an optical line terminal (OLT). An example embodiment verifies the failure is due to a faulty optical transmitter in the ONT and not a different network fault, such as a fiber optic line cut or power outage. Through the use of the example embodiment, a service provider can determine in an automated manner which specific ONT of a PON is malfunctioning.

To understand the problem further, greater details of the operations of a passive optical network (PON), including an optical receiver in an OLT, are discussed in reference to FIGS. 18-25B.

FIGS. 18-25B illustrate example embodiments of an aspect of the present invention in which a malfunctioning ONT is detected by looking for a presence of a modulated or unmodulated upstream optical signal when no signal should be present on an upstream communications path. Further illustrated is a manner for determining a malfunctioning ONT by looking for an inappropriate presence of unmodulated or very low level modulated upstream optical signal when no signal should be present on the upstream communications path.

An optical network terminal (ONT) can malfunction in such a way that it sends a continuous stream of light (e.g., low level, such as less than 1OdBm) up to a shared fiber of an optical distribution network (ODN). This can adversely affect communications between ONTs on the ODN and an optical line terminal (OLT). Using existing error detection techniques, such as those described in various passive optical network (PON) protocols, this type of ONT malfunction may not be detected. Even if it is detected (e.g., resulting from system failure), the ONT malfunction (i.e., output of continuous light at a low level) may not be identified,, and field service engineers may spend a great deal of time inspecting a receiver in the OLT, fiber optic cables between the ONTs and OLT, and any relays or junctions between the ONTs and OLT. Moreover, the amount of continuously outputted light which can cause communications errors has been found to be very low. So, unless field service engineers are sensitive to the source of the communications errors, hours of lost network services can result.

Detection of an ONT sending a low level continuous stream of light up to a shared fiber of an ODN may be done several ways. One method may involve individually disconnecting ONTs from the ODN to determine if there is a single ONT or multiple ONTs causing the problem. With this method, however, the problem may not be corrected in a timely fashion. Additionally, this method requires considerable customer downtime. In another method, the OLT may be disconnected from the ODN, and the ODN may be examined with additional test equipment.

Accordingly, what is needed is a method or corresponding apparatus for diagnosing problems on an ODN which detects, prior to establishing layer 2 communications, a malfunctioning ONT by looking for an inappropriate presence of a modulated or unmodulated upstream optical signal when no signal should be present on the upstream communications path. Furthermore, after establishing layer 2 communications with any number of ONTs, a malfunctioning ONT may be detected by looking for an inappropriate presence of an unmodulated or very low level modulated upstream optical signal when no signal should be present on the upstream communications path.

As used herein, a modulated upstream optical signal is a signal which conveys information (i.e., communicates upstream communications data) and is • interchangeably referred to herein as an "input signal"). The input signal may be either a "zero-bit input signal" (i.e., communicates a zero-bit) or a '"one-bit input signal," i.e., communicates a one-bit. In contrast, an unmodulated upstream optical signal is a signal which does not convey information (i.e., communicates no upstream communications data) and is interchangeably referred to herein as a "no- input signal.'^"

Further, power levels associated with a zero-bit input signal or a one-bit input signal are referred to herein as a "zero-bit input signal power level" or a "one- bit input signal power level," respectively. Additionally, a power level associated with a no-input signal is referred to herein as a "no-input signal power level." In a PON system, multiple ONTs transmit data to an OLT using a common optical wavelength and fiber optic media. Field experience has demonstrated that a malfunctioning ONT can send an optical signal up to the OLT at inappropriate times, resulting in the OLT not being able to communicate with any of the ONTs on the ODN. A typical PON protocol provides some functionality for detecting this problem, but is limited only to inappropriate modulated signals. Consequently, the following ONT malfunctions are not being detected. An example ONT malfunction not being detected involves an ONT sending a continuous upstream signal (modulated or unmodulated) up the fiber prior to attempting to establish communications with an OLT on an ODN. Another example ONT malfunction occurs when an ONT sends an unmodulated light signal up the fiber at an inappropriate time while attempting to establish communications or after having established communications with an OLT on an ODN. Consequently, an ability to detect whether a network contains an ONT with such a malfunction may depend on an ability to detect an unmodulated light signal.

While an OLT must be able to detect the presence of a modulated signal (or an input signal) in order to function as a node in a communications path, the ability to detect an unmodulated signal (or a no-input signal), however, is not required for operation. In accordance with example embodiments of the invention, the ability to detect an unmodulated upstream signal may improve the ability of the OLT to detect error conditions in upstream communications between ONTs and the OLT, as discussed hereinafter.

As such, in part, a difference between detecting a modulated versus an unmodulated upstream signal is that an optical receiver (or transceiver) does not have the ability to detect an unmodulated signal. In some cases, the optical receiver may not be able to detect or communicate the presence of an unmodulated upstream signal.

In other cases, even though the presence of an unmodulated signal may indicate a system problem,. the presence of an unmodulated signal may not actually result in a problem in upstream communications between ONTs and an OLT. Sometimes the presence of an unmodulated upstream signal is removed by signal conditioning circuitry on the optical receiver (or transceiver). The unmodulated upstream signal adds a "directed current (DC) offset" to a modulated upstream signal. The DC offset may be subsequently removed from the modulated upstream signal without corrupting it. Current experience, however, indicates that the effect of an unmodulated upstream signal on a modulated upstream signal varies from optical receiver to optical receiver. Additionally, the effect of the unmodulated upstream signal depends on the brightness or amplitude of the unmodulated upstream signal. FIG. 18 is a network diagram of an example passive optical network (PON) 1801. The PON 1801 includes an optical line terminal (OLT) 1802, wavelength division multiplexers 1803a-n, optical distribution network (ODN) devices 1804a-n, ODN device splitters (e.g., 1805a-n associated with ODN device 1804a), optical network terminals (ONTs) (e.g., 1806-n corresponding to ODN device splitters 1805a-n), and customer premises equipment (e.g., 1810). The OLT 1802 includes PON cards 1820a -n, each of which provides an optical feed (1821a-n) to ODN devices 1804a-n. Optical feed 1821a, for example, is distributed through corresponding ODN device 1804a by separate ODN device splitters 1805a-n to respective ONTs 1806a-n in order to provide communications to and from customer premises equipment 1810.

The PON 1801 may be deployed for fiber-to-the-business (FTTB), fiber-to- the-curb (FTTC), and fϊber-to-the-home (FTTH) applications. The optical feeds 1821a-n in PON 1801 may operate at bandwidths such as 155 Mb/sec, 622 Mb/sec, 1.25 Gb/sec, and 2.5 Gb/sec or any other desired bandwidth implementations. The PON 1801 may incorporate asynchronous transfer mode (ATM) communications, broadband services such as Ethernet access and video distribution, Ethernet point-to- multipoint topologies, and native communications of data and time division multiplex (TDM) formats. Customer premises equipment (e.g., 1810) which can receive and provide communications in the PON 1801 may include standard telephones (e.g., Public Switched Telephone Network (PSTN)), Internet Protocol telephones, Ethernet units, video devices (e.g., 181 1), computer terminals (e.g., 1812), digital subscriber line connections, cable modems, wireless access, as well as any other conventional device. A PON 1801 includes one or more different types of ONTs (e.g., 1806a-n).

Each ONT 1806a-n, for example, communicates with an ODN device 1804a through associated ODN device splitters 1805a-n. Each ODN device 1804a-n in turn communicates with an associated PON card 1820a-n through respective wavelength division multiplexers 1803a-n. Wavelength division multiplexers 1803a-n are optional components which are used when video services are provided.

Communications between the ODN devices 1804a-n and the OLT 1802 occur over a downstream wavelength and an upstream wavelength. The downstream communications from the OLT 1802 to the ODN devices 1804a-n may be provided at 622 megabytes per second, which is shared across all ONTs connected to the ODN devices 1804a-n. The upstream communications from the ODN devices 1804a-n to the PON cards 1820a-n may be provided at 155 megabytes per second, which is shared among all ONTs connected to ODN devices 1804a-n.

Error conditions in upstream communications between an optical line terminal (OLT) and optical network terminals (ONTs) often result in layer 2 communication errors, for example, errors in ranging or normalization parameters. One such error condition in upstream communications is the presence of an unmodulated signal (or a no-input signal) on an upstream communications path.- An example solution to this problem may include detecting the presence of an unmodulated signal on the upstream communications path, identifying whether the detected unmodulated signal leads to a layer 2 communications error, and communicating the error condition so that it may be corrected. An unmodulated signal on the upstream communications path may be detected by measuring a power level associated with the unmodulated signal. For the sake of readability, the power level associated with the unmodulated signal is referred to herein as a "no-input signal power level" and is used throughout this disclosure.

FIG. 19 illustrates three power levels: a minimum logical one input signal power level 1920, a maximum logical zero input signal power level 1925, and a maximum no-input signal power level 1930. The terms logical one and logical zero are interchangeably referred to herein as a one-bit and a zero-bit.

In general, when the power level of an input signal is above the minimum logical one input signal power level 1920, the input signal is designated as a logical one input signal. When the power level of an input signal is below the maximum logical zero input signal power level 1925, the input signal is designated as a logical zero input signal. When the power level of an input is below the minimum logical one input signal power level 1920 but above the maximum logical zero input signal power level 1925, the input signal is indeterminate, i.e., the input signal is neither a logical one input signal nor is the input signal a logical zero input signal.

In this way, by modulating or otherwise changing the power level of an input signal, the input signal can either convey a logical one input signal or a logical zero input signal. Moreover, by modulating the power level of an input signal, the input signal conveys information. Accordingly, upstream communications between an ONT and OLT on an upstream communications pathway is accomplished by modulating the power level of an input signal to an optical transmitter generating optical signals.

In contrast, when the power level of a signal is not modulated, the signal conveys no information. This is the case when there are no upstream communications between an ONT and an OLT on an upstream communications pathway. In this disclosure, the term no-input signal is used to describe a signal whose power level is not modulated. Furthermore, the terms unmodulated signal and no-input signal are used interchangeably throughout this disclosure.

When the power level of a no-input signal is below the maximum no-input signal power level 1930, a no-input signal is said to be valid or non-faulty. More specifically, a no-input signal with a power level less than the maximum no-input signal power level 1930 does not or is less likely to cause an error condition. On the other hand, when the power level of a no-input signal is above the maximum no- input signal power level 1930, the no-input signal is said to be invalid or faulty. In contrast to a no-input signal with a power level less than the maximum no-input signal power level. 1930, a no-input signal with a power level greater than the maximum no-input signal power level 1930 does or is more likely to cause an error condition (described later in greater detail).

Still referring to FIG. 19, consider the following illustrative example. The minimum logical one input signal power level 1920 is +3dBm (decibel-milliwatt), the maximum logical zero input signal power level 1925 is -5dBm, and the maximum no-input signal power level 1930 is -4OdBm.

An input signal 1932 with a series of power levels 1935 is received during a grant timeslot 1940. During the grant timeslot 1940. the input signal 1932 has power levels which at times are greater than +3dBm and at times are less than - 5dBm. Thus, the series of power levels 1935 in the input signal 1932 designates a series of logical ones and logical zeros. Before the grant timeslot 1940, a first no- input signal portion 1945a of the input signal 1932 has a power level less than - 4OdBm. As such, the first no-input signal portion 1945a of the input signal 1932 is not faulty, i.e., validly conveys no information.

In contrast, after the grant timeslot 1940, a second no-input signal portion 1945b of the input signal 1932 has a power level greater than -4OdBm, e.g., a "faulty no-input signal level" 1950. In this case, the second no-input signal portion 1945b of the input signal 1932 is faulty, i.e., invalidly conveys no information. Discussed later in greater detail, a no-input signal having a power level, such as the faulty no- input signal power level 1950, may lead to problems in upstream communications, e.g., errors in ranging and normalization parameters. FIG. 2OA illustrates upstream communications between an OLT 2005 and communicating ONTs 2010a-n over an upstream communications path 2015. Upstream communications begins when the communicating ONTs 2010a-n transmit upstream communications data 2020a-n on the upstream communications path 2015. Upstream communications data 2020a-n are then combined on the upstream communications path 2015 by a splitter/multiplexer 2025. Upstream communications data 2020a-n are transmitted by the communicating ONTs 2010a-n at respective predefined times and in the case of a time division multiplexing (TDM) communications protocol, placed into individual timeslots 2030a-n of an upstream communications frame 2035. The OLT 2005, via the upstream communications path 2015, receives the upstream communications frame 2035. The OLT 2005 may then demultiplex (i.e., separate) the upstream communications frame 2035 into individual timeslots 2030a- n. As a result, the OLT 2005 receives respective upstream communications data 2020a-n from each communicating ONT 2010a-n. FIG. 2OB is a network block diagram illustrating how an OLT 2005 may measure a power level of a no-input signal (or a no-input signal power level) on an upstream communications path 2015 at a time there are no upstream communications between the OLT 2005 and communicating ONTs 2010a-n. The no-input signal power level on the upstream communications path 2015 may be measured at a time the OLT 2005 is ranging an ONT 2011 or at another time there are no upstream communications on the upstream communications path 2015, e.g., when the OLT 2005 is immediately rebooted and before any ONTs are ranged. In an example embodiment, the OLT 2005 may instruct all communicating ONTs 2010a-n to halt upstream communications in order to range the ONT 201 1. With upstream communications from the communicating ONTs 2010a-n halted, the no-input signal power level on the upstream communications path 2015 should be small, (e.g., a power level below the maximum no-input signal power level 1930 of FIG. 19) or have no value. Typically, once halted, any power present on the upstream communications path 2015 is caused by, for example, very low level leakage of optical transmitters (e.g., laser diodes) in transmitter units of the communicating ONTs 2010a-n or due to typical optical noise developed or imparted onto the upstream communications path 2015.

The OLT 2005 may send the ONT 201 Ia ranging request 2040. The ONT 201 1, in turn, may respond with a ranging response 2045. During the ranging, the no-input signal power level on the upstream communications path 2015 is measured during period(s) the ranging response 2045 is not on the upstream communications path 2015. As such, the no-input signal power level is not increased by a signal representing the ranging response 2045. If the no-input signal power level is greater than, for example, the maximum no-input signal power level 1930 of FIG. 19, the ONT 201 1 is faulty.

The ranging exchange between the OLT 2005 and the ONT 201 1 may occur over a period of time known as a ranging window (not shown, but discussed below in reference to FIG. 23B). The measured no-input signal power level on the upstream communications path 2015 may be averaged over an un-allocated grant window (not shown). In addition to measuring a no-input signal power level during the un-allocated grant window, a no-input signal power level may also be measured before any ONTs have been ranged, e.g., when the OLT 2005 is rebooted.

FIG. 2OC is a network block diagram in which upstream communications between an OLT 2005 and communicating ONTs 2010a-n are carried over an upstream communications path 2015 . In addition to the communicating ONTs 2010a-n, there is a non-communicating ONT 2013. Upstream communications begin with the communicating ONTs 2010a-n sending upstream communications data 2020a-n via the upstream communications path 2015. The non-communicating ONT 2013 may have no-data to send. Consequently, rather than sending upstream communications data 2020, nothing is sent, denoted by a "no-data" indicator 2023. For purposes of explaining aspects of the invention, the "no-data" indicator 2023 indicates a timeslo.t portion that is neither filled with an "idle" signal or a substantive upstream communications signal. The upstream communications data 2020a-n and the no-data indicator 2023 are then combined by splitter/multiplexer 2025. The upstream communications data 2020a-n and the no-data indicator 2023 are transmitted in their respective timeslots 2030a-n of upstream communications frame 2035.

