JP4490182B2 - Communication apparatus and packet communication method - Google Patents

Description

  INDUSTRIAL APPLICABILITY The present invention is suitable for application to a parent device of a system in which a parent device and a child device are synchronized by time information and execute error correction code processing, such as a PON system such as E-PON and GE-PON. It relates to the device.

  The Ethernet (registered trademark) PON system (EPON) stipulated in IEEE Draft P802.3ah connects a master device (OLT: Optical Line Terminal) and a plurality of slave devices (ONU: Optional Network Unit) with an optical transmission medium. By transmitting and receiving data, the PON system accommodates the Ethernet (registered trademark) service. The OLT is installed in the communication network operator station, and the ONU is installed in the subscriber's home or outdoors. The ONU accommodates one to a plurality of subscriber terminals. GE-PON is an EPON system having a gigabit transmission rate.

  In such a PON system, upstream signals from a plurality of ONUs are time-division multiplexed and transmitted to the OLT. To perform such multiplexing, it is necessary to synchronize the time between the OLT and each ONU. There is. For this reason, according to the IEEE Draft P802.3ah standard, the OLT transmits the current value of its own time information as time stamp information to the ONU, and the ONU updates its own time information according to the received time stamp value. It is shown.

  In these EPON and GE-PON, an error correction code (ECC) is added to a packet transmitted on a passive optical branch network (PON) shared by a slave device to improve transmission characteristics digitally. Application of form is under consideration.

  According to the IEEE Draft P802.3ah specification, the FEC (Forward Error Correction) function for executing error correction code processing is a physical layer (PCS) that is a lower layer of a data link layer that is a second layer. (Including Physical Coding Sublayer), PMA (Physical Medium Attachment), and PMD (Physical Medium Dependent)). The PCS sublayer has a function of encoding a received bit string from the data link layer and passing it to the PMA sublayer.

  The data link layer of the second layer has an LLC (Logic Link Control) sublayer, a Multi-Point MAC Control sublayer, and the like, and the parent device establishes communication with a plurality of child devices in the Multi-Point MAC Control sublayer. Time division multiple access control is performed.

  FIG. 15 shows a configuration of a parent device and a child device configured according to the IEEE Draft P802.3ah standard. The PON control unit in the parent device is a part that is included in the Multi-Point MAC Control sublayer and performs control for establishing PON communication. The parent device includes a PON control unit 100 having a multiplexing unit 110, a time stamp insertion unit (TS insertion unit) 120, and an FEC encoding unit 130. Here, only the functional configuration in the downlink direction is shown. .

  According to the IEEE Draft P802.3ah standard, as described above, in order to establish time division multiple access, time information is synchronized between the parent device and the child device, and the parent device manages the upstream data transmission time of each child device. To do. For this reason, the TS insertion unit 120 of the parent device accommodates time information as a time stamp value in a control packet transmitted from the parent device to each child device, and transmits the time information in the downlink direction. On the other hand, the child device synchronizes the time by setting its own timer value according to the time information transferred from the parent device, and when the time information is next input, the value is a constant value with its own timer value. If they are different from each other, control is performed to reset the link establishment state of the PON, and the communication is temporarily interrupted. Thus, in the PON system, time synchronization is positioned as a very important function.

  In the multiplexing unit 110 of the PON control unit 100, the control packet is multiplexed with the data packet input from the upstream side of the parent device, but it is necessary to include time information in the control packet as described above. After the multiplexing process in the unit 110, the time information is input to the control packet as a time stamp value. This is because the waiting time with the data packet occurs indefinitely during multiplexing, so if the time information is input into the control packet before multiplexing processing, the time until the packet actually arrives at the child device varies This is because accurate time synchronization cannot be achieved between the parent device and the child device. When synchronization is lost, when the upstream data from each child device arrives at the parent device, it is input out of timing controlled by the parent device, and the data of each child device collides normally. Communication is not possible. For this reason, the TS insertion unit 120 for inputting time information is arranged at the subsequent stage of the multiplexing unit 110.

  According to the IEEE Draft P802.3ah specification, in addition to the above control packet, an OAM (Operation Administration and Maintenance) packet containing maintenance operation information is also generated by the PON control unit 100 and multiplexed on downlink data. Yes. Unlike the control packet, the OAM packet is not used for PON access control, and therefore the time stamp value is not inserted.

