Classifications

• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04J—MULTIPLEX COMMUNICATION
  • H04J3/00—Time-division multiplex systems
  • H04J3/02—Details
  • H04J3/08—Intermediate station arrangements, e.g. for branching, for tapping-off
  • H04J3/085—Intermediate station arrangements, e.g. for branching, for tapping-off for ring networks, e.g. SDH/SONET rings, self-healing rings, meashed SDH/SONET networks
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L12/00—Data switching networks
  • H04L12/28—Data switching networks characterised by path configuration, e.g. LAN [Local Area Networks] or WAN [Wide Area Networks]
  • H04L12/42—Loop networks
  • H04L12/437—Ring fault isolation or reconfiguration
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L12/00—Data switching networks
  • H04L12/64—Hybrid switching systems
  • H04L12/6402—Hybrid switching fabrics

Definitions

• This application relates to communicating data between TDM and packet based networks.
• a time division multiplexing ("TDM") medium carries a signal that represents multiple constituent signals in defined time slots.
• a TI signal is a TDM signal that may represent up to 24 constituent voice signals.
• the TI signal has a transmission rate of 1.54 Mbps, with each constituent voice signal having a rate of 64 kbps.
• TI relies on a specific kind of TDM known as synchronous TDM, according to which a frame is periodically generated and includes a constant number of time slots, wherein each time slot has a constant length.
• the time slots can be identified by position in the frame, e.g., time slot 1, time slot 2, and each constituent signal is assigned a reserved time slot.
• a TI frame has 24 time slots per frame, numbered 1-24, and has one leading bit for framing purposes. Each time slot can carry 8 bits, such that the frame length is 193 bits.
• the frame repetition rate is 8000 Hz, which results in the 1.54 Mbps transmission rate for a TI signal.
• two TI signals can carry signals representing 24 simultaneous calls between two endpoints such as an office building and a telephone company switching office. At each endpoint, for each call, the incoming ("listening") portion of the call may be extracted from the incoming TI signal and the outgoing ("speaking") portion of the call may be inserted into the outgoing TI signal. Since each call is assigned a reserved time slot in each direction, the call signals are delivered by the TI signals with little or no latency.
• a packet-switching medium operates as follows. A sequence of data is sent from one host to another over a network. The data sequence is segmented into one or more queued packets, each with a data load and with a header containing control information, and each packet is routed through the network.
• a common type of packet switching is datagram service, which offers little or no guarantees with respect to delivery. Packets that may belong together logically at a higher level are not associated with each other at the network level. A packet may arrive at the receiver before another packet sent earlier by the sender, may arrive in a damaged state (in which case it may be discarded), may be delayed arbitrarily (notwithstanding an expiration mechanism that may cause it to be discarded), may be duplicated, or may be lost.
• a well known example of a packet-switching medium is the Internet.
• Typical concerns include reliability issues that the packets reach their destination without being dropped and predictability issues that the packets reach their destination without excessive delay.
• reliability and predictability are important. It has been found that dropped packets can distort voice quality if there is a considerable delay in the retransmission of those packets, and that excessive delay in the delivery of the packets can introduce inconsistencies in voice quality.
• a method and a system are provided for use in emulating time division multiplexing connections across a packet switch network.
• a method of packet circuit emulation (“PACE") is provided for use in communicating data between a time division multiplexing (TDM) network and a packet based network.
• a clock signal is derived from data packets containing data originally transmitted over the TDM network.
• Outgoing TDM data is obtained using the data packets and the clock signal.
• the outgoing TDM data has timing characteristics of the data originally transmitted over the TDM network.
• Data extracted from the data packets may be collected in a first-in first-out buffer memory (“FIFO").
• the clock signal may be derived in part from fill level information for the FIFO.
• an adaptive clock recovery mechanism is provided that is well suited for use with packet based networks. Adaptive heuristics are provided to reduce end to end latency. Jitter and wander are reduced, to approach or meet Bellcore requirements. Also provided are automatic recovery from protection switches and automatic adaptation to changes in the network such as increases in load. A capability of operation with or without a stratum clock is provided.