The OLT 2005, via the upstream communications path 2015, receives the upstream communications frame 2035. The OLT 2005 then demultiplexes (or separates) the upstream communications frame 2035 into individual timeslots 2030a-n. Consequently, the OLT 2005 receives from each communicating ONT 2010a-n upstream communications data 2020a-n. The OLT 2005 also receives the no-data indicator 2023 from the non-communicating ONT 2013. While the OLT 2005 is "receiving" the no-data indicator 2023 in the timeslot

2030c of the upstream communications frame 2035, a no-input signal power level on the upstream communications path 2015 may be measured. In another example embodiment, a no-input signal power level may be measured on an upstream communications path at a time there are no upstream communications for least a portion of at least one timeslot in an upstream communications frame.

In contrast to the previous example, the non-communicating ONT 2013 may send an "idle" signal (not shown) or a message indicating there is no data to be sent (not shown). In this situation a no-input signal power level on the upstream communications path 2015 cannot be measured. FIG. 21A is an example embodiment of the invention in which an upstream communications frame 2105 has n number of timeslots 21 10a-n. Each timeslot 21 10a-n grants (or allocates) a time for upstream communications 21 15 (referred to herein as t , ). It is during the t_s|M 2115 that upstream communications data is communicated from an ONT to an OLT. In the upstream communications frame 2105, an "unused" timeslot (i.e., a timeslo_.t without upstream communications data) defines a time for no-upstream communications 2120 (referred to herein as t _u|M). It is during the t _uici 2120 that a no-input signal power level on an upstream communications path may be measured. An unused timeslot such as t _mei2120 may occur in networks with more timeslots than ONTs.

In this example embodiment, the t 2120 is equal to the I ₁ 2115. As such, if the t _]ot is 1.2 μs, for example, the no-input signal power level on an upstream communications path may be measured for as long as 1 .2 μs.

FIG. 21B is another example embodiment illustrating a time for no-upstream communications 2120 (referred to herein as t ,) optionally equal to some whole multiple of a time for upstream communications 21 15 (referred to herein as t_slol). For example, if the t_s)oι 2115 is 1.2 μs, the t _t 2120 may be two, three, etc., times the length of the t_s|o( 2115. Accordingly, a no-input signal power level on an upstream communications path is measured for 2.4μs, 3.6μs, etc., where the longer time typically results in improved accuracy of the power level measurement. FIG. 21 C is yet another example embodiment in which a time for no- upstream communications 2120 (referred to herein as t _met) is equal to some fraction of a time for upstream communications 21 15 (referred to herein as t ). For example, if the t_s]_ot 21 15 is 1.2 μs, the t _uie| 2120 may be a quarter, one and half, etc. times the length of the t_s|ot 2115. Accordingly, a no-input signal power level on an upstream communications path may be measured for 0.3 μs, l.δμs, etc.

In still yet other example embodiment, a no-input signal power level on an upstream communications path may be measured during a time there are no upstream communications (e.g. , t _uιel 2120 or when no communications frames are communicated in an upstream direction) and then averaged, resulting in an averaged measurement, to increase noise immunity. By measuring a no-input signal power level on an upstream communications path at a time there are no upstream communications, an error condition of very small optical power levels can be detected. Having detected such an error condition, a determination may be made as to whether the error condition may lead to layer 2 communications errors, such as errors in the ranging or normalization parameters.

FIG. 22 illustrates a ratio between a one-bit input signal power level 2205 and a zero-bit input signal power level 2210. This ratio is referred to herein as an extinction ratio 2215. The extinction ratio 2215 is a measure of a contrast (or a distinction) between power levels of input signals designating a one-bit input signal and a zero-bit input signal. For example, if the extinction ratio 2215 is large, the distinction between a one-bit input signal power level and a zero-bit input signal power level is also large. Because the distinction between the power levels is large, an optical receiver has an easier task in detecting an input signal as either a one-bit input signal or a zero-bit input signal. In contrast, if the extinction ratio 2215 is small, the distinction between a one-bit input signal power level and a zero-bit input signal power level is also small, and an optical receiver has a more difficult task in detecting an input signal as either a one-bit input signal or a zero-bit input signal.

A similar ratio may be said to exist between the zero-bit input signal power level 2210 and a no-input signal power level 2220. This ratio is referred to herein as a no-input extinction ratio 2225. Like the extinction ratio 2215, the no-input extinction ratio 2225 is a measure of a contrast (or a distinction) between a power level of an input signal designating a zero-bit input signal and a power level of a no- input signal. For example, if the no-input extinction ratio 2225 is large, the distinction between a zero-bit input signal power level and a no-input signal power level is also large. Because the distinction between power levels is large, an optical receiver has an easier task in detecting a zero-bit input signal or a no-input signal. In contrast, if the no-input extinction ratio 2225 is small, the distinction a zero-bit input signal power level and a no-input signal power level is also small, and an optical receiver has a more difficult task in detecting a zero-bit input signal or a no-input signal.

Difficulties in distinguishing between a no-input signal and a zero-bit input signal may also lead to difficulties in distinguishing between a one-bit input signal and a zero-bit input signal. As a consequence, there may be an increase in the number of bit errors which occur during normal communications. As such, it desirable to have a no-input extinction ratio which is sufficiently large enough to prevent such bit errors. FIG. 23 A is a power level diagram illustrating a no-input signal 2305 which has a power level at time \_mUai 2310 equal to a power level at time t_fmaI 2315. The power level of the no-input signal 2305 (i.e., no-input signal power level) may be integrated (or added) by an integrator 2320 (or other electronics) in an optical power receiver (or transceiver) to produce an integrated no-input signal power level 2325. The integrator 2320 integrates from time t_jnjtja] to time t_fma|, resulting in an integrated no-input signal power level at t_fina| 2330 being greater than an integrated no-input signal power level at t_injtial 2335, as is expected. The longer the period of integration time, the higher the integrated no-input signal power level 2325 is ramped (or increased). Consequently, over time, a no-input extinction ratio (see FIG. 22) becomes smaller, and it is more difficult to distinguish a no-input signal from a zero- bit input signal. Further, the higher the integrated no-input signal power level at i_nitial 2335, the more significant the resulting integrated no-input signal power level

2325 becomes over time and the smaller a no-input extinction ratio becomes over the same time.^"

FIG. 23B is a diagram illustrating how a transmitted optical power level from a faulty ONT affects measurement during ranging of an ONT by an OLT. A message diagram 2300a illustrates an exchange of ranging messages between an OLT 2301 and an ONT 2302 during a ranging window 2355. A transmitted power level versus time plot 2300b illustrates the ONT 2302 transmitting a no-input signal power level 2303 during the ranging window 2355. A received power level versus time plot 2300c illustrates the OLT 2301 receiving the no-input signal power level 2303, which has been integrated by an integrator 2304 in a receiver (not shown) of the OLT 2301, as an integrated no-input signal power level 2345.

The transmitted power level versus time plot 2300b indicates that the no- input signal power level 2303 may be constant during the ranging window 2355, • where the constant level may be a normal low level (e.g., -4OdBm) or a faulty high level (e.g.. between -3OdBm and -25dBm, or higher). The integrated no-input signal power level 2345 ramps up from an integrated no-input signal power level at time i_nitial 2340 to an integrated no-input signal power level at time t_f|nal 2350 over the ranging window 2355.

In operation, while the no-input signal power level 2303 is being integrated over the ranging window 2355, the OLT 2301 sends a ranging request 2360 to the ONT 2302. The ONT 2302. in turn, responds with a ranging response 2365. The OLT 2301 , having sent the ranging request 2360, receives the ranging response 2365 from the ONT 2302 during the ranging window 2355 or it reports a ranging error.

Typically, the receiver of the OLT 2301 is reset between adjacent upstream timeslots to accommodate power levels which vary from ONT to ONT. During ONT ranging, however, an upstream timeslot is effectively enlarged to accommodate variability in supported fiber lengths, i.e., more than one timeslot is used for the ranging window 2355. For example, the ONT 2302 may be located up to 20 kilometers away from the OLT 2301. To accommodate this distance, the duration of the ranging window 2355 is set sufficiently long enough to allow the ONT 2302 located 20 kilometers away from the OLT 2301 to receive the ranging request 2360 and the OLT 2301 to receive the ranging response 2365.

When the duration of the ranging window 2355 is set for a long period of time, the receiver of the OLT 2301 is not reset during this period of time. As a result, no-input signal power levels from non-transmitting ONTs on the ODN have more time to be integrated by the receiver of the OLT 2301 , thus increasing the integrated no-input signal power level 2345. This increase has a negative impact on a signal condition circuitry in the receiver of the OLT 2301. In other words, the longer the duration of the ranging window 2355, the greater the effects of a small no-input extinction ratio (see FIG. 22). Consequently, it may be difficult to distinguish between a zero-bit input signal power level and a one-bit input signal power level possibly leading to upstream communications problem(s).

In one embodiment of the present invention, prior to ranging an ONT, an OLT instructs communicating ONTs to halt upstream communications. Despite upstream communications being halted, there still may be a no-input signal from one or more halted ONTs causing a "faulty no-input signal power level" (see FIG. 19). Consequently, the faulty no-input signal power level may be integrated, causing the integrated no-input signal power level 2345 to increase further.

FIG. 24A is a block diagram of an example OLT 2405 in communication with an ONT 2410. In this particular example, the OLT 2405 has a PON card 2415. The PON card 2415 includes a processor 2420 communicatively coupled to a receiver 2425 and a transmitter 2430. Alternatively, the receiver 2425 and the transmitter 2430 may be integrated into a single transceiver (not shown). In the direction toward from the OLT 2405, the receiver 2425 (or transceiver) receives upstream communications 2435. The processor 2420 subsequently processes the upstream communications 2435. In the opposite direction toward the ONT 2410, the processor 2420 sends, via the transmitter 2430 (or transceiver), downstream communications 2440.

FIG. 24B is a block diagram which illustrates an example processor 2445, supporting example embodiments of the invention, operating in a PON card of an OLT. The processor 2445 may include a measurement unit 2450, a comparison unit 2455, and a notification generator 2460. Alternatively, some or all of the aforementioned components may not be co-located with the processor 2445, but may be remotely located connected via a communications bus (not shown).

In operation of this example embodiment, the measurement unit 2450 may measure a power level of a no-input signal 2401 on an upstream communications path. The measurement unit 2450 may include an integrator, such as the integrator 2320 of FIG. 23 A, or other electronics to measure the power level of the no-input signal 2401. A measured no-input signal power level 2402 may be compared against a threshold value 2403 by the comparison unit 2455. A result 2404 from the comparison unit 2455 is communicated to the notification generator 2460. The notification generator 2460 may generate a notification if the communicated result 2404 indicates the measured no-input signal power level 2402 exceeds the threshold 2403. Keeping the integrated no-input signal power levels of FIGS. 23 A and 23 B in mind, it should be understood that the comparison unit 2455 may compare a maximum, an average (at multiple times or over a length of time), or a portion of the measured no-input signal power level 2402 against the threshold 2403. The threshold 2403 against which the measured no-input signal power level

2402 is compared may be determined or defined in multiple ways. For example, the threshold 2403 may be set to a value equal to a "tolerable no-input signal power level" multiplied by a number of ONTs in communication with the OLT. Field experience may indicate a no-input signal power level of -2OdBm to -3OdBm per ONT often leads to problems in upstream communications. Based on such experience, the tolerable no-input signal power level may be -4OdBm. Therefore, in an example network having thirty-two ONTs communicating with an OLT, the threshold may be calculated as -4OdBm multiplied by thirty-two. Additionally, losses between the ONTs and the OLT (i.e., ODN losses) may be accounted for in calculating the threshold. In another example embodiment, the tolerable no-input signal power level may be less than a zero-bit input signal power level specified for the ONTs. One skilled in the art will readily appreciate that the value of the tolerable no-input signal power level may not be fixed (i.e., set to the same level for all communications networks, but rather may depend on characteristics of a communications network.

The threshold 2403 may alternatively represent a maximum power level corresponding to a fault associated with upstream communications in a non- communicating state. In another example embodiment, the threshold 2403 may be less than a sum of a zero-bit input signal power level of each ONT offset by respective losses between the ONTs and the OLT. It should be understood that the threshold 2403 may be predetermined based on a configuration of a passive optical network or determined based on some other metric.

Continuing to refer to FIG. 24B, the notification generator 2460 may generate a remote notification 2465 which is sent over a network 2466 to, for example, a remote user or remote management system 2467. Alternatively, the notification generator 2460 may generate a local notification 2470, which is presented locally to, for example, a local user or local management system 2471. It should be understood that the remote notification 2465 may be any form of signal (e.g., analog, digital, packet, and so forth), data values, including in header or load portions of packets, and so forth. The local notification 2470 may also be any form of signal or may be audio or visual alarms to alert an operator at a console at the OLT that an error as described herein had occurred.

FIG. 25A is a flow diagram illustrating an example process 2500 for diagnosing a problem on an ODN. A no-input signal power level on an upstream communications path may be measured (2505) at a time no upstream communications are on the upstream communications path. The measured no-input signal power level may be compared (2510) against a threshold. If the measured no- input signal power level on the upstream communications path is greater than the threshold, a notification may be issued (2515) to alert an operator (or management system) that the threshold is exceeded. If, however, the measured no-input signal power level on the upstream communications path is not greater than the threshold, the process 2500 may return to begin measuring (2505) the no-input signal power level. FIG. 25B is a flow diagram illustrating a process 2520 for diagnosing a problem on an ODN in accordance with an example embodiment of the invention. A no-input signal power level on an upstream communications path may be measured (2525) at a time no upstream communications are on the upstream communications path. In this example embodiment, the no-input signal power level is measured during a time for no upstream communications (t _u)c(). In reference to

FIGS. 21A-21C, the time for no upstream communications (t _uie() may be equal to a time for upstream communications (t_slot). Alternatively, the time for no upstream communications (t _uiet) may be equal to a whole multiple or fraction of the time for upstream communications (t_slol). Next, a threshold may be calculated (2530). In this example embodiment, the threshold is equal to a number of ONTs on the ODN multiplied by a tolerable no-input signal power level. The tolerable no-input signal power level may be estimated based on system modeling, equal to a value measured at a time known not be experiencing an error condition (e.g., initial system set-up), and so forth. The measured no-input signal power level on the upstream communications path may be compared (2535) against the calculated threshold. If the measured no-input signal power level is greater than the calculated threshold, a notification may be issued (2540) that the calculated threshold is exceeded. If. however, the measured no-input signal power level on the upstream communications path is less than the calculated threshold, the process 2520 may wait (2545) for the time for no upstream communications (t ) to reoccur. After waiting, the process 2520 may once again measure (2525) the no-input signal power level on the upstream communications path.

The above description referring to FIGS. 18-25B describes diagnosing problems on a time division multiple access (TDMA) optical distribution network

(ODN), such as a passive optical network (PON). An example method may include: (i) measuring no-input signal power level on a communications path configured to carry upstream communications between multiple optical network terminals (ONTs) and an optical line terminal (OLT) in a passive optical network (PON) at a time no upstream communications are on the communications path from the ONTs to the OLT; (ii) comparing the measured no-input signal power level to a threshold; and (iii) generating a notification in an event the threshold is exceeded. Through the use of this method, faults in optical transmitters, such as bad solder joints, can be determined. Such faults may cause errors in parameters, such as ranging or normalization parameters, associated with communications. By determining the faults, the time required to resolve communications errors can be reduced.