  Here, in the IEEE Draft P802.3ah specification, as described above, the processing performed by the PON control unit 100 belongs to the second data link layer, and the error correction function is assigned to the first physical layer (PCS sublayer). Therefore, according to this rule, the FEC encoding unit 130 is arranged at the subsequent stage of the PON control unit 100. To be precise, the transmission path encoding unit exists in the preceding stage, and the blocks for performing the PMA sublayer and PMD sublayer processing exist in the subsequent stage, but are omitted here.

  The FEC encoding unit 130 performs an error correction code calculation process on the data packet and control packet transmitted from the PON control unit 100, adds an error correction code region to each packet, and outputs the packet. In the above IEEE Draft P802.3ah specification, the FEC encoding unit 130 uses Reed-Solomon code RS (255, 239), and a 16-byte code for every 239 bytes in the packet input to the FEC encoding unit 130 An area is added.

  When error correction code processing is performed in this way, data for the code area increases, and data loss occurs unless the actual data throughput is reduced. In the IEEE Draft P802.3ah specification, there is no description regarding this throughput control.

  In Patent Document 1, a plurality of input buffer units stop sending cells to the output buffer unit based on a back pressure signal output from the output buffer unit when the output buffer unit is congested. An invention relating to an ATM switch in which transmission of cells exceeding a predetermined transmission rate is stopped is disclosed.

  In the above-mentioned E-PON (GE-PON) system, if the invention of Patent Document 1 is applied, it becomes possible for the parent device to send a downstream packet without data loss. That is, a back pressure signal is generated based on the congestion state in the FEC encoding unit 130, and a flow control unit that controls the flow rate of the data packet is provided in the preceding stage of the PON control unit 100. The flow control unit is based on the back pressure signal. This is a method for controlling the output rate of data packets. When the output rate is limited in the flow control unit, if the input rate is high, a congestion state occurs in the flow control unit in the same manner as described above. For the flow control function for solving this problem, IEEE802.3x There is a method established in the standard, but detailed description thereof is omitted here.

IEEE Draft P802.3ah JP 2002-57669 A

  When the FEC function is applied to downlink transmission in E-PON (GE-PON), the IEEE Draft P802.3ah rule does not disclose a mechanism for adjusting the data amount increase for the FEC code area. When the function of the flow control unit is added by applying this technique, a configuration for transmitting a downlink packet without data loss can be obtained.

  However, the configuration based on the combination of the prior arts has a problem that the synchronization of time information between the parent device and the child device cannot be established, which is indispensable for the E-PON (GE-PON) system defined in IEEE Draft P802.3ah. . Hereinafter, this will be described with reference to FIGS. 16-1, 16-2 and FIG.

  FIG. 16A illustrates a case where a multiplexing unit is sending a data packet when the control packet arrives at the multiplexing unit in the multiplexing process of the data packet and the control packet in the PON control unit of the parent device (the multiplexing waiting is performed). FIG. 16-2 shows a case where the control unit arrives at the multiplexing unit and is in an idle state where no processing is performed by the multiplexing unit (when there is no multiplex waiting). Yes. As a process of the multiplexing unit, it is impossible to stop sending in the middle of a packet that has been sent once because erroneous data transfer is performed. Accordingly, as shown in FIG. 16A, when the control packet arrives at the multiplexing unit while the multiplexing process is being performed on the data packet by the multiplexing unit, the data packet is transmitted as the control packet. Is output after this data packet. On the other hand, when the control packet arrives at the multiplexing unit and the multiplexing unit is in an idle state, it is multiplexed and output as it is.

  The upper three packets in FIG. 17 are time charts showing packet state changes after (a) multiplexing, (b) TS insertion, and (c) FEC encoding when there is multiplexing waiting. The lower three packets are time charts showing packet state changes after (d) multiplexing, (e) TS insertion, and (f) FEC encoding when there is no multiplex waiting. Reference numerals (a) to (f) in FIG. 17 correspond to reference numerals (a) to (f) in FIGS. 16A and 16B.

  The time Tda1 from when the control frame is output from the multiplexing unit to when it is output from the TS insertion unit is the same regardless of whether there is multiple waiting or when there is no multiple waiting. There is no problem until you arrive at the club. However, when passing through the FEC encoder, a difference occurs between these two conditions.

