JP2012213238A - Station side line concentrator and access control method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a station side line concentrator which can effectively use bandwidth in uplink.SOLUTION: PON IF parts 2 are connected to each of a plurality of PON networks. An uplink IF part 1 controls transmission timing of user data from the plurality of PON networks so that the user data which is received by the plurality of PON IF parts 2 is densely arranged in uplink. Thus, bandwidth can be effectively used in the uplink.

Description

  The present invention relates to a PON (Passive Optical Network) that is a medium-shared communication in which a plurality of home-side devices share a medium and transmit data, and in particular, an EPON that transmits data as an Ethernet (registered trademark) frame. The present invention relates to a station side line concentrator, an access control apparatus, and an access control method that accommodate multiple lines (Ethernet (registered trademark) PON) and multiplex them on an upper network (hereinafter referred to as uplink).

  In recent years, the Internet has become widespread, and users can access various information on sites operated in various parts of the world and obtain the information. Along with this, devices capable of broadband access such as ADSL (Asymmetric Digital Subscriber Line) and FTTH (Fiber To The Home) are rapidly spreading.

  As prior art related to this, there are techniques disclosed in Patent Document 1 and Non-Patent Document 1 below. In the station side line concentrator disclosed in Patent Document 1, a buffer memory, a receiving unit, a transmitting unit, and a PON IF are provided corresponding to each of the plurality of PON transmission paths A, B,. The control unit generates a control frame so that normal frames received via the PON transmission paths A, B,..., Z do not compete with each other, and sends them to the PON transmission paths A, B,.

  Non-Patent Document 1 defines an EPON access control protocol (MPCP (Multi-Point Control Protocol)) and an OAM (Operations, Administration and Maintenance) protocol, and a registration method for a new home-side device using an MPCP message. , A bandwidth allocation request, a transmission instruction, and the like are described.

JP 2004-253881 A

IEEE Std 802.3ah (registered trademark) -2004

  FIG. 21 is a diagram illustrating an example of an access control method in the station-side line concentrator disclosed in Patent Document 1. FIG. 21 shows a case where user data frames from the PON line 1 and the PON line 2 are multiplexed and transmitted to the uplink. By overlapping the laser off period in the PON line 1 and the laser on period in the PON line 2, the band is used effectively.

  However, in the uplink, the user data frame from the PON line 1 and the user data frame from the PON line 2 are not used as much as the laser on period, the synchronization period, and the management frame transmission period in the PON line 2. Bands are generated, and it cannot be said that the bands are fully utilized effectively.

  For example, when the transfer rate of the PON line 1 and the transfer rate of the PON line 2 are different, more unused bandwidths in the uplink are generated.

  The present invention has been made to solve the above problems, and an object of the present invention is to provide a station side line concentrator, an access control apparatus, and an access control method capable of effectively using a bandwidth in the uplink. That is.

  According to one aspect of the present invention, a station-side concentrator that accommodates a plurality of passive optical networks, the receiving means connected to each of the plurality of passive optical networks, and received by the plurality of receiving means Interface means for controlling the transmission timing of user data from a plurality of passive optical networks so that the arranged user data is densely arranged in the uplink.

  The interface means controls the transmission timing of user data from a plurality of passive optical networks so that user data received by a plurality of receiving means are densely arranged in the uplink, so that the bandwidth is effectively used in the uplink. It becomes possible to do.

  Preferably, the interface unit controls the transmission timing so that the difference between the reception timing by the reception unit and the transmission timing to the uplink is reduced.

  Since the interface means controls the transmission timing so that the difference between the reception timing by the reception means and the transmission timing to the uplink is small, the uplink band can be used more effectively.

  More preferably, all the uplink transmission rates of the passive optical network and the uplink uplink transmission rate are the same, and the interface means does not accumulate user data received by a plurality of receiving means in the buffer, and does not accumulate the user data. Send to uplink.

  Since the interface means transmits user data received by a plurality of receiving means to the uplink through without being accumulated in the buffer, the buffer becomes unnecessary and the cost of the apparatus can be reduced.

  More preferably, the interface means overlaps the burst signal of the first passive optical network and the burst signal of the second passive optical network, and the time for the overlap is equal to the second passive optical network. The reception timing is controlled so as to be at least a part of the time excluding the time corresponding to the user data frame period and the laser off period from the burst length.

  The interface means overlaps the burst signal of the first passive optical network with the burst signal of the second passive optical network, and the overlap time is determined from the burst length of the second passive optical network by the user. Since the reception timing is controlled to be at least a part of the time excluding the time corresponding to the data frame period and the laser off period, it is possible to continuously transmit user data frames in the uplink.

  More preferably, there is a passive optical network including a home-side device having an uplink transfer rate slower than an uplink uplink transfer rate among a plurality of passive optical networks, and the interface means has an uplink transfer rate higher than that of the uplink. When user data from a late home device is up-converted and transmitted to the uplink, the end of user data from the home device with a slow uplink transfer rate is the last of user data that is up-converted and sent to the uplink The reception timing is controlled so as to be earlier.

  There is a passive optical network including a home-side device having an uplink transfer rate slower than an uplink uplink transfer rate among a plurality of passive optical networks, and the interface unit is a home-side device having an uplink transfer rate slower than the uplink. When user data from is up-converted and transmitted to the uplink, the end of user data from the home side device having a low uplink transfer rate is earlier than the end of user data that is up-converted and transmitted to the uplink As described above, the reception timing is controlled as described above, so that the uplink band can be used more effectively.