FIGS. 26-42 illustrate example embodiments of an aspect of the present invention in which a transport layer ranging window has a longer duration than a physical layer ranging window. The transport layer ranging window defines a range within which an optical network terminal (ONT) can respond to a ranging request without affecting upstream communications from other ONTs on a passive optical network (PON), and the physical layer ranging response defines a time within which a receiver in a optical line terminal (OLT) is enabled to receive a ranging response from the ONT. By keeping the transport layer ranging window sufficiently long in duration, ranging responses from ONTs at unspecified ranges can be captured. By shortening the physical layer ranging window, errors due to noise or faulty ONT output power can be reduced.

As previously described, diagnosing a passive optical network (PON) for problems may involve detecting, prior to establishing layer 2 communications, a malfunctioning optical network terminal (ONT). A malfunctioning ONT may be detected by looking for an inappropriate presence of a modulated or unmodulated upstream optical signal when no signal should be present on the upstream communications path. The inappropriate presence of such signals may cause a power level associated with these signals (i.e., a no-input signal power level) to be' integrated over time by an integrator in a receiver to produce an integrated no-input signal power level. As expected, over time the integrated no-input signal power level increases, causing a no-input extinction ratio to become smaller. Consequently, it becomes more difficult to distinguish a no-input signal from a zero- bit input signal, possibly leading to bit errors. In less severe cases, a higher than expected no-input signal power level may result in erroneous settings of parameters used in connection with upstream communications.

The effect of integrating a no-input signal power level is particularly significant when ranging an ONT. While ranging, an integrator (or other electronics) in an optical line terminal (OLT) receiver (or transceiver) may integrate (or otherwise calculate) a no-input signal power level for an extended period of time. Accordingly, what is needed is a method or a corresponding apparatus for ranging an ONT in a passive optical network in a manner minimizing the aforementioned effects caused by the inappropriate presence of an unmodulated or modulated optical signal on the upstream communications path or other times when such presence causes adverse effects, directly or indirectly on upstream communications. It should be understood that alternative embodiments may be employed in situations involving downstream communications. International Telecommunication Union (ITU) specification 983.1 , Section

8.4.2.5.2, describes shortening a ranging window when the location of an ONT to be ranged is known. With a priori knowledge, a ranging window may be shortened to correspond to a known distance between the OLT and the ONT. According to shortening the ranging window done in prior art systems by reducing a transport layer ranging window (layer 2) and a physical layer ranging window (layer 1 ) in equal amounts.

By shortening a ranging window, disruption to communicating ONTs is minimized. Since communicating ONTs are disabled from communicating in the upstream direction during ranging, the shorter the ranging window, the shorter the amount of time upstream communications must be halted. Consequently, the negative impact of ranging on the throughput of communicating ONTs is lessen by using a shortened ranging window.

In contrast, when the location of the ONT is unknown, it possible the ONT is located at a possible maximum distance (e.g., 20 Km) away from the OLT. As such, to accommodate this maximum distance, a maximal ranging window must used. FIG. 26 is a message diagram illustrating an OLT 2605 ranging an ONT 2610. To range the ONT 2610, the OLT 2605 transmits a ranging request 2615. The ONT 2610, in response to the transmitted ranging request 2615, transmits a ranging response 2620. The OLT 2605, having received the ranging response 2620, determines a metric associated with the ranging response 2620 for use in connection with upstream communications between the OLT 2605 and the ONT 2610. For example, a round-trip time 2625 may be determined, where the determined round- trip time 2625 represents a time from when the OLT 2605 transmits the ranging request 2615 to the time the OLT 2605 receives the ranging response 2620. It should be understood that ranging cycles may be calculated other ways, such as a one-way trip time of a ranging request 2615 or a ranging response 2620. In one embodiment, the OLT 2605 sets at least one parameter, used in connection with upstream communications between the OLT 2605 and the ONT 2610, based on at least one metric associated with the ranging response 2620. For example, the OLT 2605, based on the round-trip time 2625, may set an equalization delay 2630. The OLT 2605 may then send the ONT 2610 the equalization delay 2630 or command the ONT 2610 to set an internal parameter based on the equalization delay 2630. In an example embodiment, the equalization delay 2630 is conveyed via a message 2635. During later communications with the OLT 2605 in this example embodiment, the ONT 2610, in turn, waits for a time according to the equalization delay 2630 before sending upstream communications data 2640. In one embodiment, the ONT 2610 uses the equalization delay 2630 to have the upstream communications data 2640 reach in the OLT 2605 during a predefined timeslot relative to upstream communications data from other ONTs (not shown), as known in the art.

Previously described in reference to FIGS. 2OB and 23B, a ranging window is further described in FIGS. 27 A and 27B.

FIG. 27A illustrates, in connection with a transport layer ranging window 2705, an OLT 2715 transmitting a ranging request 2720 and an ONT 2725, among a group of ONTs 271 Oa-n, transmitting a ranging response 2730. During the transport layer ranging window 2705, upstream communications from communicating ONTs 271 Oa-n are halted (or set in a "quiet" state).

In FIG. 27A, the transport layer ranging window 2705 starts with the OLT 2715 transmitting a last-bit 2735 of the ranging request 2720. Presented differently, the ranging request 2720 is transmitted before the transport layer ranging window 2705 begins. Alternatively, referring to FIG. 27B, a transport layer ranging window 2755 starts with an OLT 2765 transmitting a first-bit 2785 of a ranging request 2770, so the ranging request 2770 and ranging response 2780, are transmitted during the transport layer ranging window 2755. As such, the duration of the transport layer ranging window 2705 of FIG. 27A may be shorter than the duration of the transport layer ranging window 2755 of FIG. 27B. Consequently, upstream communications from ONTs 2710a-n of FIG. 27A may be halted for a shorter period of time than upstream communications from ONTs 2760a-n of FIG. 27B. In other embodiments, the transport layer ranging window 2705 and 2755 of FIGS. 27 A and 27B, respectively, are the same duration or the transport layer ranging window 2755 of FIG. 27B is shorter than the one of FIG. 27 A.

For the remainder of this disclosure, a ranging request is described as being transmitted during a transport layer ranging window unless otherwise specified. It is noted, however, that example embodiments of the invention are not limited to a transport layer ranging window starting with a first bit of a ranging request being transmitted. Example embodiments of the invention are also applicable to a transport layer ranging window starting with transmission of a last-bit of a ranging request. As described previously in reference to FIG. 23B, an integrated no-input signal power level may ramp (or increase with time) over a transport layer ranging window. Due to the duration of the transport layer ranging window and ramping the integrated no-input signal power level over this duration, a no-input extinction ratio (see FIG. 22) may be small. In such a case, it may be difficult to distinguish a no- input signal from a zero-bit input signal, possibly leading to upstream communications problems. To minimize this potential source of communication errors, example embodiments of the invention monitor for a ranging response during a portion of the transport layer ranging window rather than during the entire transport layer ranging window. FIG. 28 illustrates a transport layer ranging window 2805 having, for example, a duration of 100 μs (microseconds). A physical layer ranging window 2810 within the transport layer ranging window 2805 may have a duration that is based, in part, on a duration of an expected ranging response 2815. For example, the duration of the physical layer ranging window 2810 may be twice the duration of the ranging response 2815. As such, the duration of the physical layer ranging window 2810 is 10 μs if the duration of the ranging response 2815 is 5 μs. In another example, the duration of the physical layer ranging window 2810 may be some multiple of the duration of the ranging response 2815 plus some time for a delimiter or other overhead (not shown) associated with transmitting the ranging response 2815.

FIG. 29A illustrates a transmitted power level versus time plot 2900a in which, during a transport layer ranging window 2901 , an ONT (not shown) transmits a no-input signal power level 2905. A received power level versus time plot 2900b further illustrates, during the transport layer ranging window 2901 , an OLT (not shown) receiving the transmitted no-input signal power level 2905.

The transmitted power level versus time plot 2900a indicates the transmitted no-input signal power level 2905 may be constant during the transport layer ranging window 2901. The constant level may be a normal no-input level (e.g., less than - 4OdBm) or a faulty low-level (e.g., between -3OdBm and -25dBm. or higher).

The received power level versus time plot 2900b illustrates the duration of the transport layer ranging window 2901 as being from t_ιmtιa| to t_f,_naι, and the duration of a physical layer ranging window 2902 as being from ti to t₂. The duration of the transport layer ranging window 2901 is greater than the duration of the physical layer ranging window 2902, i.e., the time from t_inιtiai to t_f,_naι is greater than the time from ti to t₂.

In general, the effect of any noise on the receiver increases the longer the physical layer ranging window 2902 is open and decreases the shorter the physical layer ranging window is opened. For purposes of illustrating the effects of noise in a hardware sense, examples in terms of an integrator integrating noise are presented herein, including immediately below. However, the example is not intended to be restrictive in any way. During the physical layer ranging window 2902 {i.e. , from ti to tV), monitoring for ranging response may be enabled. While the monitoring is enabled, a ranging response received during the physical layer ranging window 2902 may be processed. Additionally, while the monitoring is enabled, the transmitted no-input signal power level 2905 is received and integrated by an integrator 2906a (or other electronics) in a receiver (or transceiver) of the OLT. Consequently, a power level measured from Ti to T₂ increases over time (or ramps) due to integration. This power level, which may be measured while monitoring is enabled, is referred to herein as an integrated power level associated with monitoring for a ranging response (e.g., 2920 and 2935).

In contrast, during a first disabled period 2910a (i.e.. from tj_ni_tj_aι to ti) or 2910b (i.e., from t₂ to tf,_nai), monitoring for a ranging response may be disabled. While the monitoring is disabled, a ranging response received may not be processed. Additionally, while monitoring is disabled, the transmitted no-input signal power level 2905 is received, but may not be integrated by the integrator 2906a. Consequently, power levels measured from tj_nj_tj_a| to tj and from t? to t_f,_naι remain substantially unchanged (e.g., 2917a and 2917b). . At ti_bi_al, the transmitted no-input signal power level 2905 is received by the

OLT at an initial-power level 2915, which is about the no-input power level output by the ONT, less transmission or other losses. Also at tj_ni_ti_ai, the integrator 2906a is reset by a reset command 2907a or other mechanism. During the first disabled period 2910a, the transmitted no-input signal power level 2905 received by OLT is not integrated. As such, the transmitted no-input signal power level 2905 received by the OLT between ti_bi_al and ti remains non-integrated from the initial-power level 2915. .

At t|, the transmitted no-input signal power level 2905 is received by the OLT at a first-power level 2925. Since the transmitted no-input signal power level 2905 is not integrated during the first disabled period 2910a, the initial-power level 2915 and the first-power level 2925 are substantially equal. During the physical layer ranging window 2902, however, the transmitted no-input signal power level 2905 received by OLT is integrated. As such, an integrated power level associated with monitoring for a ranging response 2920 ramps from the first-power level 2925 at ti to a second-power level 2930 at t₂.

At t₂, the transmitted no-input signal power level 2905 received by the QLT at the second-power level 2930. During the second disabled period 2910b, the transmitted no-input signal power level 2905 received by the OLT is not integrated. As such, the transmitted no-input signal power level 2905 received by the OLT from t₂ to t_fi_nai remains substantially unchanged from the second-power level 2930.

In comparison, if monitoring during the transport layer ranging window 2901 {i.e. , from t_initai to tf_inai) is enabled, an integrated power level associated with monitoring for a ranging response 2935 (represented by a dashed line) ramps from the initial-power level 2915 at t,_nitιaι to a final-power level 2940 at t_fina|. Since the duration of the transport layer ranging window 2901 is longer than the duration of the physical layer ranging window 2902, there is more time for the integrated power level to increase. Consequently, the measured second-power level 2930, at the end of the physical layer ranging window 2902, is less than the final-power level 2940 that would have been measured at the end of the transport layer ranging window 2901 if the physical layer ranging window 2902 was the same length as the transport layer ranging window 2901. Accordingly, the above described consequences of having a small no-input extinction ratio may be minimized by enabling monitoring for a ranging response during a physical layer ranging window rather than during an entire transport layer ranging window.

Alternatively, in FIG. 29B, in a received power level versus time plot 2900c, during a first disabled period 2960a (from t_iπmai to ti) or 2960b (from t₂ to tr,_nai), monitoring for ranging response may be disabled in a different manner as compared to FIG. 29A. In the embodiment of FIG. 29B, while the monitoring is disabled, a ranging response received may not be processed (e.g., by hardware, firmware, or software), but the transmitted no-input signal power level 2905 may be integrated by an integrator 2906b. Consequently, power levels measured from t,_mtiai to ti and from t₂ to to_nal increase over time (or ramp). These power levels, which may be measured while monitoring is disabled, are referred to herein as integrated power levels (e.g., 2963 and 2978), in comparison to the integrated power level associated with monitoring for a ranging response (2920 and 2935) for FIG. 29A.

At t_mui_ai, the transmitted no-input signal power level 2905 is received by an OLT at an initial-power level 2965, which is about the no-input power level output by an ONT, less transmission or other losses. During the first disabled period 2960a, the transmitted no-input signal power level 2905 received by the OLT is integrated. As such, beginning at t_muιa], an integrated power level 2963 ramps from the initial-power level 2965 to a first-power level 2970 at t_\.

At T|, the integrator 2906b is reset by a reset command 2907b. Resetting the integrator 2906b resets the integrated power level 2963 from the first-power level 2970 to a reset power level 2975. During the physical layer ranging window 2902, the transmitted no-input signal power level 2905 received by the OLT is integrated. As such, beginning at ti, an integrated power level 2973 associated with monitoring for a ranging response ramps from the reset power level 2975 to a second-power level 2980 at t₂. At t₂, the transmitted no-input signal power level 2905 is received by the

OLT at the second-power level 2980. During the second disabled period 2960b, the transmitted no-input signal power level 2905 received by the OLT is integrated. As such, beginning at t₂, an integrated power level 2978 ramps from the second-power level 2980 to a final-power level 2985 at t_finai. An ONT may be located up to some distance away from an OLT, for example 20 Km. To accommodate such distance, the duration of a transport layer ranging window is set sufficiently long enough to allow the ONT to receive a ranging request, within which an ONT can respond to a ranging request without affecting upstream communications from other ONTs on the ODN, and the OLT to receive a ranging response. As such, the ranging request may be located in time anywhere within the transport layer ranging window. Consequently, the issue is what portion of the transport layer ranging window to monitor for (or to otherwise locate), in time, a ranging response. One approach may be to repeatedly transmit a ranging request and monitor for a ranging response, where physical layer ranging window(s) is/are located in the transport layer ranging window at different location(s) each cycle until the location, in time, of the ranging response is found within the transport layer ranging window.

FIG. 30 is a series of timing diagrams illustrating dynamically adjusting a physical layer ranging window in an iterative manner within a transport layer ranging window 3001 to locate a ranging response 301 Oa-c. By way of example,

FIG. 30 illustrates a binary search. One skilled in the art will readily recognize other types of searches are equally applicable, for example, a search based on a hash algorithm. Furthermore, one skilled in the art will readily recognize a transport layer ranging may be approximated through dynamic adjustment of the duration, delay, number, or combination thereof, of a physical layer ranging window. Hardware, firmware, or software may be employed to support or execute the search as understood in the art.