  That is, assuming that the delay time from when the control frame is output from the TS insertion unit to when the FEC encoding unit is output is Tdb1 when there is no waiting, and Tdb0 when there is waiting, Tdb0> Tdb1 It becomes. This is because, when waiting occurs and a control packet is attached after the data packet, the control packet is delayed by an extra Tf1 by the code region by the addition of the code region in the FEC encoder, and passes through the FEC encoder. Because it becomes. That is, if the length of the code area of the data packet preceding the control packet is Tf1, Tdb0 = Tdb1 + Tf1. Since Tf1> 0, Tdb0> Tdb1.

  As described above, according to the conventional technique, when a redundant code such as FEC is added to a packet stream obtained by multiplexing a control packet that accommodates time information and a data packet, the control frame is transmitted due to the addition of the code area. Since the transfer delay varies, there is a problem in that synchronization of time information between the parent device and the child device cannot be established, and proper E-PON (GE-PON) communication cannot be performed between the parent device and the child device.

  The present invention has been made in view of the above. When a redundant code such as FEC is added to a packet stream obtained by multiplexing a control packet and a data packet containing time information, an addition of a code area is performed. Therefore, it is possible to reliably prevent the control frame transfer delay from fluctuating and to ensure that the time information is synchronized between the parent device and the child device so that the control frame is always output with a constant transfer delay. It is an object of the present invention to provide a communication device and a packet communication method capable of establishing proper PON communication between a parent device and a child device.

  In order to solve the above-described problems and achieve the object, the present invention provides a multiplexing unit that time-division-multiplexes and outputs a control packet to a data packet, and a time at which time information is inserted into the control packet output from the multiplexing unit. A communication apparatus comprising: an information insertion unit; and a redundant encoding unit that outputs a data packet output from the multiplexing unit and a control packet in which time information is inserted, and outputs a redundant code. When outputting a packet, the output unit includes a speed adjustment unit that performs a speed adjustment to add a free time corresponding to the redundant code area added by the encoding unit to the packet after the packet, and the redundant encoding unit includes: A redundant code is inserted into a free time formed after each packet.

  In the present invention, when outputting a data packet or a control packet from the multiplexing unit, an empty time corresponding to a redundant code area added thereafter to each data packet and control packet is added after each packet, and then redundant encoding is performed. At the time of processing, a redundant code is inserted in the free time formed after each packet.

  According to the present invention, the rate adjustment unit controls the packet output rate from the multiplexing unit in consideration of the redundant code region by providing an empty region for the redundant code region after each packet. Control packet processing delay does not occur during redundant encoding processing even when control packet multiplex waiting occurs in this part, so that control packet transfer delay regardless of whether there is packet waiting in the multiplex part Time fluctuation can be suppressed. Therefore, the control frame is always output with a certain transfer delay, and the time information can be reliably synchronized between the parent device and the child device, and communication can be performed between the parent device and the child device. Appropriate and stable communication without interruption can be achieved.

  Embodiments of a communication apparatus according to the present invention will be described below in detail with reference to the drawings. Note that the present invention is not limited to the embodiments.

Embodiment 1 FIG.
Embodiment 1 of the present invention will be described with reference to FIGS. In the communication system to which the invention of the first embodiment is applied, a transmission device having a time information transfer function, and a reception device having a function of updating the time of its own timer based on time information sent from the transmission device; It has. FIG. 1 shows the configuration of the transmission device, and FIG. 2 shows the configuration of the reception device. In the first embodiment, the speed adjustment unit 4 is connected to the multiplexing unit 1 of the transmission apparatus, and after the packets (data packets and control packets) output from the multiplexing unit 1 by the speed adjustment unit 4, the redundancy is thereafter performed. The code adding unit 3 provides vacant time (gap) for the redundant code area added to the packet, and then performs timing control to output the next data packet or control packet.

  As illustrated in FIG. 1, the transmission device includes a multiplexing unit 1, a TS (time stamp) insertion unit 2 as a time information insertion unit, a redundant code addition unit 3, and a rate adjustment unit 4. The multiplexing unit 1 receives a fixed-length or variable-length data packet from the outside and an internally generated control packet. The multiplexing unit 1 time-division-multiplexes the control packet into data packets. Output. The TS insertion unit 2 inserts the current time information into the control packet from the multiplexing unit 1 and outputs it. The redundant code adding unit 3 adds a redundant code to the data packet output from the multiplexing unit 1 and the control packet in which the time information is inserted, and outputs it.