• FIGS. 1-7 are block diagrams of packet and TDM data processing systems.
• a method and a system are provided for use in emulating time division multiplexing connections across a packet switch network.
• a method of packet circuit emulation (“PACE") is provided for emulating a time division multiplex (“TDM") connection across a packet based network.
• the method and system may be provided in a node having a TDM connection and a packet network connection, such as an Optical Services Activation Platform (“OSAP”) node by Appian Communications, Inc., described in co-pending U.S. Patent Application Serial No. 09/546,090, which is hereby incorporated by reference in its entirety.
• OSAP Optical Services Activation Platform
• Fig. 7 illustrates an example system 710 in which a first OSAP 714 provides a connection between a TDM system 712 and a packet based network system 716, and a second OSAP 718 provides a connection between packet based network system 716 and a TDM system 720.
• PACE to support data transfer from TDM system 712 to TDM system 720, PACE relies on TDM to packet logic in first OSAP 714 and packet to TDM logic in second OSAP 718.
• PACE maps a TI data stream onto a constant packet stream and at the other end de-maps the packet stream onto another TI data stream which PACE provides with at least some of the clocking characteristics of the original TI stream.
• a physical user interface is provided for PACE that is a TI interface having TI compliant electrical and timing characteristics.
• TI is an example; other technologies may be used instead of, or in addition to, TI.
• PACE addresses a jitter problem inherent in packet networks and recovers the source clock. Variations in the phase of a signal are known as jitter or wander. Rapid variations are known as jitter. Slow variations are known as wander.
• jitter refers to periodic or stochastic deviations of the significant instants of a digital signal from its ideal, equidistant values. Otherwise stated, jitter is produced when the transitions of a digital signal occur either too early or too late when compared to a perfect squarewave.
• TDM data has relatively stringent requirements with respect to jitter and wander, compared to packet-based networks.
• Packet based networks tend to exhibit more jitter than TDM networks due to packetization and queuing; TDM networks such as SONET and TI are less susceptible to jitter as a result of carrying data in time slots that are nearly evenly spaced in time.
• the act of transferring TDM data to a packet introduces jitter because packet transmission causes a block of data to be sent all at once instead of in a stream of evenly spaced portions.
• Queuing introduces jitter by introducing variable delays in the transmission of TDM packets. TDM packets may have to wait to be sent after other, large packets are sent. Congestion and variations in data load can cause changes in delay, increasing the jitter.
• synchronization is affected by a source clock that produces an electrical timing signal to be distributed.
• the source clock is a highly stable clock, generally referred to as a stratum 1 source.
• a stratum 1 source is a highly reliable timing source, e.g., one that has a free running inaccuracy on the order of one-part in 10 11 .
• a TDM network provides traceability of timing back to a stratum 1 source to help ensure timing stability.
• timing distribution in rings and fully-connected meshes are considered as an overlay of linear paths to each mode from a master clock, i.e., as a tree structure.
• Each location with a stratum 1 clock source has other locations subtended from it, with respect to a timing distribution.
• Clock quality is defined by several parameters, such as filtering capability, transient suppression, and holdover accuracy during any loss of stratum 1 connection.
• These subtended, or 'slave,' clocks form a quality hierarchy and are described as stratum 2, 3E, and 3.
• stratum 2, 3E, and 3 Under normal circumstances (i.e., no network failures), every slave clock in the network is synchronized, or traceable, to a stratum 1 clock. Accordingly, each slave clock has at least two potential reference inputs, typically prioritized as primary and secondary.
• PACE includes two main functions: a TDM to packet conversion function and a packet to TDM conversion function.
• FIG. 1 shows an example data flow in TDM to packet conversion logic 110, which has the following characteristics.
• TI data 114 is collected (“recovered") from a Line Interface Unit ("LIU") 112 in conjunction with a recovered clock 116 associated with the recovered TI data.
• LIU Line Interface Unit