The transport layer ranging window 3001 may be approximated by a first-half physical layer ranging window 3015 and a second-half physical layer ranging window 3016. In a first iteration 3003a, a ranging request 3005a is transmitted, but a ranging response 3010a is not received during the first-half physical layer ranging window 3015. In a second iteration 3003b, a ranging request 3005b is transmitted, ; and a ranging response 3010b is received during the second-half physical layer ranging window 3016. Accordingly, the ranging response 3010b is located, in time, during a second-half of the transport layer ranging window 3002.

To locate a ranging response in time with more accuracy, the second-half of the transport layer ranging window 3002 may be approximated by a first-quarter physical layer ranging window 3020 and a second-quarter physical layer ranging window (not shown). In a third iteration 3003c, a ranging request 3005c is transmitted, and a ranging response 3010c is received during the first-quarter physical layer ranging window 3020. Accordingly, the ranging response 301 Oc is located, in time, during a first-quarter of the second-half of the transport layer ranging window 3002. Presently differently, the ranging response 3010c is located, in time, during a third-quarter of the transport layer ranging window 3001.

It should be understood that the example illustrated in FIG. 30 is a simplified example. In practice, hundreds or thousands of attempts to locate a ranging response 301 Oa-c may be performed.

One skilled in the art will readily recognize the transport layer ranging window 3001 may be even further divided to locate a ranging response, in time, with more accuracy. The number of times a transport layer ranging window is divided in order to locate a ranging response, in time, may depend on the duration of the ranging response. For example, to locate a ranging response of 5 μs, a transport layer ranging window of 100 μs may be divided up to sixteen times to locate the ranging response, in time. In addition to dynamically adjusting the physical layer ranging window within the transport layer ranging window, a transport layer ranging window may also be approximated by shifting one or more physical layer ranging windows.

FIG. 31 is a series of timing diagrams illustrating another example of searching for a ranging response by dynamically adjusting a position, in time, of the physical layer ranging window within a transport layer ranging window. In FIG. 31 , in a first iteration 3103a, during or otherwise in connection with a transport layer ranging window 31 10a, a ranging request 31 15a is transmitted, but a ranging response 3120a is not received during a physical layer ranging window 3125a. In a second iteration 3103b, during a transport layer ranging window 31 1 Ob, a ranging request 31 15b is transmitted. A physical layer ranging window 3125b is shifted, in time, with respect to the first iteration physical layer ranging window 3125a, but a ranging response 3120a remains not received during the shifted physical layer ranging window 3125b. In an nth iteration 3103«, during a transport layer ranging window 31 10«, a ranging request 31 15« is transmitted. A physical layer ranging window 3125« is shifted, in time, with respect to previous physical layer ranging windows. In this nth iteration 3103«, a ranging response 3120« is received during the shifted physical layer ranging window 3125«.

Having found the ranging response 3120« during the physical layer ranging window 3125«, transmitting^' a ranging request, monitoring for a ranging response, and shifting a physical layer ranging window may or may not repeat. In one example embodiment, the transmitting, monitoring, and shifting repeat at least until a ranging response is received during a physical layer ranging window. In another example embodiment, the transmitting, monitoring, and shifting repeat for a fixed, variable or otherwise predetermined number of repetitions. In addition to shifting a physical layer ranging window non-incremental Iy within a transport layer ranging window, a physical layer ranging window may be shifted incrementally across the transport layer ranging window.

In both FIGS. 30 and 31 , the physical layer ranging window is set slightly longer in duration than the expected duration of a ranging response to keep a metric, calculated by integrating a no-input signal power level, to an acceptable error level, where the acceptable error level is one within which parameter(s) based upon the metric and used for upstream communications during normal operations do not adversely affect the upstream communications.

FIG. 32 A is a series of timing diagrams illustrating a search technique in which a physical layer ranging window is shifted across a transport layer ranging window in equal steps. In FIG. 32A, in a first iteration 3201a, during a transport layer ranging window 3203a, a ranging request 3205a is transmitted, but a ranging response 3215a is not received during a physical layer ranging window 3210a. In a second iteration 3201b, a ranging request 3205b is transmitted, and a physical layer ranging window 3210b is shifted, in time, relative to the previous physical layer ranging window 321Oa, across a transport layer ranging window 3203b by a shift increment 321 1. A ranging response 3215b is not received during the shifted physical layer ranging window 3210b. In a third iteration 3201 c, a ranging request 3205c is transmitted, and a physical layer ranging window 3210c is again shifted, in time, relative to the previous physical layer ranging window 3210b, across a transport layer ranging window 3203c by the shift increment 321 1. Again, a ranging response 3215c is not received during the shifted physical layer ranging window 3210c.

In an wth iteration 3201«, a ranging request 3205« is transmitted in a transport layer ranging window 3203«, and a ranging response 3215« is received during a physical layer ranging window 3210« shifted, in time, relative to a previous («-l)th physical layer ranging window (not shown) by the shift increment 3211.

In this embodiment, the shift increment 321 1 shifts the physical layer ranging window 3210a-n, in time, by an amount equal to some whole number multiple of the duration of the physical layer ranging window 3210a-n. For example, a physical layer ranging window of 10 μs may be shifted, in time, incrementally by 10 μs, 20 μs, 30 μs, etc. across the transport layer ranging window. FIG. 32B is a series of timing diagrams which collectively illustrate another example embodiment of searching for a ranging response within a transport layer ranging window. A physical layer ranging window 3240a-n is shifted, in time, by a shift increment 3241. The shift increment 3241 is some fraction of the physical layer ranging window 3240a-n. For example, a physical layer ranging window of 10 μs may be shifted, in time, incrementally by 5 μs, 15 μs, 25 μs, etc. across the transport layer ranging window 3233a-n. From a first iteration through an nth iteration 3231a-n, a ranging request 3235a-n is transmitted, and a physical layer ranging window 3240a-n is shifted by the shifting increment 3241 at least until a ranging response 3245a-n is received during a physical layer ranging window, which occurs in this example during the nth physical layer ranging window 324On.

In some cases, the physical layer ranging window (PLRW) shift techniques of FIGS. 32A and 32B may not result in successful ranging should a ranging response being partially aligned with the physical layer ranging window. One way of preventing this is to overlap any two adjacent ranging windows by an amount greater than or equal to the ranging response to ensure that if the tail end of the ranging response is just missed (e.g., by one PLRW), the very beginning of the ranging response is not missed by the next PLRW.

In general if the PLRW is Y times the duration ranging response (RR) (i.e., PLRW - Y x RR), the PLRW can be shifted no more that RRJ(Y -1) in order to guarantee that, if the very last bit of the ranging response is truncated by the current position of the PLRW, the next shifted PLRW does not truncate the very first part of the ranging response.

For example, in reference to FIG. 32A, if the duration of the PLRW is equal to two times (2x) the duration of the ranging response, there is a shift of less than RR of the PLRW. In the FIG. 32B example, if the duration of the PLRW is equal to one and one-half times (1.5x) the duration of the ranging response, there is a shift of less than one-half (0.5) of the RR to a subsequent PLRW.

FIG. 32C is a series of timing diagrams which collectively illustrate another example embodiment of searching for a ranging response within a transport layer ranging window. A physical layer ranging window 3270a-n is shifted, in time, by a variable shift increment 3272. The variable shift increment 3272 shifts the physical layer ranging window 3270a-n, in time, by some amount. The amount shifted may be random or pseudo-random. Optionally, the physical layer ranging window 3270a-n may be shifted, in time, by an amount according a geometric series, a logarithmic series, or other series. As such, from a first iteration through an nth iteration 3271a-n, a ranging request 3255a-n is transmitted, and the physical layer ranging window 3270a-n is shifted in transport layer ranging windows 3273a-n by the variable shifting increment 3272 at least until a ranging response 3275a-n is received during a physical ranging window, which occurs in this example in the nth physical layer ranging window 3270«.

FIG. 33 is a series of timing windows which illustrate an example technique of adjusting timing of a physical layer ranging window in an event only part of a ranging response is received during a physical layer ranging window. In an (n-l)th iteration 330In-I. during a transport layer ranging window 3305a, a ranging request 3310a is transmitted. During the transport layer ranging window 3305a, a ranging response 3315, in part, is received during a physical layer ranging window 332Oa. The portion of the ranging response 3315 received during the physical layer ranging window 3320a is referred to herein as a received portion 3325, while a remaining portion not received is referred to herein as a non-received portion 3330.

In an «th iteration 3301«, during a later transport layer ranging window 3305b, a ranging request 331Ob is transmitted, and a physical layer ranging window 3320b is shifted, in time. The physical layer ranging window 3320b is shifted, in time, by an amount expected to result in receiving a ranging response 3316 in full during the physical layer ranging window 3320b. For example, the physical layer ranging window 3320b may be shifted, in time, relative to the («-l)th physical layer ranging window 3320a, by an amount equal to the non-received portion 3330. Alternatively, the physical layer ranging window 3320b may be shifted, in time, by an amount greater than the non-received portion 3330. In addition to shifting, in time, the physical layer ranging window, in another embodiment, the duration of a physical layer ranging window may be lengthened, after a portion of the ranging response is received, by an amount expected to allow the ranging response to be received during the physical layer ranging window.

FIG. 34 is a series of timing diagrams which illustrate a search technique for monitoring for a ranging response by adjusting a length of a physical layer ranging window in a dynamic manner. In a first iteration 3401a, during a transport layer ranging window 3405a, a ranging request 3410a is transmitted. During the transport layer ranging window 3405a, a ranging response 3415a is not received during a physical layer ranging window 3420a. In this embodiment, the duration of the physical layer ranging window 3420a is lengthened at least until a ranging response is received during the lengthened physical layer ranging window.

In an nth iteration 3401«, during a transport layer ranging window 3405«, a ranging request 3410« is transmitted, and a physical layer ranging window 3420« is shown in a lengthened state relative to the length of the physical layer ranging window 3420a of the first iteration 3401a. During the transport layer ranging window 3405«, a ranging response 3415« is received during the lengthened physical layer ranging window 342On.

In another embodiment, in addition to lengthening the duration, once the timing of the ranging response 3415« is known to be within the transport layer ranging window 3405« and the physical layer ranging window 3420«, the physical layer ranging window 3420« can be shortened to reduce noise or integration effects associated with monitoring for the ranging response 3415«.

FIG. 35 is a series of timing diagrams illustrating use of a series of physical layer ranging windows to monitor for a ranging response. In a first iteration 3501a, during a transport layer ranging window 3505a, a ranging request 3510a is transmitted from an OLT to an ONT. During the transport layer ranging window 3505a, a ranging response 3515a from the ONT is not received by the OLT during a series of physical layer ranging windows 3520a, which includes multiple physical layer ranging windows 3525a-d.

In an «th iteration 3501«, during a transport layer ranging window 3505«, a ranging request 3510« is transmitted, and a series of physical layer ranging windows 3520« is shown shifted relative to the series of physical layer ranging windows 3520a of the first iteration 3501a. During the transport layer ranging window 3505«, a ranging response 3515« is received during a physical layer ranging window 3525d in the shifted series of physical layer ranging windows 3520«.

Each series of the physical layer ranging windows 3520a-n may be defined by more than one physical layer ranging window 3525a-d. During each window 3525a-d in the series of physical layer ranging windows 3520a-n, monitoring is enabled (described above in reference to FIGS. 29 A and 29B) for an amount of time equal to or for a portion of each physical layer ranging window 3525a-d. Each physical layer ranging window 3525a-d of the series of physical layer ranging windows 3520a-n may be equally "sized," i.e., similar in duration. Alternatively, each physical layer ranging window of the series of physical layer ranging windows may be differently "sized," i.e., differing in duration. As such, monitoring for a ranging response during the series of physical layer ranging windows may be enabled for regular or irregular durations within the series 3520a-n.

Additionally, in the series of physical layer ranging windows 3520a-n, between each physical layer ranging window 3525a-d, there may be gaps 3530a-c. During each gap 353Oa-c, monitoring for a ranging response is disabled (described above in reference to FIGS. 29A and 29B). In other words, between adjacent physical layer ranging windows (e.g., 3525a and 3525b) in the series of physical layer ranging windows 3520a-n, monitoring for ranging response is enabled, then disabled, then enabled again, and so on. During each gap 353Oa-c, monitoring may be reset (for example, an integrator may be "zeroed"), including at the beginning or the end of each of the gaps 353Oa-c. Furthermore, each physical layer ranging window may be equally "spaced" from one another with such a gap. That is, monitoring for a ranging response may be disabled for a similar duration between adjacent physical layer ranging windows. Alternatively, adjacent physical layer ranging windows may be unequally "spaced" from one other, thus disabling monitoring for different durations. As such, monitoring for a ranging response during a series of physical layer ranging windows may be disabled for regular or irregular durations within the series.

It should be understood that there may be more than four physical layer ranging windows 3525a-d in each series 3520a-n. For example, there may be tens, hundreds, thousands, or millions of physical layer ranging windows in each series 3520a-n depending on an expected length of ranging response, length of transport layer ranging windows 3505a-n, and implementation features.

FIG. 36A is a series of timing diagrams illustrating a shift in a series of physical layer ranging windows to locate a ranging response in full. In an (/7-l)th iteration 3633«-l, during a transport layer ranging window 3605a, a ranging request 361 Oa is transmitted. During the transport layer ranging window 3605a, a ranging response 3615 is received in part during a series of physical layer ranging windows 3620a. The part of the ranging response 3615 received is referred to herein as a received portion 3625, while the remaining portion not received is referred to herein as a non-received portion 3630. The non-received portion may fall within a gap between the physical layer ranging windows or it may arrive after the physical layer ranging windows are halted following receipt of the received portion 3625. In an «th iteration 3633/7, during a later transport layer ranging window

3605b, a ranging request 3610b is transmitted, and a series of physical layer ranging windows 3620b is shifted, in time, relative to the earlier series 3620a. The series of physical layer ranging windows 3620b is shifted, in time, by an amount expected to result in receiving a ranging response 3616 in full during a physical layer ranging window 3622 in the series of physical layer ranging windows 3620b. For example, the series of physical layer ranging windows 3620b may be shifted, in time, by an amount 3631 equal to an amount of time of the non-received portion 3630. Alternatively, the series of physical layer ranging windows 3620b may be shifted, in time, by an amount greater than the non-received portion 3630 but still allowing the ranging response 3616 to fall within the physical layer ranging window 3622.

FIG. 36B is a series of timing diagrams further illustrating shifting a series of physical layer ranging windows to locate a ranging response in full. In an (n-l )th iteration 3673«- 1, during a transport layer ranging window 3655a, a ranging request 3660a is transmitted. During the transport layer ranging window 3655a, a ranging response 3665 is received in part during a series of physical layer ranging windows 3670a. A first received portion 3675 of the ranging response 3665 is received during a first physical layer ranging window 3677a, while a remaining portion is not received during the first physical layer ranging window 3677a. The remaining portion is referred to herein as a non-received portion 3680. The non-received portion 3680 may be received during a gap 3678 (see FlG. 35) and/or during another physical layer ranging window 3677b of the series 3670a.

In an nth iteration 3673«, during a later transport layer ranging window 3655b, a ranging request 3660b is transmitted, and a series of physical layer ranging windows 3670b is shifted, in time, relative to the («-l)th iteration series 3670a. The series of physical layer ranging windows 3670b is shifted, in time, by an amount 3681 expected to result in receiving a ranging response 3666 in full during one physical layer ranging window 3672 of the series of physical layer ranging windows 3670b. For example, the series of physical layer ranging windows 3670b may be shifted, in time, by an amount 3681 equal to an amount of time of the non-received portion 3680. Alternatively, the series of physical layer ranging windows 3670b may be shifted, ih time, by an amount greater than the amount of time of the non- received portion 3680 but not more than an amount that allows for receipt within the window 3672.