  The speed adjustment unit 4 performs control for reading each multiplexed packet (each data packet and each control packet) from the multiplexing unit 1. When each packet is read from the multiplexing unit 1, after each packet, A time vacancy (gap) corresponding to at least the redundant code length added by the redundant code adding unit 3 is provided for the packet, and read control is performed to output the next data packet or control packet. That is, when each packet arrives at the multiplexing unit 1 and the multiplexing unit is sending another packet (when there is a multiplex waiting), only the amount corresponding to the redundant code length is behind the packet being sent. The read control is performed so as to output the next packet. In addition, when waiting for a redundant code does not occur even if a gap corresponding to a redundant code is included after each packet, the packet is multiplexed and output as it is. Incidentally, in FIG. 15 describing the prior art, no gap is provided between the packets output from the multiplexing unit 110, but in FIG. Is provided with a gap.

  When the packet is fixed length and the redundant code length added by the redundant code adding unit 3 is constant like an ATM cell used in A-PON, it corresponds to a predetermined redundant code length. Try to provide enough time to do what you want. On the other hand, when the data packet has a variable length and the redundant code length added by the redundant code adding unit 3 is variable according to the data packet length as in the MAC frame used for E-PON, the rate adjusting unit 4 As shown in FIG. That is, the packet length monitor 8 inspects the packet length of each packet output from the multiplexing unit 1, and notifies the control unit 9 of the inspection result. The read control unit 9 controls the output timing of the next packet based on the inspection result from the packet length monitor 8, and determines the length of the required redundant code area according to the length of each packet. Then, after a required time gap is made based on the determination result, the read timing of the control packet or data packet output following the packet is controlled.

  The packet output from the multiplexing unit 1 is input to the TS insertion unit 2, where a time stamp is inserted only in the control packet, and all packets are output to the redundant code addition unit. The redundant code adding unit 3 adds a redundant code after each input packet (each data packet and each control packet). Each packet to which a redundant code is added is output to the transmission path.

  4 are time charts showing packet state changes after (a) multiplexing, (b) TS insertion, and (c) redundancy code addition in the case of multiplexing waiting. The lower three packets are time charts showing packet state changes after (a) ′ multiplexing, (b) ′ TS insertion, and (c) ′ redundancy code addition when there is no multiplex waiting. Reference numerals (a) to (c) in FIG. 4 correspond to reference numerals (a) to (c) in FIG.

  When the control packet arrives at the multiplexing unit 1 and is in an idle state where the multiplexing unit 1 is not performing any processing (when there is no multiplex waiting), the control packet is multiplexed and output as it is. The time from when the control frame is output from the multiplexing unit 1 until it is output from the TS insertion unit 2 is Tda1, and the delay from when the control frame is output from the TS insertion unit 2 until the redundant code addition unit 3 is output Let time be Tdb1.

  On the other hand, when a data packet is being transmitted when the control packet arrives at the multiplexing unit 1 (when there is a multiplex waiting), this control packet is waited until the data packet being transmitted is transmitted, and then this data It will be output after the packet. After the data packet output from the multiplexing unit 1, a gap having a length corresponding to the redundant code area is added, and then the control packet is output from the multiplexing unit 1. Also in this case, the processing delay time from when the control frame is output from the multiplexing unit 1 to when the TS insertion unit 2 is output is Tda1 as is the case with the processing delay time when there is no multiplexing waiting.

  In addition, since a gap for the code area is secured in advance after the data packet, the redundant code adding unit 3 does not need to newly add a redundant code area as in the prior art, and is secured in advance. Since only the redundant code created for the gap is inserted, the delay time does not increase. Accordingly, in this case as well, the delay time from when the control frame is output from the TS insertion unit 2 to when the redundant code addition unit 3 is output is Tdb1 as is the processing delay time when there is no multiplex waiting. As described above, the delay time from the TS insertion of the control packet to the output of the redundant code adding unit 3 is constant regardless of whether there is waiting in the multiplexing unit 1.