  More preferably, the interface means overlaps the burst signal of the first passive optical network and the burst signal of the second passive optical network, and the time for the overlap is equal to the second passive optical network. The reception timing is controlled so as to be at least a part of the time excluding the time required to transmit the user data frame at the uplink speed and the time corresponding to the laser off period from the burst length.

  The interface means overlaps the burst signal of the first passive optical network with the burst signal of the second passive optical network, and the overlap time is determined from the burst length of the second passive optical network by the user. The reception timing is controlled to be at least a part of the time excluding the time required to transmit the data frame at the uplink speed and the time corresponding to the laser off period, so that user data frames are continuously transmitted in the uplink. It becomes possible to do.

  According to another aspect of the present invention, an access control apparatus for controlling reception timing of user data from a plurality of passive optical networks, wherein user data received from the plurality of passive optical networks is transmitted in the uplink. Means for controlling the transmission timing of user data from a plurality of passive optical networks so as to be densely arranged, and means for instructing transmission of a grant including the transmission timing to a home-side apparatus connected to the passive optical network .

  According to still another aspect of the present invention, a computer program for causing a computer to control reception timing of user data from a plurality of passive optical networks, received from the plurality of passive optical networks. A step of controlling the transmission timing of user data from a plurality of passive optical networks so that the user data is densely arranged in the uplink, and a grant including the transmission timing to a home device connected to the passive optical network And causing the computer to execute a step of instructing transmission.

  According to an aspect of the present invention, the interface unit controls the transmission timing of the user data from the plurality of passive optical networks so that the user data received by the plurality of receiving units is densely arranged in the uplink. Thus, it is possible to effectively use the bandwidth in the uplink.

It is a figure for demonstrating an example of the access control method in the station side concentrator in the 1st Embodiment of this invention. It is a block diagram which shows the structural example of the station side concentrator in the 1st Embodiment of this invention. It is a block diagram which shows the structural example of 4to1MUX31. It is a figure which shows the structural example in the case of implement | achieving the access control part 11 with software. 4 is a flowchart for explaining a procedure of initialization processing of an access control unit 11; It is a flowchart for demonstrating the procedure of an interruption process routine. It is a flowchart for demonstrating the procedure of a discovery process. It is a flowchart for demonstrating the procedure of a timeout process. It is a flowchart for demonstrating the procedure of a TOCn process. It is a flowchart for demonstrating the procedure of a message reception process. It is a flowchart for demonstrating the procedure of a register | resistor request process. It is a flowchart for demonstrating the procedure of a register | resistor confirmation process. It is a flowchart for demonstrating the procedure of a deregister process. It is a flowchart for demonstrating the procedure of a report reception process. It is a flowchart for demonstrating the procedure of a RTT update process. It is a flowchart for demonstrating the procedure of a bandwidth allocation process. 3 is a block diagram illustrating a configuration example of a PON communication unit 21. FIG. It is a figure for demonstrating an example of the access control method in the station side concentrator in the 2nd Embodiment of this invention. It is a figure which shows the example of a connection of the station side concentrator 100 in embodiment of this invention. It is a block diagram which shows the structural example of PON communication part 21 'in the 2nd Embodiment of this invention. 10 is a diagram illustrating an example of an access control method in the station-side line concentrator disclosed in Patent Literature 1. FIG.

(First embodiment)
FIG. 1 is a diagram for explaining an example of an access control method in the station-side line concentrator according to the first embodiment of the present invention. In the present embodiment, access control is performed so that the user data frame transmission period of the PON line 1 and the laser on period, synchronization period, report frame transmission period, and management frame transmission period of the PON line 2 overlap. It is. As a result, the user data frame from the PON line 1 and the user data frame from the PON line 2 are continuously transmitted to the uplink. Here, “continuously (or densely)” means that data is transmitted on the uplink at intervals of not less than the minimum IFG and less than the maximum overhead. The maximum overhead time is a time corresponding to a laser on period, a synchronization period, a report frame, and a management frame transmission period.

  In FIG. 1, the report frame, the management frame, and the user frame are transmitted in this order after the synchronization period, but the order of the frames is arbitrary. Access control may be performed so as to overlap by a time excluding the time corresponding to the laser data off period and the user data frame period including IFG and preamble from the burst length (time) of the PON line 2. In this case, the user data frames may overlap between the previous and next bursts, but the user data frames are prevented from colliding because they are arranged in the uplink after delay adjustment by a FIFO described later. In addition, the overlapping time may be a time shorter than the maximum in the above example.

  FIG. 2 is a block diagram illustrating a configuration example of the station-side line concentrator according to the first embodiment of the present invention. This station side concentrator includes an uplink IF (Interface) unit 1 and N PON IF units 2. In FIG. 2, N = 16 and 16 PON lines i (i = 1 to 16) are concentrated. However, the number of PON lines is not limited to this.

  The PON line is not limited to Passive. In general, branching is performed by an optical coupler, but an active optical switch may be used instead. Here, the PON line means a line that accommodates a plurality of home devices by MPCP.