• the LIU may provide optical to electrical conversion, framing detection and synchronization, and the ability to extract timing from the incoming signal, and use it as a receive clock for incoming data.
• the LIU may detect alarms, frame errors, parity errors, and far end errors, including framing errors.
• TI framer logic 122 monitors the TI signal for error conditions.
• each packet has a fixed size payload, i.e., each packet has the same number of bits.
• the fixed payload size is an example; a payload size that is not fixed may be used instead of, or in addition to, the fixed payload size.
• Packetizing logic 124 handles peer signaling (e.g., sequence numbering and trail tracing) by providing a PACE header as described below. A packet is thus created having the data as the payload of the packet and a PACE header; the packet is sent to the packet-based network via a switch.
• example packet to TDM conversion logic 210 includes packet interface logic 212, data smoothing logic 214, and TDM logic 216.
• the packet interface logic has packet data extraction logic 218 that extracts the payload data from the packet and sends it to a FIFO buffer 220.
• Packet interface logic 212 also handles peer signaling and monitoring via the received PACE header, which includes information for round trip delay processing, sequence numbers, trail trace messages, status monitoring, and redundant data recovery.
• round trip delay measurements are made by using RTsnd and RTrcv bits in a PACE header.
• a trail trace message includes a 16 byte string and is assembled by packet interface logic 212 from the received packets. Packet interface logic 212 also inserts a message into a trail trace message field.
• Data smoothing logic 214 includes two stages. The first stage helps to remove jitter caused by the packet network. The second stage helps to remove jitter inherent in the implementation logic and data path and also tracks the wander associated to the source. The second stage provides standards compliant timing at its output.
• the first stage of the data smoothing logic includes clock generation and FIFO centering logic 222.
• the TI source clock rate is recovered from the rate of the data entering the FIFO.
• the FIFO fill level i.e., the difference between the producer "write” and consumer “read” pointer points of the FIFO
• FIFO control logic 224 are used to track the clock rate.
• the FIFO fill level remains stable if the output clock rate from FIFO 220 is the same as the incoming data rate at the input of FIFO 220.
• the output clock is adjusted to maintain the FIFO fill level at the FIFO center.
• the output clock rate is increased until the FIFO fill level stabilizes at the FIFO center. If the FIFO level falls below the FIFO center, the output clock rate is decreased until the FIFO fill level stabilizes at the FIFO center. Accordingly, especially over the long term, the output clock rate tracks the clock rate of the source.
• the FIFO fill level may be filtered before it is used to control the output clock, to help prevent sudden changes in the data rate from affecting the output clock.
• a low pass digital filter 310 used for this purpose filters differences in write and read pointers detected by pointer difference logic 312. The filtered FIFO level tracks the clock rate of the data source and its associated wander.
• the filtered FIFO fill level may be used as follows, in an example of filtering.
• One way to filter the fill level is to first calculate the derivative of the fill level, which indicates the change of the fill level over a time period. Jitter caused by the packet based network and packetization generally results in relatively large changes to the derivative of the fill level. Differences between the data rate and the output clock generally results in relatively small changes to the derivative of the fill level.
• a low pass filter attenuates the large changes in the derivative of the fill level, which produces a value that largely or mostly reflects the difference between data rate and output clock.
• Some of the packet related jitter has low frequency components that are not filtered out by the low pass filter and cause some wander in the filtered fill level. If such frequencies are under 10 hertz, the wander is acceptable at the TDM interface.
• the output clock when the filtered fill level is positive, the output clock needs to be increased; if the filtered fill level is negative, the output clock needs to be decreased; if the filtered fill level is zero, the output clock does not need to be adjusted.
• the amount of change that is needed on the output clock depends on the filtered fill level value and the sampling rate.
• the output of the FIFO is clocked with a gapped clock (i.e., a clock stream at a nominal frequency which may be missing clock pulses at arbitrary intervals for arbitrary lengths of time).