In an alternative embodiment, the series of physical layer ranging windows 3670b may be replaced with a subset or just one physical layer ranging window once timing of the ranging response within the transport layer ranging window 3655b is approximately known.

An ability to detect a partial response may be related to noise reduction gained by decreasing a size of the physical layer ranging window. In such a case, an optional, generalized, search methodology might be as follows: 1) reduce a size of the physical layer ranging window until the presence of a ranging response can be identified and located; and 2) further shift and reduce the size of the physical layer ranging window until the ranging response can be precisely captured. The presumes that the noise sensitivity associated with detecting and locating presence of a ranging response in full or in part is less than that for completely processing a ranging response. FIG. 37 is a series of timing diagrams that superimposes effects in an OLT of integration of no-input signal power while waiting to receive a ranging response from an ONT. In a first iteration 3703a, during a transport layer ranging window 3705a, a ranging request 371 Oa is transmitted. During the transport layer ranging window 3705a, the OLT monitors for a ranging response 3715a during a physical layer ranging window 3720a. During the physical layer ranging window 3720a, the OLT integrates and measures a power level 3725a associated with monitoring for a ranging response. The measured power level 3725a associated with monitoring for a ranging response may exceed a threshold 3730 (discussed above in reference to FIG. 24B) due to a long period of integration. If the measured power level 3725a exceeds the threshold 3730, the physical layer ranging window 3720a is reduced in duration in a next iteration. In this example embodiment, transmitting a ranging request 3710a, monitoring for a ranging response 3715a, and reducing the duration of the physical layer ranging window 3720a repeats at least until the measured power level 3725a associated with monitoring for the ranging response 3715a is below the threshold 3730.

Continuing to refer to FIG. 37, in an «th iteration 3703«, during a transport layer ranging window 3705«, a ranging request 3710« is transmitted, and a physical layer ranging window 3720« is reduced in duration relative to the physical layer ranging window 3720a of the first iteration 3703 a, and possibly other previous iterations (not shown). During the transport layer ranging window 3705«, the OLT (not shown) monitors for a ranging response 3715« during the reduced physical layer ranging window 3720«. The OLT measures a power level on an upstream communications path associated with monitoring for ranging response 3715«. In the nth iteration 3703«, the measured power level 3725« associated with monitoring for the ranging response 3715« is below the threshold 3730 due to a reduced integration time, based on the length of time of the physical layer ranging window 3720«. FIG. 38 is a block diagram of an example OLT 3805 in communication with an ONT 3807. A transmitter 3810 transmits a ranging request 3815 to the ONT 3807. A monitor unit 3820 monitors for a ranging response 3825 from the ONT 3807. Associated with the monitoring for the ranging response 3825, a determination unit 3830 determines at least one metric 3835. The at least one metric 3835 is used in connection with upstream communications between the ONT 3807 and the OLT 3805. Based on the determined metric 3835, a configuration unit 3840 sets at least one parameter 3845. The set parameter 3845 is used in connection with upstream communications between the ONT 3807 and the OLT 3805. The transmitter 3810 may send the at least one parameter 3845 to the ONT 3807 so that the ONT 3807 may further communicate (see FIG. 26).

FIG. 39 is a block diagram illustrating an example monitor unit 3905, which may be used in supporting example embodiments of the invention. The monitor unit 3905 may include a receiver 3910, a measurement unit 3920, and a control unit 3940. Alternatively, some or all of the aforementioned components may not be co- located in the monitor unit 3905, but may be remotely located and connected via a communications bus (not shown). In operation of this example monitor unit 3905, the receiver 3910 may monitor for a ranging response 3915. In monitoring for the ranging response 3915, the measurement unit 3920 may measure a power level 3925 associated with monitoring for the ranging response 3915. The measurement unit 3920 may further compare the measured power level 3925 against a threshold 3930 (discussed in detail in reference to FIG. 24B). If the threshold 3930 is exceeded 3932, a notification 3935 may be sent to the control unit 3940. In response to the notification 3935, the control unit 3940 may issue a physical layer ranging window control 3945 to the receiver 3910. The receiver 3910 may respond by shifting, in time, at least one physical layer ranging window (not shown). Alternatively, the receiver 3910 may respond by enlarging or reducing the duration of at least one physical layer ranging window.

FIG. 40 is a flow diagram illustrating an example process 4000 for ranging an ONT in a passive optical network in accordance with an example embodiment of the invention. In connection with (i.e., before or during) a transport layer ranging window, a ranging request is transmitted (4005) from an OLT. The process 4000 monitors (4010) for a ranging response during at least one physical layer ranging window in the transport layer ranging window. The process 4000, from monitoring (4010), determines (4015) at least one metric associated with the monitored ranging response. The determined metric is used in connection with upstream communications between the ONT in the OLT.

FIG. 41 is a flow diagram illustrating an example process 4100 for ranging an ONT in a passive optical network in accordance with an example embodiment of the invention. In connection with (i.e., before or during) a transport layer ranging window, a ranging request is transmitted (4105) from an OLT. Monitoring for a ranging response is enabled (41 10) for an amount of time equal to a physical layer ranging window, or less than a physical layer ranging window where possible in some network applications. An integrator (or other circuitry) used to measure a power level associated with monitoring for a ranging response is reset (4115) at the beginning of a physical layer ranging window or the start of the monitoring. If a ranging response is detected or otherwise received in full (4120) during a physical layer ranging window, at least one metric associated with the ranging response is determined (4125).

If the ranging response is not received in full (4120), but is received in part

(4130), a physical layer ranging window is shifted (4135) by an amount expected to receive the ranging response in full during a physical layer ranging window in a later transport layer ranging window. The process 4100 returns to transmit (4105) a ranging request during a next transport layer ranging window.

If the ranging response is not received in full (4120) and not received in part

(4130), a physical layer ranging window is shifted (4140). The physical layer ranging window may be shifted incrementally (in whole number, fractional, or variable increments) across a transport layer ranging window. After shifting (4140) a physical layer ranging window, if the transport layer ranging window is not yet covered (4145) by the physical layer ranging window (i.e., monitoring across the transport layer ranging window is not complete and a ranging response has not yet been found), the process 4100 returns. The process 4100 returns to transmit (4105) a ranging request, to enable (41 10) monitoring for a ranging response, and to reset

(41 15) the integrator (or other circuitry) used to measure the power level associated with monitoring for the ranging response.

FIG. 42 is a flow diagram illustrating an example process 4200 for ranging an ONT in a passive optical network in accordance with an example embodiment of the invention. In connection with (i.e., before or during) a transport layer ranging window, a ranging request is transmitted (4205) from an OLT. Monitoring for a ranging response is enabled (4210) for an amount of time equal to at least one physical layer ranging window. An integrator (or other circuitry) used to measure a metric, such as a power level, associated with monitoring for a ranging response is reset (4215) at the beginning of a physical layer ranging window or the staπ of the monitoring.

If a ranging response is monitored (4220) or otherwise received during a physical layer ranging window, a metric, such as power level, associated with the monitoring is measured (4225). If the measured metric does not exceed a threshold

(4230), at least one metric associated with the ranging response is determined

(4235). The ONT is consequently ranged. If, however, the measured metric exceeds a threshold (4230). a physical layer ranging window may be reduced in duration (4240). In this example embodiment, the process 4200 repeats at least until the metric associated with monitoring for a ranging response, measured during the reduced physical layer ranging window, is less than the threshold.

If a ranging response is not received (4220) during a physical layer ranging window, the physical layer ranging window may be enlarged in duration (4245). The process 4200 repeats at least until a ranging response is received during the enlarged physical layer ranging window. The systems of FIGS. 38 and 39 and flow diagrams of FIGS. 40-42 may be implemented in the form of software, firmware, or hardware. If implemented in software, the software may be any applicable software language that can be stored on a computer readable medium, such as RAM or ROM, or distributed via a computer network. A general purpose or application specific processor may load and execute the software, causing the processor to be configured to operate in a manner as disclosed herein.

The above description referring to FIGS. 26-42 describes ranging an optical network terminal (ONT) in a passive optical network (PON). An example method may include: (i) transmitting a ranging request from an optical line terminal (OLT) to an ONT in connection with a transport layer ranging window; (ii) monitoring for a ranging response from the ONT during at least one physical layer ranging window within the transport layer ranging window, the transport layer ranging window having a duration longer than the physical layer ranging window; and (iii) determining at least one metric associated with the ranging response for use in connection with upstream communications between the ONT and the OLT. The metric(s), used in connection with upstream communications, are accurately determined, and communications faults during normal operations are thus reduced.

FIGS. 43-51 illustrate example embodiments of an aspect of the present invention in which an optical receiver of an optical line terminal (OLT) is reset at about a time a ranging signal from an optical network terminal (ONT) is expected to be received to minimize the effects caused by an inappropriate presence of an unmodulated or modulated optical signal on an upstream communications path. During standard ranging, a receiver of an OLT is reset at a time which corresponds to a closest distance the ONT can be from the OLT (e.g., a time corresponding to a real distance of 1 kilometer (km) or an "ideal" distance of Okm). In contrast, when using a rogue tolerant ranging method according to an embodiment of the present invention, resetting of the receiver of the OLT is delayed by a time delay, such as an equalization delay (Te) stored for each ONT. In this way, resetting of the receiver is delayed (e.g., by delaying when a reset signal is sent) until just before a ranging response from the ONT is expected to be received. In other words, a time to reset a receiver of an OLT may be delayed until just before a ranging response from an ONT is expected to be received.

The time to reset the receiver of the OLT may be based on a previous successful ranging attempt, presumably before a rogue ONT was added to an optical distribution network (ODN), such as a passive optical network (PON). Such a time may be incremented in an iterative manner, for example, from minus 20 bit-times to plus 20 bit-times before or after the time to allow for variations. Each bit-time may be, for example, 6 nanoseconds at 155 megahertz (MHz). In other words, a time to reset a receiver may be changed to allow correct communication to an ONT when a rogue ONT is also present on the ODN.

When standard ranging fails to establish communication with an ONT, the rogue tolerant ranging method according to an embodiment of the present invention may be used. If the rogue tolerant ranging method succeeds (i.e., an ONT is successfully ranged), this indicates to an operator that one or more rogue ONTs are present and affecting the ODN. Such rogue ONTs can be identified and removed at a later time without further loss of service to other ONTs on the ODN. The rogue tolerant ranging method allows all ONTs on the ODN, including a rogue ONT, to communicate with the OLT, even in the presence of the rogue ONT.

The rogue tolerant ranging method, unlike existing error detection techniques (e.g., those described in the various PON protocols), detects and identifies the aforementioned rogue ONT malfunctions. Moreover, no specialized test equipment is used to overcome these malfunctions; the OLT can be configured in hardware, software, or combination thereof, to test and adjust for the rogue ONT(s). FIG. 43 illustrates an example optical line terminal (OLT) 4300 to tolerate a fault condition otherwise affecting ranging of an ONT. The OLT 4300 includes an OLT receiver 4305, determining unit 4310, time delay changing unit 4315, and resetting unit 4320. At about a time the OLT receiver 4305 is expected to receive a ranging signal 4306 (e.g., a ranging response) from an ONT being ranged (not shown) the OLT receiver 4305 is reset by the resetting unit 4320. In one embodiment, the time the OLT receiver 4305 is reset by the resetting unit 4320 is based on an equalization delay assigned to the ONT previously. In another embodiment, the time the OLT receiver 4305 is reset by the resetting unit 4320 is based on a time previously determined by a successful ranging attempt.

Whether a ranging attempt is successful is determined by the determining unit 4310. The determining unit 4310 determines whether ranging is successful by, for example, measuring a no-input signal power level on a communications pathway. FIG. 44 is a diagram illustrating how a transmitted optical power level on a communications pathway from a faulty ONT affects whether an ONT is successful ranged by an OLT. A message diagram 4400a illustrates an exchange of ranging signals or otherwise messages (e.g., a ranging grant (or ranging request) and a ranging response (or ranging cell)) between an OLT 4401 and an ONT 4402 during a ranging window 4420. A transmitted power level versus time plot 4400b illustrates the ONT 4402 transmitting a no-input signal power level 4403 during the ranging window 4420. The no-input signal power level 4403 may be, for example, a power level of a rogue ONT or power levels of non-transmitting ONTs. A received power level versus time plot 4400c illustrates the OLT 4401 receiving the no-input signal power level 4403, which has been integrated by an integrator 4404 in a receiver (not shown) of the OLT 4401, as an integrated no-input signal power level 4405.

The transmitted power level versus time plot 4400b indicates that the no- input signal power level 4403 may be constant during the ranging window 4420, where the constant level may be a normal low level (e.g., -4OdBm) or a faulty high level (e.g., between -3OdBm and -2OdBm, or higher). The integrated no-input signal power level 4405 ramps up from an integrated no-input signal power level at time t_mi_ti_ai 4410 to an integrated no-input signal power level at time t_f,_na| 4415, over the ranging window 4420.

In operation, while the no-input signal power level 4403 is being integrated over the ranging window 4420, the OLT 4401 sends a ranging grant 4425 to the ONT 4402. The ONT 4402, in turn, responds with a ranging response 4430. The OLT 4401, having sent the ranging grant 4425, receives the ranging response 4430 from the ONT 4402 during the ranging window 4420 or it reports a ranging error.

Typically, the receiver of the OLT 4401 is reset between adjacent upstream timeslots to accommodate power levels which vary from ONT to ONT. During ONT ranging, however, an upstream timeslot is effectively enlarged to accommodate variability in supported fiber lengths, i.e., more than one upstream timeslot is used for the ranging window 4420. For example, the ONT 4402 may be located up to 20 kilometers away from the OLT 4401. To accommodate this distance, the duration of the ranging window 4420 is set sufficiently long enough to allow the ONT 4402 located 20 kilometers away from the OLT 4401 to receive the ranging grant 4425 and the OLT 4401 to receive the ranging response 4430.

When the duration of the ranging window 4420 is set for a long period of time, the receiver of the OLT 4401 is not reset during this period of time. As a result, a no-input signal power level, such as power level of rogue ONT on the ODN, have more time to be integrated by the receiver of the OLT 4401, thus increasing the integrated no-input signal power level 4405.

As the received power level versus time plot 4400c illustrates, integrating the no-input signal power level 4403 over a long period of time causes the integrated no-input signal power level 4405 to ramp (or increase). Consequently, over time, it may be more difficult to distinguish a zero-bit input signal (i.e., a zero bit) from a one-bit input signal (i.e., a one bit) possibly causing ranging errors and/or may lead to upstream communications problem(s)

Rather than using a typical ranging window, such as the ranging window 4420, to determine when to reset a receiver of an OLT, in one embodiment of the present invention, the receiver is reset at about a time a ranging response from an ONT is expected to be received. Changing the time the receiver is reset may be referred to as a "dynamic reset." Through the use of the dynamic reset, the amount of time a power level of rogue ONT is integrated may be limited, thereby reducing the adverse effects associated with integrating such a power level. In this way, the ranging techniques according to this and other embodiments of the present invention tolerate a fault condition otherwise affecting ranging of an ONT. In some instances, however, resetting a receiver at about a time a ranging response from an ONT is expected to be received by an OLT does not result in successful ranging of the ONT. For example, a time between a time a ranging response from an ONT is expected to be received by an OLT and a time a ranging response from an ONT is actually received is large, possibly in terms of a time window or relative to a sensitivity of a particular receiver with respect to an amount of power a rogue ONT adds to an optical fiber link. Consequently, despite resetting the receiver at about the time the ranging response from the ONT is expected to be received, a power level is integrated sufficiently long enough to affect ONT ranging adversely. In another example, a time a ranging response from an ONT is actually received occurs before a time a ranging response from the ONT is expected to be received. Again, despite resetting the receiver at about the time the ranging response from the ONT is expected to be received, a power level is integrated sufficiently long enough to affect ONT ranging adversely. In such instances, a time to reset a receiver is changed (described later in greater detail).