  On the other hand, as shown in FIG. 2, the receiving apparatus includes a redundant code removal unit 5, a separation unit 6, and a TS extraction unit 7. The receiving device receives the packet from the transmitting device via the transmission path and inputs it to the redundant code removing unit 5. The redundant code removing unit 5 removes the redundant code from each input packet (each data packet and each control packet), and sends the packet from which the redundant code has been removed to the separating unit 6. The separation unit 6 separates the control packet and the data packet and sends the control packet to the TS extraction unit 7. The separated data packet is sent to a packet processing unit (not shown) in the subsequent stage. The TS extraction unit 7 extracts a time stamp that is time information inserted by the transmission device from the input control packet.

  The TS extraction unit 7 compares the extracted time information with the current timer value of the timer that is operating based on the time information extracted from the previously input control packet, and the difference is larger than a predetermined threshold value. In this case, the time information extracted this time is invalidated, the communication link with the transmission device is disconnected, and then the process of reestablishing the communication link is started. If the difference is smaller than a predetermined threshold as a result of the comparison, the timer value of the timer is updated with the time information extracted this time.

  In the receiving apparatus, the redundant code removing unit 5 performs processing for removing redundant codes, and the separating unit 6 also performs processing for dividing the packet into two paths. The packet processing delay time does not vary depending on the amount of packets, that is, the data rate. As described above, in the transmission apparatus, the delay time from the TS insertion of the control packet to the output of the redundant code adding unit 3 is constant regardless of whether or not the multiplexing unit 1 waits. Therefore, in this system, since there is no variation in the processing delay time from the TS insertion unit 2 of the transmission device to the TS extraction unit 7 of the reception device, the control packet is caused by the delay processing time in these units. The situation where the comparison result between the time information and the timer value in the TS extraction unit 7 exceeds the threshold value and the communication link is disconnected does not occur.

  As described above, according to the first embodiment, the speed adjustment unit 4 in the transmission apparatus provides a space for the redundant code area after each packet, so that the packet output rate from the multiplexing unit 1 is considered in the redundant code area. Therefore, even when the multiplexing waiting for the control packet occurs in the multiplexing unit 1, the processing delay of the control packet does not occur in the redundant code adding unit 3, so that the packet waiting in the multiplexing unit 1 can be performed. Regardless of whether or not there is, fluctuations in the transfer delay time of the control packet can be suppressed. Therefore, the control frame is always output with a constant transfer delay, the synchronization of time information can be reliably established between the transmission device and the reception device, and between the transmission device and the reception device, Appropriate and stable communication without interruption due to disconnection of the communication link can be achieved.

Embodiment 2. FIG.
Next, a second embodiment of the present invention will be described with reference to FIGS. In the second embodiment, the present invention is applied to a parent device of a PON system such as E-PON (GE-PON). The point that is applied to the PON system and the point that the flow control unit is provided are the major differences between the first embodiment and the second embodiment. In this case, for convenience, only one child device is shown, but a plurality of child devices may be connected to the parent device. In this case, only the downlink processing function of the parent device and the slave device is described, and the description of the uplink processing function is omitted.

  In FIG. 5, the parent device (OLT) includes a PON control unit 16 including a multiplexing unit 11, a TS insertion unit 12, and a speed adjustment unit 14, an FEC encoding unit 13, and a flow control unit 15. .

  The flow control unit 15 has a buffer for accumulating data packets. The flow control unit 15 monitors the input flow rate of the data packet, and when there is an input exceeding the allowable range, makes a data stop request to the upstream device. The flow control unit 15 stops sending data packets to the multiplexing unit 11 when a back pressure signal, which will be described later, is input from the speed adjustment unit 14, and accumulates in the buffer when the back pressure signal is not input. The read data packet is read out and output to the multiplexing unit 11 of the PON control unit 16 to execute control for reducing the output rate of the data packet. A data packet from the flow control unit 15 is input to the multiplexing unit 11, and a control packet including time information and an OAM packet that is one of the control packets are input. Packets, OAN packets, and data packets are time-division multiplexed and output. A control packet including time information and an OAM packet which is one of the control packets are internally generated in the parent device. Hereinafter, a control packet including time information is simply referred to as a control packet.