  The case where the uplink is 10GE (10 Gigabit Ethernet (registered trademark)) and the PON line is 10GEPON (10 Gigabit Ethernet (registered trademark) PON) will be described. The uplink is GE (Gigabit Ethernet (registered trademark)). The PON line may be GEPON (Gigabit Ethernet (registered trademark) PON).

  The uplink IF unit 1 is received by the access control unit 11 that performs uplink access control of the uplink and each PON line, the uplink transmission / reception unit 12 that transmits / receives a frame to / from the uplink, and the uplink transmission / reception unit 12. The DEMUX 13 that outputs the downlink signal to each PON IF unit 2 and the concentrator 14 that multiplexes the uplink signal from each PON IF unit 2 and outputs the multiplexed signal to the uplink transmitting / receiving unit 12.

  The PON IF unit 2 terminates the data link between the PON transmission / reception unit 22 that transmits / receives signals to / from the PON line and the home-side device connected to the PON line, and relays the main signal to the uplink. PON communication unit 21 for performing

The DEMUX 13 transmits the downlink signal received by the uplink transceiver unit 12 to each PON.
Output to the IF unit 2. The DEMUX 13 may transmit the downlink signal from the uplink transmission / reception unit 12 to each PON IF unit 2 as it is, or analyze the contents of the downlink frame to identify the transmission destination of the frame, and selectively select the PON IF. You may make it send to the part 2. FIG. As an identification method, it may be identified by a VLAN (Virtual Local Area Network) tag or may be identified by a MAC (Media Access Control) address.

  The concentrator 14 is composed of a plurality of 4to1 MUXs 31 that multiplex four lines, and realizes a desired multiplexing number N by connecting the plurality of 4to1 MUXs 31 in multiple stages. In the present embodiment, 16 to 1 MUX is realized by connecting five 4 to 1 MUXs 14 in multiple stages.

  FIG. 3 is a block diagram illustrating a configuration example of the 4to1 MUX 31. This 4to1 MUX 31 receives four input units 41 connected to each PON IF unit 2, four FIFOs (First In First Out) 42 connected to each input unit 41, and a notification of the amount of accumulated frames from each FIFO 42. The MUX control unit 43 that controls the output of the frame stored in the FIFO 42, the multiplexer 44, and the output unit 45 that outputs the frame from the multiplexer 44 to the next-stage 4to1 MUX 31 or the uplink transmission / reception unit 12.

  The input unit 41 converts the high-speed serial signal from the PON IF unit 2 into a low-speed parallel internal signal and accumulates it in the FIFO 42. The internal signal can be 64 bits and 156.25 Mbps. The FIFO 42 notifies the MUX control unit 43 of the frame accumulation status.

  The MUX control unit 43 monitors the state of the FIFO 43, extracts a frame exclusively from the FIFO 2 in which the frame is stored, and outputs the frame to the output unit 45 while controlling the multiplexer 44.

  The output unit 45 converts the frame received from the multiplexer 44 into a high-speed serial signal, for example, an XAUI (10 Gbps Attachment Unit Interface) signal, and outputs the converted signal.

  As will be described later, since the access control unit 11 performs access control so that the uplink signal of the PON line can be directly output and output to the uplink, the capacity of the FIFO may be small. In other words, this FIFO avoids temporary collisions due to fluctuations depending on accuracy and differences in the arrangement of user data frames within a burst, and absorbs a slight shift in clock frequency between input and output. It is. If the access control unit 11 has an allowance for the allocation interval between the PON lines, the FIFO 42 of the 4To1MUX 31 and the MUX control unit 43 can be deleted and passed from the input unit 41 to the multiplexer 44. At this time, the multiplexer 44 may be controlled by the input unit 41 or may simply be a logical sum (a logic 0 signal is output during a period in which no data is output).

  The PON communication unit 21 and the access control unit 11 are connected by a control line, and the MPCP frame is transmitted and received by the PON communication unit 21, but the MPCP message passes through the PON communication unit 21 and the access control unit 11 terminates. . At this time, the PON communication unit 21 may add a time stamp at the time of PON reception or overwrite a time stamp at the time of PON transmission. By doing so, RTT and time stamp drift do not depend on the internal processing delay of the access control unit 11 and become more accurate. It is assumed that the access control unit 11 and each PON communication unit 21 have a common clock.

  FIG. 4 is a diagram illustrating a configuration example when the access control unit 11 is realized by software. Needless to say, the access control unit 11 may be realized by hardware.

  The access control unit 11 includes a central processing unit (CPU) 51, a read only memory (ROM) 52, a random access memory (RAM) 53, a shared memory 54, an input / output unit 55, and a clock timer 56. Including.

  The function of the access control unit 11 is realized by the CPU 51 executing the access control program stored in the ROM 52. The RAM 53 holds various information used when performing access control. Details of the various information will be described later.

  The input / output unit 55 is connected to the PON IF unit 2 and controls message input / output. When the message signal received by the PON IF unit 2 is input, the input / output unit 55 converts the message signal into an internal format and adds it to the input message queue (Qin) of the shared RAM 54. At this time, if Qin changes from an empty state to a non-empty state, an interrupt is issued to the CPU 51.

  Further, the input / output unit 55 outputs the message stored in the output message queue (Qeg) in the shared RAM 54 to the PON IF unit 2 in accordance with an instruction from the CPU 51.

  The clock timer 56 has a function as a clock for measuring the current time (ctime) and a counter function for counting a predetermined time. The ctime value of the watch timer 56 is referred to by the CPU 51.