Additional techniques for determining whether ranging is successful are described in reference to FIGS. 1-17.

Returning to FIG. 43, in an event the determining unit 4310 determines (e.g., via a ranging result 4307) ranging is unsuccessful; the determining unit 4310 communicates its results via a determination message 431 1 to the time delay changing unit 4315. The time delay changing unit 4315, in turn, changes the time to reset the receiver of the OLT, such as via a time to reset a receiver message 4316.

In one embodiment, the time delay changing unit 4315 is configured with an adder (not shown) adapted to add a delay to the time when a ranging response from an ONT is expected to be received by an OLT. In another embodiment, the time delay changing unit 4315 is configured with a subtracter (not shown) adapted to subtract a delay from the time when a ranging response from an ONT is expected to be received by an OLT. In yet another embodiment, the time delay changing unit 4315 is configured with an incrementer (not shown) adapted to increment a delay in an iterative manner within a range of delays to delay the time to reset the receiver of the OLT and to compensate for variations in an equalization delay due to physical conditions expected to be experienced by an optical distribution network. In this way, the time to reset the receiver of the OLT is changed by the delay.

At the time to reset the receiver on the OLT, the resetting unit 4320 resets the OLT receiver 4305, such as via a reset signal 4321.

In FIG. 45, an optical line terminal (OLT) (not shown) with an OLT time line 4505 ranges an optical network terminal (ONT) (not shown) with an ONT time line 4510. At a time T_inιuai 4515, the OLT sends a ranging grant 4520 to the ONT. At a time T_eXp_ected 4525, a ranging response 4530 from the ONT is expected to be received by the OLT. In expectation, a receiver (not shown) of the OLT is reset at a time T_rcset 4545. In this example, the receiver is reset at about a time the ranging response 4530 is expected to be received. That is, the time T_expe_cι_ed 4525 and time T_rcsci 4545 occur about the same time.

In one embodiment, a receiver is reset at a time T_rcse! and disabled at a time T_di_sabi_cΦ Between the time T_rcsct and the time Td,sabied is an expected ranging response time T_rangmg _resp_onse, which is typically at least as long as a ranging response message or signal. Disabling the receiver at Tdj_sabied limits the effects of post-integration by an integrator (not shown) which may interfere with ONT ranging and/or may lead to upstream communications problem(s).

In this example, rather than at the time T_expected 4525, the OLT actually receives the ranging response 4530 at a time T_acuiai 4535. Between the time T_eXpected 4525 and the time T_acnjai 4535, in a typical optical receiver manner, the receiver of the OLT integrates a power level of a rogue ONT for a time T,_megrate 4540, which may extend further along the OLT time line 4505 to an upper bound of a typical ranging window (e.g., a time equivalent to ranging an ONT 20 kilometers from the OLT). By not resetting the receiver of the OLT at the time T_in,tιai 4515, but at about the time T_eXp_e_ted 4525 (e.g., at the time T_reSet 4545), in some embodiments, the amount of time the receiver integrates is limited or otherwise shortened to the time

I integrate 4540. FIGS. 46A and 46B are timing diagrams illustrating changing a time to reset a receiver of an OLT in an event ONT ranging is unsuccessful.

In FIG. 46A, an OLT operating according to an OLT time line 4605a ranges an ONT operating according to an ONT time line 4610a. At a time T_inj_lιaι 4615 a, the OLT sends a ranging grant 4620a to the ONT. At a time T_CXpCCtcd 4625a, a ranging response 4630a from the ONT is expected to be received by the OLT. In expectation, a receiver of the OLT is reset at about the time T_expcclcd 4625a. Rather than at the time T_=xpcctcd 4625a, the ranging response 4630a is actually received by the OLT at a time T_actuai 4635a. In this example, despite resetting the receiver at about the time T_expecte_d 4625a in a first ranging attempt, ranging is unsuccessful. In a second ranging attempt, the time to reset the receiver is changed by adding a delay 4640a to the time T_eχ_pecι_ed 4625a. With the delay 4640a added, the receiver is reset at a time T_reSct 4645a, and ranging is successful. With the ONT successfully ranged, the time T_reset 4645a may be optionally stored. In others words, in an event ranging is successful, the time T_reset 4645a is stored. As such, the receiver of the OLT in subsequent ranging attempts is not reset at the time T_CXpccted 4625a, but at the time T_rcSct 4645a.

In an alternative embodiment, resetting a receiver of an OLT at about a time a ranging response is expected to be received is based on a time which resulted in a successful ranging attempt previously.

In FIG. 46B, an OLT operating according to an OLT time line 4605b ranges an ONT operating according to an ONT time line 4610b. At a time Tj_πitιa| 4615b, the OLT sends a ranging grant 4620b to the ONT. At a time T_CXp_ected 4625b, a ranging response 4630b from the ONT is expected to be received by the OLT. In expectation, a receiver of the OLT is reset at about the time T_expCcted 4625b. Rather than at the time T_expected 4625b, the ranging response 4630b is actually received by the OLT at a time T_actuai 4635b.

In this example, despite resetting the receiver at about the time T_expecι_ed 4625b in a first ranging attempt, ranging is unsuccessful. In a second ranging attempt, the time to reset the receiver is changed by subtracting a delay 4640b from the time T_eXpe_cιed 4625b. With the delay 4640b subtracted, the receiver is reset at a time T_re_set 4645b, and ranging is successful. With the ONT successfully ranged, the time T_resei 4645b may be optionally stored. In others words, in an event ranging is

■ successful, the time T_reSet 4645b is stored. As such, the receiver of the OLT in subsequent ranging attempts is not reset at the time T_eXpected 4625b, but at the time

Tresei 4645b. In an alternative embodiment, resetting a receiver of an OLT at about a time a ranging response is expected to be received is based on a time which resulted in a successful ranging attempt previously.

To ensure upstream communications sent from an ONT is received by the OLT in a correct time slot, relative to upstream communications from other ONTs, data is delayed at least for an equalization delay before being sent. Equalization delays are assigned -to ONTs to equalize logical distances between the OLT and ONTs, making every ONT appear equidistant from the OLT. Since physical distances from the OLT vary from ONT to ONT, the equalization delays also vary from ONT to ONT. Based on an equalization delay assigned to a given ONT, a time a ranging response from the given ONT is expected to be received can be calculated or otherwise determined. As such, resetting a receiver about the time the ranging response from the ONT is expected to be received may be based on the equalization delay for the given ONT. However, an equalization delay for a given ONT varies, for example, as physical conditions experienced (or expected to be experienced) by an optical distribution network (ODN) change. For example, temperature variations cause fiber optic cables to lengthen and shorten, effectively causing the ONT to be further away from or closer to an OLT in optical path distance. Accordingly, to ensure the OLT receives upstream communications in the correct time slot, an equalization delay for a given ONT may be updated with some periodicity. Consequently, a time a ranging response from an ONT is expected to be received by an OLT and a time a ranging response from an ONT is actually received by the OLT may differ throughout a day or from season to season. Generally speaking, to accommodate such variations, a time to reset a receiver of an OLT may be delayed (or advanced) in increments. FIG. 47 illustrates an OLT operating according to an OLT time line 4705 ranging an ONT operating according to an ONT time line 4710. At a time T_inma| 4715, the OLT transmits a ranging grant 4720 to the ONT. The OLT expects to receive a ranging response 4725 from the ONT at about a time T_expected 4730 based on an equalization delay (not shown) known for the ONT. Due to variations, however, the ONT transmits the ranging response after an equalization delay Te_actua| 4735, which differs from the equalization delay known for the ONT. Consequently, the OLT receives the ranging response 4725 not at the time T_eXp_ected 4730, but rather at a time T_acniai 4740. To accommodate such variations in an equalization delay, a time to reset a receiver of the OLT is changed.

In FIG. 47, a time to reset a receiver of the ONT T_reset 4745 is delayed for one or more delay increments 4750a, 4750b...4750«, generally 4750a-/?. In one embodiment, a size (or duration) of the delay increments 4750 depends on a transmission rate and is measured in "bit times." A "bit time" is an amount of time needed to eject one bit at a given rate of transmission. For example, transmitting at rate 155.52 Megabits per second (Mbps), one bit is ejected every 6 nanoseconds. Thus, at 155.52 Mbps, one bit time is equal to 6 nanoseconds per bit. As another example, at 1 Gigabits per second (Gbps), one bit time is equal to 1 nanoseconds per bit. In another embodiment, a size (or duration) of the delay increments 4750 depends on an overall system tolerance window. For example, the overall system tolerance window may be defined or otherwise configured to be plus or minus 100 nanoseconds. Accordingly, a duration of each delay increment is some portion of the plus or minus 100 nanoseconds. Continuing Io refer to FIG. 47, the time T_reseι 4745 (i.e., the time to reset the receiver) is delayed for two delay increments, viz., 4750a and 4750b. That is, from the time T_expcctCd 4730 (i.e., the time the ranging response is expected to be received), two delay increments elapse before resetting the receiver. In this example, the time T_reset 4745 is delayed for whole number multiples of the delay increments 4750. In another embodiment, a time to reset a receiver is delayed for something less than whole number multiples of delay increments, e.g., 1-1/2 delay increments, 2-3/4 delay increments, and so forth. In FIG. 48 A, due to a variation, transmitting a ranging response 4805a is delayed for an actual equalization delay Te_acmai 4810a. Consequently, the ranging response 4805a is actually received at a time T_actuai 4815a. Based on an equalization delay known to an OLT, however, the ranging response 4805a is expected to be received at a time T_eXpected 4820a. In this instance, the time T_actuai 4815a occurs in time before the time T_eXpected 4820a.

In a first ranging attempt, resetting a receiver of the OLT is advanced by n number of delay increments from the time T_expecιed 4820a, and the receiver is reset at a time T_reSet 4825a-l . In this example, the first ranging attempt is unsuccessful, i.e., the ranging the ONT is unsuccessful. In an event ranging is unsuccessful in a next ranging attempt, the time to reset the receiver of the OLT is incremented (i.e., a time at which the receiver of the OLT is reset is incremented).

In a second ranging attempt, a time at which the receiver of the OLT is reset is advanced (not shown) by n-\ number of delay increments from the time T_cxpCctcd 4820a. In this example, the second ranging attempt is unsuccessful. In a third ranging attempt, a time at which the receiver of the OLT is reset is advanced by n-2 delay increments from the time T_expeclCd 4820a and the receiver is reset at a time T_reSei 4825a-2. In this example, the third ranging attempt is successful.

In FIG. 48B, due to a variation, transmitting a ranging response 4805b is delayed for an actual equalization delay Te_actua| 481 Ob. Consequently, the ranging response 4805b is actually received by the OLT at a time T_actuaι 4815b. Based on an equalization delay known to an OLT, however, the ranging response 4805b is expected to be received at a time T_cxpccιed 4820b. In this instance, the time T_acιuai 4815b occurs after the time T_expcclCd 4820b. In a first ranging attempt, resetting a receiver of the OLT is advanced by zero number of delay increments from the time T_eχ_pected 4820b and the receiver is reset at a time T_reSet 4825b- 1. In this example, the first ranging attempt is unsuccessful, i.e., the ranging the ONT is unsuccessful. In an event ranging is unsuccessful in a next ranging attempt the time to reset the receiver of the OLT is incremented. In a second and a third ranging attempt, the time to reset the receiver of the

OLT is advanced (not shown) by 1 and 2 number of delay increments from the time T_CXp_ccιed 4820, respective. In this example, the second and the third ranging attempt are unsuccessful. In a fourth ranging attempt, the time to reset the receiver of the OLT is advanced by 3 delay increments from the time T_cxpcctCd 4820b and the receiver is reset at a time T_resct 4825b-2. In this example, the fourth ranging attempt is successful. With the ONT successfully ranged, the time T_reSeι 4825b-2 may be optionally stored. In others words, in an event ranging is successful, the time T_rese[ 4825b-2 is stored. As such, the receiver of the OLT in subsequent ranging attempts is not reset at the time T_eXpectcd 4820, but at the time T_reSet 4825b-2.

FIG. 48A illustrates in an event a ranging response is actually received before a time a ranging response is expected to be received (e.g., T_eXpe_αed 4820a), a time to reset a receiver (e.g., T_reSeι 4825a- 1) is iteratively incremented by advancing the time to reset a receiver by n number of delay increments from the time T_expec,_ed. FIG. 48B illustrates in an event a time a ranging response is actually received after a time a ranging response is expected to be received (e.g., T_eXpec_ted 4820b), a time to reset a receiver (e.g., T_reSei 4825b- 1) is iteratively incremented by delaying the time to reset a receiver by n number of delay increments from the time T_expecιcd. In contrast to FIGS. 48A and 48B, in an event a ranging response is actually received before or after a time a ranging response is expected to be received (Texpccted). a time to reset a receiver (T_reSet) is iteratively incremented by both advancing and delaying the time T_reSet by n number of delay increments from the time Texpecied-

In FIG. 49, transmitting a ranging response 4905 is delayed for an actual equalization delay Te_actuai 4910. Consequently, the ranging response 4905 is actually received at a time T_actuai 4915. Based on a known equalization delay, however, the ranging response 4905 is expected to be received at a time T_expecled 4920. To accommodate such variation a time to reset a receiver is changed by iteratively incrementing a delay with a range of delays. For purposes of describing this and other embodiments, delay increments advancing a time to reset a receiver of an OLT (T_reSet) so that that the time (T_reSe_t) occurs in time before a time a ranging response from an ONT is expected to be received (T_expect_ed) are referred to hereinafter as "negative" delay increments. Conversely, delay increments delaying a time to reset a receiver of an OLT (T_rcSct) so that the time T_reSeι occurs in time after the time T_expected are referred to hereinafter as "positive" delay increments. One skilled the art will readily acknowledge the choice of labels is arbitrary and is not intended to be limiting.

Continuing to refer to FIG. 49, a range of delay increments 4923 includes n number of negative delay increments and tn number of positive delay increments. In a first ranging attempt, the time to reset the receiver of the OLT is advanced by n number of negative delay increments from the time T_expc_cted 4920, and the receiver is reset at a time T_reSeι 4925-1. In this example, the first ranging attempt is unsuccessful, i.e., ranging of the ONT is unsuccessful. In an event ranging is unsuccessful; in a next ranging attempt the time to reset the receiver of the OLT is changed by incrementing to a next delay increment within the range of delay increments 4923. In an n ranging attempt, the time to reset the receiver of the OLT is advanced by zero number of negative delay increments from the time T-_xpcct-d 4920, and the receiver is reset at a time T_resct 4925-2. In this instance, resetting the receiver at about the time the ranging response is expected to be received does not result in successful ranging. In («+2)^th ranging attempt, the time to reset the receiver of the OLT is delayed by 2 positive delay increments from the time T_eXp_ect_ed 4920, and the receiver is reset at a time T_rese_t 4925-3. In this example, the third ranging attempt is successful. With the ONT successfully ranged, the time T_reSet 4925-3 may be optionally stored. In others words, in an event ranging is successful, the time T_resel 4925-3 is stored. As such, the receiver of the OLT in subsequent ranging attempts is not reset at the time T_expecι_d 4920, but at the time T_rescl 4925-3.