  The speed adjustment unit 14 performs control to read out each multiplexed packet (each data packet, each control packet, and each OAM packet) from the multiplexing unit 11, and when reading each packet from the multiplexing unit 11, After the packet, a time vacancy (gap) corresponding to the redundant code length added by the redundant code adding unit 13 is provided for the packet, and read control is performed so as to output the next packet. That is, when each packet arrives at the multiplexing unit 11 and the multiplexing unit 11 is sending another packet (when there is a multiplex waiting), the amount corresponding to the redundant code length is added after the packet being sent. Read control is performed so as to output the next packet with only a gap. On the other hand, if there is no waiting between the packets even if a gap corresponding to the redundant code is included after each packet, the packet is multiplexed and output as it is.

  In this case, as shown in FIG. 3, the speed adjustment unit 14 includes a packet length monitor 8 and a read control unit 9, and the packet length monitor 8 determines the packet length of each packet output from the multiplexing unit 1. And the reading control unit 9 is notified of the inspection result. The read control unit 9 controls the output timing of the next packet based on the inspection result from the packet length monitor 8, and each packet corresponds to the FEC code area required according to the length of each packet. The packet reading operation of the multiplexing unit 11 is controlled so that an empty area is provided after the data is read out. Therefore, a gap is provided between the packets output from the multiplexing unit 11 as shown in FIG. Further, in this case, the speed adjustment unit 14 detects whether or not a predetermined amount or more of data from the flow control unit 15 is accumulated in the buffer in the multiplexing unit 11, and the predetermined amount or more of data is accumulated. When the multiplexing unit 11 enters the congested state, a back pressure signal is output to the flow control unit 15.

  The TS insertion unit 12 inserts the current time information (time stamp) into the control packet from the multiplexing unit 11, performs transmission path coding on all packets, and then sends all packets to the FEC coding unit 13. Output. A FEC (Forward Error Correction) encoding unit 13 calculates an error correction code for each input packet, and adds the calculated error correction code as a code area after each packet. Each packet to which the FEC code is added is output to the transmission path.

  In FIG. 5, the child device (ONU) includes an FEC decoding unit 21, a separation unit 23, and a PON control unit 22 having a TS extraction unit 24. The child device receives the packet from the parent device via the transmission path and inputs the packet to the FEC decoding unit 21. The FEC decoding unit 21 decodes each input packet, performs error correction processing using the error correction code inserted in the code area, deletes the code area, and performs PON control on the packet from which the code area has been deleted Sent to the separation unit 23 of the unit 22. The separation unit 23 of the PON control unit 22 separates the control packet, the OAM packet, and the data packet, and sends the control packet to the TS extraction unit 24. The separated data packet is sent to a packet processing unit (not shown) in the subsequent stage, and the OAM packet is sent to an OAM processing unit (not shown) in the subsequent stage.

  The TS extraction unit 24 extracts a time stamp from the input control packet. Further, the TS extraction unit 24 compares the extracted time stamp with the current timer value of the timer operating based on the time stamp extracted from the previously input control packet, and the difference is determined from a predetermined threshold value. If it is larger, the time information extracted this time is invalidated, the communication link with the parent device is disconnected, and then the process of reestablishing the communication link is started. If the difference is smaller than a predetermined threshold as a result of the comparison, the timer value of the timer is updated with the time stamp value extracted this time, and the time stamp value extracted this time is used as the output timing of the upstream data. The standard.

  FIG. 6A illustrates a case where the multiplexing unit 11 is transmitting a data packet when the control packet arrives at the multiplexing unit in the multiplexing process of the data packet and the control packet in the multiplexing unit 11 of the parent device (multiplex waiting). FIG. 6-2 shows a case where the control packet arrives at the multiplexing unit and is in an idle state where no processing is performed by the multiplexing unit (when there is no multiplexing waiting). ) Shows a packet arrangement.

  As shown in FIG. 6A, when there is a multiplex waiting, the data packet interval is clogged in the preceding stage of the multiplex unit 11, but the multiplex unit is operated by the operation of the speed adjustment unit 14 connected to the multiplex unit 11. As for the packet output from 11, a gap period corresponding to each FEC code area is provided after all packets including the control packet, the OAM packet, and the data packet. On the other hand, as shown in FIG. 6B, if there is no waiting between packets even if a gap corresponding to the code area is included after each packet, the packet is multiplexed and output as it is.