  When the clock timer 56 functions as a discovery timer (TD) and a logical link timer (TLij: i = 1, 2,..., N: jε {LLID of PON IF i} described later), the count value is the CPU 51. When the value set by is reached, the CPU 51 is interrupted to notify the count out.

  5 to 15 are flowcharts for explaining the processing procedure of the access control unit 11. Here, various information held in the RAM 53 will be described.

  The final allocation time (TEi: i = 1, 2,..., N) is based on an overlap between the laser off period (Toff) and the laser on period (Ton), and is the time obtained by subtracting Toff minutes from the final time of the grant period. It is. In FIG. 1, for example, the end time (laser-off start time) of the user data frame of the PON line 1 is the final allocation time. This final allocation time is held in the RAM 53 for every N PON IFs.

  When the grant period does not include the minimum IFG (inter-frame gap) after the last frame, the minimum IFG is added to the final allocation time. Note that Toff may not be deducted on the assumption that there is no overlap.

  The total final allocation time (TEz) is a time when only one is held in the RAM 53, and is basically the final allocation time of the last allocated PON IF k, but there may be exceptions. This exception will be described later.

  The discovery order list (dLST) holds the order of the PON IFs for performing the discovery process in a list format, and is set as 1 → 2 →... → N → 1, for example.

  The logical link information (LLTij: i = 1, 2,..., N: jε {LLID of PON IF i}) is information held in the RAM 53 in a table format for each logical link, and the logical link state (LLstat) ), Report state (RPstat), management frame report information (RPm), user data report information (RPu), and RTT (Round Trip Time) information. The report information may simply reflect the latest report, but may be in a queue format.

  The logical link status (LLstat) is information indicating the status of the logical link, and includes unregistered, registered, and registered status.

  The report state (RPstat) is information indicating whether the report of the logical link is valid or invalid.

  The management frame report information (RPm) is information indicating the time required for sending out the reported management frame.

  The user data report information (RPu) is information indicating the time required to transmit the reported user data frame.

  The access control unit 11 is provided with an allocation order between PON IFs and an allocation order of logical links within the PON IF. Allocation order may be simple round robin, round robin considering weight, round robin hierarchized to differentiate allocation period, skipping allocation to logical links that have been allocated too much based on bandwidth calculation, etc. It may be.

  The allocation order is organized as a general scheduling technique, and the present invention does not depend on a specific scheduling technique. However, the effect of the present invention can be enhanced by avoiding consecutive assignments to the same PON IF.

  Further, the present invention is independent of the timing of allocation, and may be allocated every time a report frame is received, or may be allocated after collecting report frames in a concentrated manner. .

  FIG. 5 is a flowchart for explaining the procedure of the initialization process of the access control unit 11. First, the input message queue Qin and the output message queue Qeg are emptied, the logical link information LLTij (i = 1, 2,..., N: j∈ {LLID of PON IF i}) is all NULL, and the final allocation time TEi is set. All are set to the current time ctime, the entire final allocation time TEz is set to ctime, and 1 → 2 →... → N → 1 is set in the discovery order list dLST (S11).

  Next, a value obtained by dividing the discovery cycle discovery_interval by N is set in the discovery timer TD (S12). This TD is an interval when the discovery process is sequentially performed for N PON lines, and the discovery process for the next PON line is performed every time TD elapses.

  Next, the process waits until an interrupt occurs (S13). When an interrupt occurs (S13, Yes), the process proceeds to an interrupt processing routine (S14).

  FIG. 6 is a flowchart for explaining the procedure of the interrupt processing routine. First, the process branches depending on the type of interrupt (S21). If the interrupt indicates that Qi is not empty (S21, Qin is not empty), the process proceeds to message reception processing (S22). If the interrupt indicates TD countout (S21, TD expires), the process proceeds to discovery processing (S23). On the other hand, when the interrupt indicates the count out of TLij (S21, TLij has expired), the process proceeds to timeout processing (S24).

  Note that the priority of interrupts is highest when Qin is not empty, and becomes lower in the order of TD expiration and TLij expiration.

  FIG. 7 is a flowchart for explaining the procedure of the discovery process. First, assuming that the head of dLST is k, a discovery gate message for PON IF k is constructed and placed in Qeg. At this time, the start time is based on TEk or the later of the current time (S31).

  Next, the end of the discovery period is set as a new TEk, and dLST is advanced to the next PON IF unit (S32). Then, the discovery timer TD is set (S33), and the process ends.

  FIG. 8 is a flowchart for explaining the procedure of the timeout process. First, it is determined whether LLstat of LLTij is being registered (S41). If not registered (S41, No), the time-out information is recorded in LLTij. Then, the number of timeouts TOCij within a predetermined period is obtained with reference to the past timeout information. Here, the maximum allowable value is TOCmax (S42).

  Next, TOCij and TOCmax are compared. If TOCij is equal to or lower than TOCmax (S43, No), the process proceeds to TOCn processing shown in FIG. If TOCij is larger than TOCmax (S43, Yes), deregister processing is performed (S44).

  In step S41, if the LLstat of LLTij is being registered (S41, Yes), deregister processing is performed (S44).