In an alternative embodiment, resetting a receiver of an OLT at about a time a ranging response is expected to be received is based on a time which resulted in a successful ranging attempt previously. FIGS. 48A, 48B, and 49 illustrate changing a time to reset a receiver in a

"forward" direction in time. For example in FIG. 48 A, in a first ranging attempt, the time to resetthe receiver is advanced by n number of delay increments, and the receiver is reset at the time T_rcsct 4825-1. Then in a second ranging attempt, the time to reset the receiver of the OLT is advanced by /7-1 number of delay increments, and the receiver is reset at the time T_rese_t 4825-2. The time T_resel 4825- 1 occurs before the time T_reSei 4825-2. One skilled in the art, however, will readily recognize embodiments of the present invention are not limited to this example.

For example, in a first ranging attempt, a time to reset a receiver of an OLT is delayed by n number delay increments from a time a ranging response from an ONT is expected to be received (T_eXpected)- In a second ranging attempt, resetting the receiver is delayed by «-1 number of delay increments from the time T_expcctC(j, and so on. With each successive ranging attempt, a time to reset a receiver (T_rcsct) occurs earlier in time. That is to say, a time to reset a receiver is changed in a "backwards" direction in time relative to the time T_expectCd in successive ranging attempts.

In another example, in a first ranging attempt, a time to reset a receiver of an OLT is delayed by n number of delay increments from a time a ranging response from an ONT is expected to be received (T_expccted). In the case of n being equal to zero, the receiver is reset at about the time the ranging response from the ONT is expected to be received. In a second ranging attempt, resetting the receiver is delayed by n number of delay increments in one direction in time. In a third ranging attempt, resetting the receiver is delayed by n number of delay increments in the other direction in time, and so on. With each successive ranging attempt, a time to reset a receiver (T_reSet) occurs either earlier or later in time. That is to say, a time to reset a receiver starts at a "middle time" and can be shifted relative to the middle time in either directions in time in successive ranging attempts.

In yet another example, in a first ranging attempt, a time to reset a receiver of an OLT is delayed by n delay increments from a time a ranging response from an

ONT is expected to be received (T_expcclC(i). In a second ranging attempt, resetting the receiver is delayed by n/1 delay increments from the time T_cxpected, and so on. With each successive ranging attempt, a time to reset a receiver (T_reSet) is halved.

In still another example, in a first ranging attempt, a time to reset a receiver of an OLT (T_r__set) is delayed by any number of delay increments from a time a ranging response from an ONT is expected to be received (T_expecιed)- In a second ranging attempt, the time T_rcsc( is delayed by any number delay increments from the time T_exp_ected, and so on. That is to say, the time to reset a receiver of an OLT is randomized.

In still yet another example, a time to reset a receiver of an OLT is delayed from a time a ranging response from an ONT is expected to be received (T_eXpecιcd) by a delay which has been calculated or otherwise determined.

In FIG. 50, a^' flow diagram 5000 illustrates ranging an ONT. Ranging the ONT starts (5002). A receiver of an OLT is reset (5005) at about a time a ranging response from the ONT is expected to be received. By doing so, a fault condition affecting ranging of the ONT is tolerated, and traffic and communications are uninterrupted by a rogue ONT. Ranging the ONT ends (5007). The ONT is ranged.

In FIG. 51, a flow diagram 5100 illustrates identifying a fault condition. A ranging attempt using a standard ranging window is determined (5105) successful or not. If determined (5105) successful, there is no fault condition to be identified, and the flow diagram 5100 ends. If determined (5105) unsuccessful, however, in a next ranging attempt, a receiver of an OLT is reset (51 10) at a time a ranging response from an ONT is expected to be received (T_cxpected).

Whether the next ranging attempt is successful is determined (51 15). If determined (51 15) successful, a fault condition is identified and the flow diagram 5100 ends. If determined (51 15) unsuccessful, however, in a next ranging attempt, a time to reset a receiver of an OLT (T_reset) is changed (5120). With the time T_reset changed (5120), the receiver of the OLT is reset (5125) at the time T_reseι.

Whether the next ranging attempt is successful is determined (5130). If determined (5130) successful, a fault condition is identified and the flow diagram 5100 ends. If determined (5130) unsuccessful, however, the flow diagram further determines (5135) whether to continue changing the time T_resei-

Whether the flow diagram 5100 determines (5135) to continue changing the time T_reset may be limited by, for example, a number of instances configured or otherwise permitted. By way of example, the number of instances is limited to 20 and, as such, the time T_reSet is changed (5120) 20 times before the time T_rese, is no longer changed.

In another example, the time T_reset is changed (5120) until a range of times is tried or otherwise covered. By way of example, the time T_reSet is changed (5120) by 1 to 100 nanoseconds. That is, the time T_rcsct is changed (5120) by 1 nanosecond in a first ranging attempt, by 2 nanoseconds in a second ranging attempt, and so forth. The time T_resc, continues to change (5120) until the time T_rcsct is changed by 100 nanoseconds. If the flow diagram 5100 determines (5135) not to continue changing the time T_reSet, a fault condition is identified and the flow diagram 5100 ends. If however, the flow diagram 5100 determines (5135) to continue changing the time T_rcsct, the time T_rcSci is incremented (5140). The flow diagram 5100 continues and the receiver of the OLT is reset (5125) at the time T_rcSet- Changing (5120) the time T_reset and resetting (5125) the receiver of the OLT at the time T_reseι in a next ranging attempt continues until the flow diagram 5100 either determines (5130) that a next ranging attempt is successful or further determines (5135) not to continue changing the time T_rcsel. In either instance, a fault condition is identified. In FIG. 51, the flow diagram 5100 illustrates incrementing (5140) the time to reset a receiver of an OLT (T_resct) so that in each successive ranging attempt, the receiver is reset (5125) at a later and later time. In an alternative embodiment (not shown), a time to reset a receiver of an OLT is decremented so that in each successive ranging attempt, the receiver is reset at an earlier and earlier time. The above description referring to FIGS. 43-51 describes ranging an ONT while tolerating to a fault condition. A fault condition of a continuous stream of light up a shared fiber from an optical network terminal (ONT) to an optical line terminal (OLT) may adversely affect ranging of the ONT by the OLT. In an example embodiment, an optical receiver of an optical line terminal (OLT) is reset at about a time a ranging signal from an ONT is expected to be received. Through the use of the example embodiment, an ONT can be ranged in the presence of a rogue ONT causing the fault condition. Moreover, the example embodiment enables the rogue ONT to be ranged in a presence of the fault condition and an Optical Distribution Network (ODN), which includes the OLT and the rogue ONT, to continue to support communications in a presence of the fault condition.

While this invention has been particularly shown and described with references to preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention encompassed by the appended claims.

Although several embodiments are described in terms of optical elements, other embodiments may be applied to other networks, such as wired or wireless networks. For example, the OLT and ONTs may correspond to routers and servers in an electrical network. In addition, although described as "cards" herein, it should be understood that PON cards, OLT cards, or ONT cards may be systems or subsystems without departing from the principles disclosed hereinabove.

It should be understood that elements of the block diagrams, network diagrams, and flow diagrams described above may be implemented in software, hardware, or firmware. In addition, the elements of the block diagrams and flow diagrams described above may be combined or divided in any manner in software, hardware, or firmware. If implemented in software, the software may be written in any language that can support the embodiments disclosed herein. The software may be stored on any form of computer-readable medium, such as random access memory (RAM), read only memory (ROM), compact disk read only memory (CD- ROM), and so forth. In operation, a general purpose or application specific processor loads and executes the software in a manner well understood in the art.

Claims

CLAIMS What is claimed is:

1. A method of identifying a passive optical network failure comprising: identifying a control optical network terminal (ONT) from among multiple ONTs in a passive optical network, the control ONT functioning normally with a normal, non-data, output signal level; identifying a test ONT from among the multiple ONTs, the test ONT potentially malfunctioning with an above normal, non-data, output signal level; and determining the test ONT is actually malfunctioning by attempting to range the control ONT and the test ONT and observing both ONTs fail to range.

2. The method of claim 1 wherein identifying a control ONT comprises: attempting to range the multiple ONTs; identifying ONTs that fail to range; disabling outputs of the ONTs that fail to range; enabling the output of a given ONT that failed to range; attempting to range the given ONT individually; and defining the given ONT as a control ONT if it successfully ranges.

3. The method of claim 2 wherein identifying a control ONT further comprises verifying the control ONT is not malfunctioning by successfully ranging another ONT individually that also failed to range.

4. The method of claim 1 further comprising identifying multiple control ONTs by: attempting to range the multiple ONTs; identifying ONTs that fail to range; disabling outputs of the ONTs that fail to range; enabling the output of ONTs individually that failed to range; attempting to range the ONTs individually; and defining ONTs successfully range as control ONTs.

5. The method of claim 4 wherein identifying multiple control ONTs further comprises identifying control ONTs at least until a condition is met, including at least one of the following conditions: a time limit, a number of control ONTs are determined, a percentage of the multiple ONTs are determined to be control ONTs, a percentage of the ONTs that failed to range are determined to be control ONTs, and a stop command from an operator is received.

6. The method of claim 1 wherein identifying a test ONT comprises: attempting to range the multiple ONTs; identifying ONTs that fail to range; disabling outputs of the ONTs that fail to range; enabling the output of a given ONT that failed to range; attempting to range the given ONT individually; and defining the given ONT as a test ONT if it again fails to range.

7. The method of claim 1 wherein identifying a test ONT comprises: defining two groups of ONTs by dividing the multiple ONTs into a first group and a second group; disabling outputs of the ONTs; enabling the outputs of the first group and attempting to range the ONTs in the first group; identifying the second group contains the test ONT by the first group successfully ranging;

identifying the first group contains the test ONT by the first group failing to range and by verifying there are not test ONTs in each group by disabling the output of the first group, enabling the output of the second group, and successfully ranging the ONTs in the second group; and identifying the test ONT by repeatedly dividing each group identified as containing the test ONT into two groups and determining which group contains the test ONT, after the remaining size of each group is one, the test ONT is the ONT of the group that fails to range.

8. The method of claim 7 wherein, responsive to determining both groups of ONTs are unable to range, identifying a test ONT further comprises attempting to range ONTs individually that did not range and defining an ONT that fails to range as a test ONT.

9. The method of claim 1 wherein determining the test ONT is actually malfunctioning further comprises disabling output of the test ONT and attempting to range the control ONT and observing the control ONT range

10. The method of claim 1 further comprising notifying an operator the test ONT is malfunctioning.

1 1. An apparatus for identifying a passive optical network fault, comprising: a control optical network terminal (ONT) identification module to monitor multiple ONTs in a passive optical network (PON) and identify a control ONT functioning normally with a normal, non-data, output signal level; a test ONT identification module to monitor the multiple ONTs and identify a test ONT potentially malfunctioning with an above normal, non- data, output signal level; and a verification module configured to determine the test ONT is actually malfunctioning by ranging the control ONT and the test ONT and observing both ONTs fail to range.

12. The apparatus according to claim 11 wherein the control ONT identification module comprises: a ranging unit to communicate with the multiple ONTs to initiate ranging of ONTs and identify whether ONTs have successfully ranged; an enabling/disabling unit to communicate with the multiple ONTs to enable or disable outputs of the ONTs; and a logic unit in communication with the ranging unit to identify a control ONT.

13. The apparatus of claim 12 wherein the control ONT identification module further comprises a verification unit in communication with the logic unit to verify the control ONT is not malfunctioning by successfully ranging another

ONT that also. failed to range.

14. The apparatus according to claim 1 1 wherein the control ONT identification module comprises: a ranging unit to communicate with the multiple ONTs to initiate ranging of ONTs and identify whether ONTs have successfully ranged; an enabling/disabling unit to communicate with the multiple ONTs to enable or disable outputs of the ONTs; and a logic unit in communication with the ranging unit to identify multiple control ONTs.

15. The apparatus according to claim 14 wherein the control ONT identification module further comprises a limiting unit in communication with the logic unit to stop the identification of multiple control ONTs if a condition is met, including at least one of the following conditions: a time limit, a number of control ONTs are determined, a percentage of the multiple ONTs are determined to be control ONTs, a percentage of the ONTs that failed to range are determined to be control ONTs, and a stop command from an operator is received.

16. The apparatus of claim 1 1 wherein the test ONT identification module comprises: a ranging unit to communicate with the multiple ONTs to initiate ranging of ONTs and identify whether ONTs have successfully ranged; an enabling/disabling unit to communicate with the multiple ONTs to enable or disable outputs of the ONTs; and a logic unit in communication with the ranging unit to identify a test

ONT.

17. The apparatus of claim 1 1 wherein the test ONT identification module comprises: a dividing unit to define a first and second group of ONTs; a ranging unit to communicate with the multiple ONTs to initiate ranging of ONTs and identify whether ONTs have successfully ranged; an enabling/disabling unit to communicate with the multiple ONTs to enable or disable outputs of the ONTs; a logic unit in communication with the ranging unit to identify the group containing the test ONT; a verification unit in communication with the logic unit to verify that only one group contains a test ONT; and a test ONT unit in communication with the dividing unit and the logic unit, the test ONT unit to have the dividing unit repeatedly divide each group containing the test ONT into two groups and to be notified by the logic unit which group contains the test ONT, and after the remaining size is one, to identify the test ONT.

18. The apparatus of claim 17 wherein the test ONT identification module further comprises a switch unit in communication with the verification unit configured upon notification from the verification unit that both sets of ONTs are unable to range to cause the test ONT unit to attempt to range the ONTs individually and to identify an ONT that fails to range as a test ONT.

19. The apparatus of claim 11 wherein the verification module is further configured to determine the test ONT is actually malfunctioning by disabling output of the test ONT and attempting to range the control ONT and observing the control ONT range.

20. The apparatus of claim 11 further comprising a notification generator in communication with the verification module to generate a notification that an

ONT is malfunctioning.

21. A passive optical network comprising: at least one optical line terminal (OLT); at least one optical network terminal (ONT) connected to the OLT by a fiber and configured to communicate with the OLT; a control ONT identification module located at the OLT to monitor multiple ONTs in a passive optical network (PON) and identify a control ONT functioning normally with a normal, non-data, output signal level; a test ONT identification module located at the OLT to monitor the multiple ONTs and identify a test ONT potentially malfunctioning with an above normal, non-data, output signal level; and a verification module located on the OLT to determine the test ONT is actually malfunctioning by attempting to range the control ONT and the test ONT and observe both ONTs fail to range.

22. A computer-readable medium containing a sequence of instructions which, when executed by a digital processor, cause the processor to: identify a control optical network terminal (ONT) from among multiple ONTs in a passive optical network (PON), the control ONT functioning normally with a normal, non-data, output signal level; identify a test ONT from among the multiple ONTs, the test ONT potentially malfunctioning with an above normal, non-data, output signal level; and determine the test ONT is actually malfunctioning by causing the processor to attempt to range the control ONT and the test ONT and observe both ONTs fail to range.