• the nominal frequency of the gapped clock is the recovered clock rate.
• the rate of change in the FIFO output clock is limited to help avoid inducing jitter into the second stage.
• FIFO centering heuristics are a set of methods used to determine the FIFO center level at various phases, including an acquire phase and a monitor phase.
• the acquire phase establishes a valid incoming signal, determines the FIFO center, and stabilizes the output data rate.
• the incoming packet stream is verified by monitoring sequence numbers in the PACE header.
• a packet stream is valid when there is no data loss.
• PACE calculates the FIFO center based on the measured round trip delays and minimum and maximum ("min-max") jitter levels.
• the round trip delay is measured by using the round trip send and round trip receive bits RTsnd, RTrcv in the PACE header. (A timer that may be associated with the packetizing logic is started at the time the RTsnd bit is set, and is sent to the packet network by the packetizing logic.
• the min-max jitter level is determined by recording the minimum and maximum FIFO level during a measurement period.
• the FIFO center is calculated presuming that the measured parameters represent a best case and that the FIFO center should be large enough to absorb worst case jitter.
• the FIFO center With respect to calculating the FIFO center, in at least one embodiment, multiple round trip delay measurements are made and the highest and lowest values are recorded. The highest value is called the Max Round Trip Delay and the lowest value is called the Min Round Trip Delay. The Max Round Trip Delay is used to calculate initial FIFO center. The initial
• FIFO center is equal to the Max Round Trip Delay divided by 2 and divided by 648 nanoseconds per bit. This is the parameter that is used when emulating TI data streams. For example, if the Max Round Trip Delay is 10 milliseconds, the initial FIFO center is equal to 7716 bits.
• the FIFO is filled with data to the FIFO center and data starts being clocked out of the FIFO.
• the lowest absolute fill level is monitored. If the lowest absolute fill level is less than the FIFO center divided by 2, the FIFO center is increased so that the difference between the FIFO center and the lowest absolute fill level is less than or equal to the original FIFO center divided by 2. For example, if the initial FIFO center is 7716 bits and the lowest absolute fill level detected is 3800 bits, the FIFO center should be increased by 116 bits (i.e. ((7716/2) -3800) * 2) to 7832 bits.
• the lowest absolute fill level is monitored again and further adjustments can be made. These steps are repeated until adjustments are not needed or for 4 times the Max Round Trip Delay (the larger the potential jitter, the longer it is desirable to monitor its jitter).
• This process establishes a FIFO center that is suitably not too large and not too small for the packet based network that is providing the transport packets.
• the FIFO center need not be manually configured, and is automatically determined at the beginning of operation and can be monitored while operational. Changes in the packet based network are automatically adjusted by either recalculating the FIFO center or slowly adjusting the FIFO center.
• the PACE traffic is monitored to determine changes in the network and to adjust to them.
• the FIFO center may be adjusted due to changes or variations in the packet-based network.
• the minimum and maximum FIFO levels and round trip delays are used to adjust the FIFO center to reduce the end to end latency and also to provide sufficient buffering to absorb changes in the network.
• PACE e.g., at packet interface 212
• PACE also detects failures to determine the need to acquire the packet stream again.
• the failures include sequence number error, loss of lock (i.e., incoming clock rate being out of range), FIFO overrun, FIFO underrun, and connectivity error (i.e., an incorrect trail trace).
• a second stage of the data smoothing logic includes a de-jitter circuit 228 for attenuation.
• the data smoothing logic can operate with very fine granularity in its clock control.
• De-jitter circuit 228 removes clock jitter (e.g., as required by Bellcore and ANSI) and tracks the wander in the clock.
• TDM logic 216 Based on the data and clock results of the data smoothing logic, TDM logic 216 outputs TDM data to an LIU.
• TDM logic provides legacy TDM interfaces and protocols including TI framing and monitoring, TI electrical interfaces, and TI timing.
• the type field may contain 0x8847.
• the SerDes header designates a port number and priority.
• the PACE header contains PACE control and status bits.
• the redundant data includes data from a previous packet.
• the data portion includes new data.
• a specific implementation of the PACE header is shown below.
• the sequence number portion is used to maintain the sequential nature of the data.