  FIG. 9 is a flowchart for explaining the procedure of the TOCn process. First, a gate message for LLIDj of PON IF i is constructed and placed in Qeg. At this time, a report compulsory instruction is issued by setting a report compulsory flag. The start time is based on TEi or the current time, and the grant length is an amount that can transmit only a report frame (S51).

  Then, the start of laser off of the grant is set as a new TEi, and a time when a certain amount of deviation time is expected is set in the timer TLij (S52), and the process is terminated.

  FIG. 10 is a flowchart for explaining the procedure of the message reception process. Note that the time stamp recorded in the message by the home side device is T1, and the time stamp added by the PON communication unit 21 when the message is received is T2.

  First, the head message is extracted from Qin (S61), and if the message type is a register request (S62, register request), register request processing is performed (S63). If the message type is register confirmation (S62, register confirmation), register confirmation processing is performed (S64).

  If the message type is a deregister request (S62, deregister request), deregister processing is performed (S65). If the message type is a report (S62, report), a report reception process is performed (S66), and a bandwidth allocation process is performed (S67).

  When the process of steps S63 to S65 or S67 ends, it is determined whether or not Qin is empty (S68). If Qin is not empty (S68, No), the process returns to step S61 to perform the subsequent processing. If Qin is empty (S68, Yes), the process is terminated.

  FIG. 11 is a flowchart for explaining the procedure of the register request process. Note that this register request is a request from a home-side device whose MAC address of PON IF i is m.

  First, a new LLID is assigned to this PON IF for which a register request has been made, and is set as LLIDj. Then, registering is set in LLstat of LLTij, RPstat is set to NULL, and (T2-T1) is set in RTT (S71).

  Next, a register message for PON IF i is constructed and placed in Qeg. At this time, DA (Destination Address) is m, and LLID is j (S72). This LLID is described in a register message and notified to the home side device, and the LLID described in the header portion of the frame is a broadcast LLID.

  Next, a gate message for PON IF i, LLIDj is constructed and placed in Qeg. At this time, the start time is based on TEi or the current time. Further, the grant length is set to an amount capable of transmitting only the register confirmation frame (S73).

  Finally, the start of laser off of the grant is set as a new TEi, and a time when a certain amount of blur time is expected is set in the timer TLij (S74), and the process is terminated.

  FIG. 12 is a flowchart for explaining the procedure of the register confirmation process. First, RTT is recalculated and RTT update processing of the logical link is performed. At this time, if the drift exceeds the specified value (S81, NG), the process is terminated.

  If the drift is within the specified value (S81, OK), LLstat of LLTij is registered, a gate message for PON IF i, LLIDj is constructed, and is put in Qeg. At this time, a report compulsory instruction is issued by setting a report compulsory flag. The start time is based on TEi or the current time. Preferably, it is better to arrange the same PON IF at a position where an interval is opened with reference to the past reservation order. The grant length is set so that only the report frame can be transmitted (S82).

  Next, the start of laser off of the grant is set as a new TEi, and a timer TLij is set at a time when a certain amount of blur time is expected (S83), and the process is terminated.

  FIG. 13 is a flowchart for explaining the procedure of the deregister process. Note that PON IF i and LLIDj are deregistered. First, a deregister message for PON IF i, LLIDj is constructed and placed in Qeg (S91). Then, LLstat of LLTij is set to NULL (S92), and the process ends.

  FIG. 14 is a flowchart for explaining the procedure of the report reception process. It is assumed that a report from PON IF i, LLIDj is received. First, RTT is recalculated and RTT update processing of the logical link is performed. At this time, if the drift exceeds the specified value (S101, NG), the process is terminated.

  If the drift is within the specified value (S101, OK), RPstat of LLTij is validated, RPm is updated to the management frame report queue queue0_report, and RPu is updated to the total of the user frame report queue queueK_report (S102). ), The process is terminated.

  FIG. 15 is a flowchart for explaining the procedure of the RTT update process. First, the new RTT is updated to (T2-T1) (S111), and it is determined whether or not the difference between the new RTT and the original RTT is greater than the maximum drift value DRIFTmax (S112).

  If the difference between the new RTT and the original RTT is equal to or less than DRIFTmax (S112, No), a new RTT is set for the LTTij RTT (S113), and the RTT update process is assumed to be successful (OK). If the difference between the new RTT and the original RTT is greater than DRIFTmax (S112, Yes), deregister processing is performed (S114), and it is assumed that the RTT processing has failed (NG).

  FIG. 16 is a flowchart for explaining the procedure of the bandwidth allocation process. It is assumed that a bandwidth is allocated to PON IF i and LLIDj. First, the sum of the time RPm required for sending the LLTij management frame and the time REPORT_length required for sending the report frame is ML, the time RPu required for sending the user data frame of LLTij is UL, and the overhead time at the beginning of the burst ( Laser on period Ton + synchronization period SyncTime) is set to OVL. Then, the smaller value of the sum of ML, UL, OVL and Toff and the upper limit value GLmax of the grant length is set as the grant length GL.

  A time obtained by subtracting OVL and ML from the overall final allocation time TEz is defined as TSz. Let TSi be the sum of the final allocation time TEi of PON IF i and the burst gap burst_gap. Also, let TSc be the sum of the current time ctime, RTT, and the processing time proc_time of the home device. The latest time among TSz, TSi and TSc is set as TS (S121).