23. A method for detecting an error condition in a passive optical network (PON), the method comprising: measuring a no-input signal power level on a communications path configured to carry upstream communications between multiple optical network terminals (ONTs) and an optical line terminal (OLT) in a passive optical network (PON) at a time no upstream communications are on the communications path from the ONTs to the OLT; comparing the measured no-input signal power level to a threshold; and generating a notification in an event the threshold is exceeded.

24. The method of Claim 23 wherein measuring the no-input signal power level includes averaging the no-input signal power level over a length of time.

25. The method of Claim 23 wherein measuring the no-input signal power level includes measuring the no-input signal power level over a length of time defined by at least a portion of at least one timeslot in an upstream communications frame.

26. The method of Claim 23 further comprising scheduling upstream communications in a manner defining a time no upstream communications are on the communications path.

27. The method of Claim 23 wherein the threshold represents a tolerable no- input signal power level multiplied by a number of ONTs in communication with the OLT.

28. The method of Claim 27 wherein the tolerable no-input signal power level is less than a zero-bit input signal power level for the ONTs.

29. The method of Claim 23 wherein the threshold represents a maximum power level corresponding to a fault associated with upstream communications in a non-communicating state.

30. The method of Claim 23 wherein the threshold is less than a sum of a zero- bit input signal power level of each ONT offset by respective losses between the ONTs and the OLT.

31. The method of Claim 23 further comprising predetermining the threshold based on a configuration of the PON.

32. The method of Claim 23 wherein generating the notification includes generating an alarm notification.

33. The method of Claim 32 wherein generating the alarm notification includes transmitting the alarm notification across a network or presenting the alarm notification locally.

34. An apparatus for detecting an error condition in a passive optical network (PON), the apparatus comprising: a measurement unit which measures a no-input signal power level on a communications path configured to carry upstream communications between multiple optical network terminals (ONTs) and an optical line terminal (OLT) in a passive optical network (PON) at a time no upstream communications are on the communications path from the ONTs to the OLT; a comparison unit which compares the measured no-input signal power level to a threshold; and a notification generator in communication with the comparison unit which generates a notification in an event the threshold is exceeded.

35. The apparatus of claim 34 further comprising a timer coupled to the measurement unit which enables the no-input signal power level to be measured over a length of time defined by at least a portion of at least one timeslot in an upstream communications frame.

36. The apparatus of claim 34 wherein the threshold represents a tolerable no- input signal power level multiplied by a number of ONTs in communication with an OLT.

37. The apparatus of claim 36 wherein the tolerable no-input signal power level is less than a zero-bit input signal power level for the ONTs.

38. The apparatus of Claim 34 wherein the threshold represents a maximum power level corresponding to a fault associated with upstream communications in a non-communicating state tolerated by the OLT.

39. The apparatus of Claim 34 wherein the comparison unit predetermines the threshold based on a configuration of the PON.

40. The apparatus of claim 34 wherein the notification generator generates an alarm notification which is transmitted across a network or is presented locally.

41. An apparatus for detecting an error condition in a passive optical network (PON), the apparatus comprising: means for measuring a no-input signal power level on a communications path configured to carry upstream communications between multiple optical network terminals (ONTs) and an optical line terminal (OLT) in a passive optical network (PON) at a time no upstream communications are on the communications path from the ONTs to the OLT; means for comparing the measured no-input signal power level to a threshold; and means for generating a notification in an event the threshold is exceeded.

42. A method for ranging an optical network terminal (ONT) in a passive optical network (PON), the method comprising: transmitting a ranging request to an ONT in connection with a transport layer ranging window; monitoring for a ranging response from the ONT during at least one physical layer ranging window within the transport layer ranging window, the transport layer ranging window having a duration longer than the physical layer ranging window; and determining at least one metric associated with the ranging response for use in connection with upstream communications between the ONT and the OLT.

43. The method of Claim 42 further comprising setting at least one parameter, used in connection with upstream communications between the ONT and

OLT, based on the at least one metric associated with the ranging response.

44. The method of Claim 42 further comprising enabling the monitoring for a ranging response for an amount of time equal to the physical layer ranging window.

45. The method of Claim 44 wherein enabling the monitoring for a ranging response includes resetting integration associated with monitoring for a ranging response at a beginning of the physical layer ranging window.

46. The method of Claim 42 further comprising: measuring a no-input signal metric on an upstream communications path during the physical layer ranging window; reducing the physical layer ranging window if the measured no-input signal metric on the upstream communications path exceeds a threshold; and repeating the transmitting, monitoring, and reducing at least until the measured no-input signal metric is less than the threshold.

47. The method of Claim 42 wherein transmitting a ranging request and monitoring for a ranging response repeats (i) at least until a ranging response is received during the physical layer ranging window, wherein monitoring for a ranging response includes dynamically adjusting the physical layer ranging window in an iterative manner, or (ii) until determining a ranging response is not within the transport layer ranging window.

48. The method of Claim 42 wherein transmitting a ranging request and monitoring for a ranging response repeats (i) at least until a ranging response is received during the physical layer ranging window, wherein monitoring for a ranging response includes shifting the physical layer ranging window within the transport layer ranging window at least until a ranging response is received during the physical layer ranging window, (ii) until determining a ranging response is not within the transport layer ranging window.

49. The method of Claim 48 wherein shifting the physical layer ranging window includes shifting the physical layer ranging window incrementally across the transport layer ranging window (i) at least until a ranging response is received during the physical layer ranging window or (ii) determining a ranging response is not within the transport layer ranging window.

50. The method of Claim 48 wherein, in an event of receiving a ranging response in part during the physical layer ranging window, shifting the physical layer ranging window includes shifting the physical layer ranging window by an amount expected to result in receiving a ranging response in full during the physical layer ranging window during a later transport layer ranging window.

51. The method of Claim 42 wherein transmitting a ranging request and monitoring for a ranging response repeats (i) at least until a ranging response is received during the physical layer ranging window, wherein monitoring for a ranging response includes lengthening the physical layer ranging window at least until a ranging response is received during the physical layer ranging window, or (ii) until determining a ranging response is not within the transport layer ranging window.

52. The method of Claim 42 wherein transmitting a ranging request and monitoring for a ranging response repeats (i) at least until a ranging response is received during the physical layer ranging window, wherein monitoring for a ranging response includes monitoring during a series of physical layer ranging windows within the transport layer ranging window, wherein the monitoring is enabled for an amount of time equal to each physical layer ranging window of the series and integration associated with the monitoring is reset at a beginning of each physical layer ranging window of the series, or (ii) until determining a response is not within the transport layer ranging window.

53. The method of Claim 52 wherein monitoring during the series of physical layer ranging windows includes shifting the series within the transport layer ranging window (i) at least until a ranging response is received during at least one physical layer ranging window of the series, or (ii) until determining a ranging response is not within the transport layer ranging window.

54. The method of Claim 52 wherein, in an event of receiving a ranging response in part during at least one physical layer ranging window, monitoring during the series of physical layer ranging windows includes shifting the series within the transport layer ranging window an amount expected to result in . receiving a ranging response in full during one physical layer ranging window of the series during a later transport layer ranging window.

55. An apparatus for ranging an optical network terminal (ONT) in a passive optical network (PON), the apparatus comprising: a transmitter to transmit a ranging request to an ONT in connection with a transport layer ranging window; a monitor unit to monitor for a ranging response from the ONT during at least one physical layer ranging window within the transport layer ranging window, the transport layer ranging window having a duration longer than the physical layer ranging window; and a determination unit to determine at least one metric associated with the ranging response for use in connection with upstream communications between the ONT and the OLT.

56. The apparatus of claim 55 further comprising a configuration unit to set at least one parameter, used in connection with upstream communications between the ONT and the OLT, based on the at least one metric associated with the ranging response.

57. The apparatus of Claim 55 wherein the monitor unit includes a control unit operatively coupled to an optical line terminal (OLT) receiver, the OLT receiver being controlled by the control unit.

58. The apparatus of Claim 57 wherein the control unit enables the OLT receiver for an amount of time equal to the physical layer ranging window.

59. The apparatus of Claim 58 wherein the control unit resets the OLT receiver at a beginning of the physical layer ranging window.

60. The apparatus of Claim 57 wherein the monitor unit includes a measurement unit to measure a no-input signal metric on an upstream communications path received by the OLT receiver during the physical layer ranging window and the control unit responds to the measurement unit by reducing the physical layer ranging window if the measured no-input signal metric on the upstream communications path exceeds a threshold, and wherein the transmitter repeats transmitting a ranging request, the monitor unit repeats monitoring for a ranging response, and the control unit repeats reducing the physical layer ranging window at least until the measured no-input signal metric is less than the threshold.

61. The apparatus of Claim 57 wherein the control unit dynamically adjusts the physical layer ranging window in an iterative manner (i) at least until the OLT receiver receives a ranging response during the physical layer ranging window, or (ii) until the control unit determines a ranging response is not within the transport layer ranging window.

62. The apparatus of Claim 57 wherein the transmitter repeats transmitting a ranging request, the monitor unit repeats monitoring for a ranging response, and the control unit shifts the physical layer ranging window within the transport layer ranging window (i) at least until the OLT receiver receives a ranging response during the physical layer ranging window, or (ii) until the control unit determines a ranging response is not within the transport layer ranging window.

63. The apparatus of Claim 61 wherein the control unit shifts the physical layer ranging window incrementally across the transport layer ranging window (i) at least until the OLT receiver receives a ranging response during the physical layer ranging window, or (ii) until the control unit determines a ranging response is not within the transport layer window.

64. The apparatus of Claim 61 wherein, in an event the OLT receiver receives a ranging response in part during the physical layer ranging window, the control unit shifts the physical layer ranging window by an amount expected to result in the OLT receiver receiving a ranging response in full during the physical layer ranging window during a later transport layer ranging window.

65. The apparatus of Claim 56 wherein the transmitter repeats transmitting a ranging request, the monitor unit repeats monitoring for a ranging response, and the control unit lengthens the physical layer ranging window (i) at least until the OLT receiver receives a ranging response during the physical layer ranging window, or (ii) until the control unit determines a ranging response is not within the transport layer window.

66. The apparatus of Claim 56 wherein, during a series of physical layer ranging windows, the control unit enables the OLT receiver for an amount of time equal to each physical layer ranging window of the series and resets the OLT receiver at a beginning of each physical layer ranging window of the series.

67. The apparatus of Claim 66 wherein the control unit shifts the series of physical layer ranging windows within the transport layer ranging window (i) at least until a ranging response is received during at least one physical layer ranging window of the series, or (ii) until the control unit determines a ranging response is not within the transport layer window.

68. The apparatus of Claim 66 wherein, in an event of the OLT receiver receives a ranging response in part during at least one physical layer ranging window, the control unit shifts the series of physical layer ranging windows an amount expected to result in the OLT receiver receiving a ranging response in full during one physical layer ranging window of the series during a later transport layer ranging window.

69. An apparatus for ranging an optical network terminal (ONT) in a passive optical network (PON), the apparatus comprising: means for transmitting a ranging request to an ONT in connection with a transport layer ranging window; means for monitoring for a ranging response from the ONT during at least one physical layer ranging window within the transport layer ranging window, the transport layer ranging window having a duration longer than the physical layer ranging window; and means for determining at least one metric associated with a ranging response for use in connection with upstream communications between the ONT and the OLT.

70. A method for ranging an optical network terminal (ONT) comprising: resetting a receiver of an optical line terminal (OLT) at about a time a ranging signal from an ONT is expected to be received to tolerate a fault condition otherwise affecting ranging of the ONT.

71. - The method of claim 70 further comprising: comparing ranging results of attempting to range the ONT using a standard ranging window and attempting to range the ONT by resetting the receiver of the OLT at about the time the ranging signal from the ONT is expected to be received; and notifying an operator of a fault condition based on comparing the ranging results.

72. The method of claim 70 wherein resetting the receiver of the OLT at about the time the ranging signal is expected to be received is based on an equalization delay assigned to the ONT previously.

73. The method of claim 70 wherein resetting the receiver of the OLT at about the time the ranging signal is expected to be received is based on a time previously determined by a successful ranging attempt.

74. The method of claim 70 further comprising: determining whether ranging the ONT is successful; and changing a time to reset the receiver of the OLT in an event ranging the ONT is unsuccessful.

75. The method of claim 74 further comprising storing the time to reset the receiver of the OLT in an event ranging the ONT is successful.

76. The method of claim 74 wherein changing the time to reset the receiver of the OLT includes adding a delay to the time when the OLT is expected to receive the ranging signal.

77. The method of claim 74 wherein changing the time to reset the receiver of the OLT includes subtracting a delay from the time when the OLT is expected to receive the ranging signal.

78. The method of claim 74 wherein changing the time to reset the receiver of the OLT includes iteratively incrementing a delay over a range of delays to delay the time to reset the receiver of the OLT and to compensate for variations in an equalization delay due to physical conditions expected to be experienced by an Optical Distribution Network (ODN).

79. The method of claim 74 wherein changing the time to reset the receiver of the OLT includes iteratively incrementing a delay by whole number delay increments to delay the time to reset the receiver of the OLT.

80. The method of claim 74 wherein changing the time to reset the receiver of the OLT includes iteratively incrementing a delay by random delay increments to delay the time to reset the receiver of the OLT.

81. The method of claim 74 wherein changing the time to reset the receiver of the OLT includes iteratively incrementing a delay by calculated delay increments to delay the time to reset the receiver of the OLT.

82. he method of claim 74 wherein changing the time to reset the receiver of the OLT includes iteratively incrementing a delay from minus bit-times to plus bit-times before or after the time the ranging signal from the ONT is expected to be received to delay the time to reset the receiver of the OLT.

83. A system for ranging an optical network terminal (ONT) comprising: a resetting unit configured to reset a receiver of an optical line terminal (OLT) at about a time a ranging signal from an ONT is expected to be received to tolerate a fault condition otherwise affecting ranging of the ONT.

84 The system of claim 83 further comprising: a comparing unit configured to compare ranging results of attempting to range the ONT using a standard ranging window and attempting to range the ONT by resetting the receiver of the OLT at about the time the ranging signal from the ONT is expected to be received; and a notification unit configured to notify an operator of the fault condition based on comparing the ranging results.

85. The system of claim 83 wherein the resetting unit is configured to reset the receiver of the OLT at a time based on an equalization delay assigned to the ONT previously.

86. The system of claim 83 wherein the resetting unit is configured to reset a receiver of the OLT at a time based on a time previously determined by a successful ranging attempt.

87. The system of claim 83 further comprising: a determining unit configured to determine whether ranging the ONT is successful; and a time delay changing unit configured to change a time to reset the receiver of the OLT in an event ranging the ONT is unsuccessful.

88. The system of claim 87 wherein the time delay changing unit is adapted to add a delay to the time when the OLT is expected to receive the ranging signal.

89. The system of claim 87 wherein the time delay changing unit is adapted to subtract a delay from the time when the OLT is expected to receive the ranging signal.

90. The system of claim 87 wherein the time delay changing unit is adapted to increment a delay in an iterative manner over a range of delays to delay the time to reset the receiver of the OLT and to compensate for variations in an equalization delay due to physical conditions expected to be experienced by an Optical Distribution Network (ODN).

91. A computer program product comprising a computer usable medium embodying computer usable code for ranging an optical network terminal (ONT), the computer program product including computer usable program code, which when executed by a processor, causes the processor to reset a receiver of an optical line terminal (OLT) at about a time a ranging signal from an ONT is expected to be received to tolerate a fault condition otherwise affecting ranging of the ONT.