• the status portion indicates far end status.
• the far end status pertains to whether or not error conditions (such as incorrect sequence number, underflow, or overflow) have been detected at the far end. When the status portion is 0, no error conditions have been detected. When the status portion is 1, one or more error conditions have been detected.
• the trail trace message includes an identifier or source used to verify end to end connectivity.
• the RTsnd portion includes the Round Trip send bit, and the RTrec portion includes the Round Trip receive bit.
• line and diagnostic loopback capability is supported.
• the loopbacks are available through the line and system side (for packetizing and de-packetizing blocks).
• WAN type diagnostic features are supported.
• PACE supports two timing modes: external and internal. Internal timing may be provided by a clock source generated in the OSAP (such as local clock 226) or a clock source that is stratum traceable.
• a clock source generated in the OSAP such as local clock 226) or a clock source that is stratum traceable.
• the external timing mode is used when timing is originated at the TI termination points external to the OSAP.
• FIG. 4 illustrates an example of this timing configuration.
• FIG. 4 shows the timing being originated at both ends.
• PACE attempts to recover the clock at both ends.
• the timing is originated at one end and the other end is loop timed.
• PACE attempts to recover the clock at both ends.
• FIG. 6 shows an internal timing configuration where the TI equipment at both ends is loop timed. In at least some cases, if both OS APs are internally timed, it is useful or preferable if their clocks are stratum traceable.
• the systems and methods described herein may be implemented in hardware or software, or a combination of both.
• the technique is implemented in computer programs executing on one or more programmable computers, such as an embedded system or other computer running or able to run VxWorks (or Microsoft Windows 95, 98, 2000, Millennium Edition, NT; Unix; Linux; or MacOS); that each include a processor such as a Motorola PowerPC 8260 (or an Intel Pentium 4) possibly together with one or more FPGAs (field programmable gate arrays, e.g., by Xilinx, Inc.), a storage medium readable by the processor (including volatile and non-volatile memory and/or storage elements), possibly with at least one input device (e.g., a keyboard), and at least one output device.
• Program code is applied to data entered (e.g., using the input device) to perform the method described above and to generate output information.
• the output information is applied to one or more output devices (e.g., a display screen of the
• each program is implemented in a high level procedural or object-oriented programming language such as C++, Java, or Perl to communicate with a computer system.
• the programs can be implemented in assembly or machine language, if desired. In any case, the language may be a compiled or interpreted language.
• each such computer program is stored on a storage medium or device, such as ROM or magnetic diskette, that is readable by a general or special purpose programmable computer for configuring and operating the computer when the storage medium or device is read by the computer to perform the procedures described in this document.
• the system may also be considered to be implemented as a computer-readable storage medium, configured with a computer program, where the storage medium so configured causes a computer to operate in a specific and predefined manner.
• the output clock may itself be filtered to help prevent sudden changes from adversely affecting the output clock.
• a round trip delay measurement may be enhanced by adding adjustment values in the PACE header, e.g., in additional fields.
• Adjustments may be made to the initial FIFO center calculation to tailor or optimize the calculation for specific networks. If it is known that the packet based network has certain characteristics (e.g., high latency but low jitter), the knowledge may be used to calculate a highly useful FIFO center.

Landscapes

• Engineering & Computer Science (AREA)
• Computer Networks & Wireless Communication (AREA)
• Signal Processing (AREA)
• Small-Scale Networks (AREA)
• Data Exchanges In Wide-Area Networks (AREA)

Applications Claiming Priority (5)

Publications (1)

Family

ID=26965319

Family Applications (2)

Family Applications Before (1)

Country Status (2)

Families Citing this family (1)

Family Cites Families (8)

• 2002
  • 2002-05-02 WO PCT/US2002/013995 patent/WO2002091681A2/en not_active Application Discontinuation
  • 2002-05-02 EP EP02734162A patent/EP1386452A2/de not_active Withdrawn
  • 2002-05-02 EP EP02734114A patent/EP1384341A2/de not_active Withdrawn

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events