  Next, the time obtained by subtracting the current time ctime from the TS is compared with the system constant TooFar for preventing the grant from leading too far (S122). If TS-ctime is not smaller than TooFar (S122, No), the process waits or sleeps until TS-ctime becomes larger than TooFar (S123), returns to step S121, and repeats the subsequent processing.

  If TS-ctime is smaller than TooFar (S122, Yes), a gate message for PON IF i and LLIDj is constructed and placed in Qeg. At this time, a report compulsory instruction is issued by setting a report compulsory flag. Further, the start time is (TS-RTT), and the grant length is GL (S124). Then, (TS + GL-Toff) is set in TEi (S125).

  Next, it is determined whether or not UL is larger than 0 (S126). If UL is 0 (S126, No), the process proceeds to step S128 without updating TEz. If UL is larger than 0 (S126, Yes), TEi is set to TEz (S127).

  Finally, the timer TLij is set at the time when a certain amount of shake time is expected in TEi (S128), and the process is terminated.

  Note that the ML may be included in the UL with priority given to simplicity. Further, the grant start time may be intentionally shifted from the best one.

  At the end of processing, the report of the logical link of LLTij is invalidated. However, when the report information is in the queue format and there is the next entry, the report queue is updated while the report is valid.

  FIG. 17 is a block diagram illustrating a configuration example of the PON communication unit 21. The PON communication unit 21 is stored in an input unit 61 for inputting a signal (internal format signal) from the DEMUX 13, a FIFO 1 (62) for storing a signal input to the input unit 61, and a FIFO 1 (62). A transmission processing unit 63 that extracts data and converts the signal format and then outputs the data to the PON transmission / reception unit 22; a reception processing unit 64 that determines and selectively outputs the type of frame input from the PON transmission / reception unit 22; A management frame processing unit 65 that receives and processes a management frame from the reception processing unit 64, an MPCP frame processing unit 66 that receives and processes an MPCP frame from the reception processing unit 64, and a user data frame received from the reception processing unit 64 An output unit 68 that converts the format and outputs the converted data to the concentrator unit 14. The PON communication unit 21 is generally realized by one or a plurality of LSIs.

  When data is stored in the FIFO 1 (62), the transmission processing unit 63 extracts the data, converts the signal format, and outputs the data to the PON transmission / reception unit 22, but the management frame processing unit 65 or MPCP frame processing If the unit 66 has a frame to be transmitted, priority is given to the output of that frame. At this time, the MPCP frame has the highest priority. Further, the time stamp may be overwritten if necessary.

  The management frame processing unit 65 processes management frames such as OAM, authentication, and encryption settings. Since the content of this process itself is irrelevant to the present invention, detailed description thereof is omitted. A management frame that is necessary as a result of this processing or voluntarily is formed, and a transmission request is sent to the transmission processing unit 63. Further, if a certain logical link cannot be maintained as a result of the processing, the MPCP frame processing unit 66 is instructed to deregister the logical link.

  The MPCP frame processing unit 66 performs processing according to the type of MPCP frame input from the reception processing unit 64. If the MPCP frame is a register request frame, authentication is performed. If the authentication result is correct, the register request message is transferred to the access control unit 11. Note that the message may be transferred to the access control unit 11 without performing authentication. Alternatively, after the message is first transferred and registered in the access control unit 11, the management frame processing unit 65 may perform authentication using another protocol.

  When the MPCP frame is a deregister request frame or a register confirmation frame, these messages are transferred to the access control unit 11. At this time, a time stamp may be added.

  When the MPCP frame is a report frame, the time stamp drift is checked. If the MPCP frame is violated, a deregister message is transferred to the access control unit 11. Note that this check may be performed by the access control unit 11. If there is no violation, the report message is transferred to the access control unit 11.

  The MPCP frame processing unit 66 may add a time stamp when transferring these messages to the access control unit 11.

  The MPCP frame processing unit 66 monitors the report frame interval for each logical link and determines connectivity. When it is determined that the logical link has been broken, or when a deregister instruction is received from the management frame processing unit 65, the deregister message of the logical link is transferred to the access control unit 11. Note that this processing may be performed by the access control unit 11.

  Note that the PON communication unit 21 may have a local access control unit (not shown). This local access control unit independently performs uplink access control of the PON line, and the access control method follows the conventional technology. The PON communication unit 21 and the MPCP frame processing unit 66 can switch whether MPCP messages are exchanged with the outside, for example, the access control unit 11 of the uplink IF unit 1 or the local access control unit, depending on the setting.

  As described above, according to the station side line concentrator in the present embodiment, the access control unit 11 performs bandwidth allocation of each PON so that user data frames from the PON line are continuously transmitted in the uplink. As a result, the bandwidth in the uplink can be used effectively.

(Second Embodiment)
FIG. 18 is a diagram for explaining an example of an access control method in the station-side line concentrator according to the second embodiment of the present invention. FIG. 18 shows a case where the transfer rate of the upstream burst signal of the PON line 1 is 10 Gbps, and the transfer rate of the upstream burst signal of the PON line 2 is 1 Gbps. In the present embodiment, access control is performed so that the user data frame transmission period of the PON line 1 and the user data frame transmission period of the PON line 2 overlap each other, so that the bandwidth in the uplink can be used more effectively. To do.

  Since the transfer rate of the PON line 2 is slower than that of the PON line 1, the user data frame sequence of the PON line 2 is up-converted to the uplink speed, and the user data frame is arranged in the uplink without a gap.

  FIG. 19 is a diagram illustrating a connection example of the station-side line concentrator 100 according to the embodiment of the present invention. As shown in FIG. 19A, this system includes a station-side concentrator 100, a home-side device A 101-1 having a transfer rate α, a home-side device B 101-2 having a transfer rate β, and a plurality of optical couplers 102. Including. The transfer rate α is different from the transfer rate β.

  As shown in FIG. 19B, the upstream optical signal (wavelength O), the downstream optical signal having the transfer rate α (wavelength C), and the downstream optical signal having the transmission rate β (wavelength L) are wavelength-multiplexed. Independent signal transmission is possible on one PON line. On the other hand, the wavelength of the upstream optical signal having the transfer rate α overlaps the wavelength of the upstream optical signal having the transfer rate beta. Therefore, as shown in FIG. 19 (c), independent signal transmission is performed on one PON line by time division multiplexing so that the optical burst signals do not collide with each other.

  For example, the home apparatus A 101-1 is for GE-PON, the home apparatus B 101-2 is for 10 GEPON, the transfer rate α is 1 Gbps, and the transfer rate β is 10 Gbps. These transfer rates are generic names, and the actual transfer rate slightly changes before and after the code format and code processing. When an error correction code is used for the PON line and the actual effective data rate decreases, the effective data rate may be considered as the rate of the PON line.

  In this embodiment, the wavelength O is 1260 to 1360 nm, the wavelength C is 1480 to 1500 nm, the wavelength L is 1574 to 1580 nm, and α <β.

  FIG. 20 is a block diagram illustrating a configuration example of the PON communication unit 21 ′ according to the second embodiment of the present invention. Compared with the PON communication unit 21 in the first embodiment shown in FIG. 17, FIFO 3 (71) and FIFO 4 (72) in which user data frames with different rates are accumulated are added, and the PON transmission / reception unit The only difference is that the functions of the input unit, transmission processing unit, reception processing unit, and output unit are different. Therefore, detailed description of overlapping configurations and functions will not be repeated.

  The PON transmitting / receiving unit 22 ′ can simultaneously transmit an optical signal having a wavelength C / rate α and an optical signal having a wavelength L / rate β. The PON transmission / reception unit 22 ′ receives the rate α signal and the rate β signal from the transmission processing unit 63 ′, converts them into an optical signal of wavelength C / rate α and an optical signal of wavelength L / rate β, and a PON line Output to.

  Further, the PON transceiver unit 22 ′ can receive a dual rate optical burst signal having a wavelength O. The PON transmission / reception unit 22 'converts the optical signal received from the PON line into an electrical signal and outputs the electrical signal to the reception processing unit 64'.

  In FIG. 20, synchronization processing corresponding to the rate α and rate β is performed in the PON transmission / reception unit 22 ′ and output to the reception processing unit 64 ′ as an electrical signal for each rate. These synchronization processes may be performed by the reception processing unit 64 '. The rate distinction may be automatically recognized from the received signal, or the rate for reception may be determined in advance based on information obtained by granting the transmission of the burst.

  The input / output signal between the DEMUX 13 and the concentrator 14 of the uplink IF unit 1 is a single rate according to the uplink rate, for example, 10 Gbps of 10 Gigabit Ethernet (registered trademark). Here, it is assumed that it is equal to β.

  The input unit 61 ′ receives a signal from the DEMUX 13 of the uplink IF unit 1, determines the logical link and its rate, and outputs a signal to the FIFO corresponding to the rate. This output signal is converted into an internal signal format corresponding to the rate.

  When data is stored in the FIFO1 (62) or the FIFO3 (71), the transmission processing unit 63 'extracts the data individually, converts the signal format, and transfers the data to the PON transmission / reception unit 22'. If the MPCP frame processing unit 66 or the management frame processing unit 65 has a frame to be transmitted, these frames are prioritized, but the rate is determined by the destination logical link of the frame, and the signal of that rate is added. Inserted. When data is flowing in the signal, it is inserted at a frame break.

  The reception processing unit 64 ′ determines the type of frame at each rate from the PON transmission / reception unit 22 ′, and accumulates it in the FIFO 4 (72) in the case of a user data frame at the rate α. The reception processing unit 64 ′ may filter inappropriate frames based on the reception time and LLID.

  In addition to receiving the user data frame from the reception processing unit 64 ′, the output unit 68 ′ extracts data from the FIFO 4 (72), converts the signal format, and outputs it to the concentrating unit 14 of the uplink IF unit 1. The start timing for extracting data from the FIFO 4 (72) is instructed by the MPCP frame processing unit 66.

  Hereinafter, differences in the processing of the access control unit 11 from the first embodiment will be described. In the present embodiment, the grant length represents time. That is, when the same grant length is given to the 1G and 10G logical links, the 10G logical link can transmit 10 times as much information as the 1G logical link.

  Discovery gates are sent in rate units. Therefore, the signal rate of register requests received during the discovery period is reserved and becomes a single rate. Further, the rate type is added to the logical link information, and when the register request is received, the rate is recorded in the rate type. It is assumed that the uplink rate is β, and there is a logical link having the same rate (β) and a logical link having a slower rate (referred to as α).

  When the logical link (L) rate type of PON IF k to be allocated next is α, the TSz calculation method is as follows. That is, since the grant length is a time unit, the calculation of the grant length is the same as that in the first embodiment. The time required to transmit the amount of data transmitted at the rate α at the UL time is defined as ULB. Let TEzn be the time when ULB is added to TEz. Let TEzn be the starting point of laser off, and let TSz be the time obtained by subtracting the grant length and burst overhead time (Ton + SyncTime).

  When the grant information is sent to the PON IF unit 2, the rate type and the output start possible time TUS are added. TUS is the new overall final allocation time minus ULB. The TUS is notified to the output unit 68 ′ of the PON communication unit 21 ′, and instructs the timing for extracting data from the FIFO 4 (72). This is to prevent FIFO underrun from occurring on the read side because writing to the FIFO 4 (72) is slow.

  In FIG. 20, the signals are separated for each rate in the upstream direction of the PON communication unit. However, since the upstream signals of the PON line are multiplexed as shown in FIG. 19, they can be integrated into a path via one FIFO. It is. In this case, the reception processing unit 64 'adjusts the writing speed in the FIFO according to the rate of the received signal.

  When α> β, the calculation of TSz and the update of the overall final allocation time are as follows. That is, since the grant length is a time unit, the calculation of the grant length is the same as that in the first embodiment. The time required to transmit the data amount transmitted at the rate α at the ML time and the UL time at the rate β is defined as MLB and ULB, respectively. A time obtained by subtracting OVL and MLB from the overall final allocation time TEz is defined as TSz.

  Assuming that the selected start time is TS, the time when OVL, MLB, and ULB are added to TS is set to TE. The new overall final allocation time is set to TE when the main allocation is UL> 0. When UL = 0, the last allocation time is not updated.

  As described above, in the station side concentrator in the present embodiment, even when there is a home-side device whose uplink rate is lower than the uplink rate, the uplink burst signal from the home-side device is transmitted early. The access control is performed, the user data frame sequence from the home side device having a low transfer rate is stored in the FIFO, the timing for sending it to the uplink is adjusted, the up-converted to the uplink speed is sent, and the uplink is sent. Since user data frames are arranged without gaps, it is possible to use the bandwidth in the uplink more effectively.

  The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

  DESCRIPTION OF SYMBOLS 1 Uplink IF part, 2 PON IF part, 11 Access control part, 12 Uplink transmission / reception part, 13 DEMUX, 14 Concentration part, 21 PON communication part, 22 PON transmission / reception part, 314 to 1 MUX, 41 Input part, 42, 62, 71, 72 FIFO, 43 MUX control unit, 44 multiplexer, 45 output unit, 51 CPU, 52 ROM, 53 RAM, 54 shared memory, 55 input / output unit, 56 clock timer, 61 input unit, 63 transmission processing unit, 64 reception Processing unit, 65 management frame processing unit, 66 MPCP frame processing unit, 68 output unit, 100 station side concentrator, 101 home side device.

Claims (8)

A station side concentrator accommodating a plurality of passive optical networks,
A plurality of receiving means connected to each of the plurality of passive optical networks;
And an interface unit that controls transmission timing of user data from the plurality of passive optical networks so that user data received by the plurality of receiving units are densely arranged in the uplink.   The station side concentrator according to claim 1, wherein the interface unit controls the transmission timing so that a difference between a reception timing by the receiving unit and a transmission timing to the uplink is reduced. All uplink transmission rates of the passive optical network and uplink uplink transmission rates are the same;
The station side concentrator according to claim 2, wherein the interface means transmits the user data received by the plurality of receiving means to the uplink through without being accumulated in a buffer.   The interface means overlaps the burst signal of the first passive optical network and the burst signal of the second passive optical network, and the overlapping time is determined by the burst length of the second passive optical network. The station side concentrator according to claim 3, wherein the reception timing is controlled so as to be at least a part of a time excluding a time corresponding to a user data frame period and a laser off period. There is a passive optical network including a home side device having an uplink transfer rate slower than the uplink uplink transfer rate among the plurality of passive optical networks,
When the interface means up-converts user data from a home device with an uplink transfer rate slower than that of the uplink and transmits the user data to the uplink, the interface means transmits the last user data from the home device with a slow uplink transfer rate. 3. The station side concentrator according to claim 2, wherein the reception timing is controlled so as to be earlier than the end of user data transmitted up-link after the up-conversion.   The interface means overlaps the burst signal of the first passive optical network and the burst signal of the second passive optical network, and the overlapping time is determined by the burst length of the second passive optical network. 6. The station-side concentrator according to claim 5, wherein the reception timing is controlled so as to be at least a part of a time excluding a time required to transmit a user data frame at an uplink rate and a time corresponding to a laser off period. . An access control device for controlling reception timing of user data from a plurality of passive optical networks,
Means for controlling the transmission timing of user data from the plurality of passive optical networks so that user data received from the plurality of passive optical networks are densely arranged in the uplink;
Means for instructing transmission of a grant including the transmission timing to a home-side apparatus connected to the passive optical network. A computer program for causing a computer to execute control of reception timing of user data from a plurality of passive optical networks,
Controlling the transmission timing of user data from the plurality of passive optical networks so that user data received from the plurality of passive optical networks are densely arranged in the uplink;
A computer program for causing a computer to execute a transmission instruction of a grant including the transmission timing to a home-side device connected to the passive optical network.

Images