  The upper three packets in FIG. 7 are time charts showing packet state changes after (a) multiplexing, (b) TS insertion, and (c) error correction code addition when there is multiplexing waiting. The lower three packets in FIG. 7 are time charts showing packet state changes after (a) ′ multiplexing, (b) ′ TS insertion, and (c) ′ error correction code addition when there is no multiplex waiting. is there. Reference numerals (a) to (c) and (a) ′ to (c) ′ in FIG. 7 correspond to reference numerals (a) to (c) and (a) ′ to (c) ′ in FIG. 6.

  When the control packet arrives at the multiplexing unit 11 and is in an idle state where the multiplexing unit 11 is not performing any processing (when there is no multiplex waiting), the control packet is multiplexed and output as it is. The time from when the control frame is output from the multiplexing unit 11 until it is output from the TS insertion unit 12 is Tda1, and the delay from when the control frame is output from the TS insertion unit 12 to when it is output from the FEC encoding unit 13 Let time be Tdb1.

  On the other hand, when a data packet is being transmitted when the control packet arrives at the multiplexing unit 11 (when there is a multiplex waiting), this control packet is waited until the data packet being transmitted is transmitted, and then this data It will be output after the packet. After the data packet output from the multiplexing unit 11, a gap having a length corresponding to the error correction code area is added, and then the control packet is output from the multiplexing unit 11. Also in this case, the processing delay time from when the control frame is output from the multiplexing unit 11 to when the TS insertion unit 12 is output is Tda1 as is the case with the processing delay time when there is no multiplexing waiting.

  In addition, since a gap corresponding to the error correction code area is secured in advance after the data packet, the FEC encoding unit 13 does not need to newly add a redundant code area as in the prior art, and is secured in advance. Since only the created error correction code is inserted into the gap, the delay time does not increase. Therefore, also in this case, the delay time from when the control frame is output from the TS insertion unit 12 to when the FEC encoding unit 13 is output is Tdb1 as is the case of the processing delay time when there is no multiplex waiting. As described above, the delay time from the TS insertion of the control packet to the output of the FEC encoding unit 13 is constant regardless of whether or not there is waiting in the multiplexing unit 11.

  As described above, according to the second embodiment, the speed adjustment unit 14 in the parent apparatus provides an empty area for the FEC code area after each packet, so that the packet output rate from the multiplexing unit 11 is considered in the FEC code area. Therefore, even if the control unit multiplex waits in the multiplexing unit 11, the control packet processing delay does not occur in the FEC encoding unit 13. Regardless of whether or not there is, fluctuations in the transfer delay time of the control packet can be suppressed. Therefore, the control frame is always output with a constant transfer delay, the synchronization of time information can be reliably established between the parent device and the child device, and between the parent device and the child device, Appropriate and stable communication without interruption due to disconnection of the communication link can be achieved.

  Further, in the second embodiment, since the speed adjustment unit 14 adds a gap region after each packet, the throughput of the data packet in the multiplexing unit 11 is reduced. 11 is in a congested state, a back pressure signal is output to the flow control unit 15 so that the flow control unit 15 performs an operation for reducing the data output rate. It can be surely prevented.

  In the second embodiment, when the multiplexing unit 11 includes a large-capacity data packet buffer, the flow control unit 15 may be omitted. In the above, the speed adjustment unit 14 outputs a back pressure signal in order to eliminate the congestion in the multiplexing unit 11 due to the throughput reduction due to the speed adjustment. However, as shown in FIG. You may make it provide the pose insertion part 17 to output. The pause insertion unit 17 is provided in the middle of the path of the upstream packet signal, and detects whether or not a predetermined amount or more of data from the flow control unit 15 is accumulated in the buffer in the multiplexing unit 11, as with the speed adjustment unit 14. When the multiplexing unit 11 is in a congested state where a predetermined amount or more of data is accumulated, the pause packet is time-division multiplexed into an upstream packet signal and transmitted to the flow control unit 15. When the flow control unit 15 receives the pause packet, it executes control to reduce the output rate of the data packet, for example, by stopping the transmission of the data packet to the multiplexing unit 11.

Embodiment 3 FIG.
Next, a third embodiment of the present invention will be described with reference to FIGS. In the third embodiment, speed adjustment for the FEC code area is performed by the flow control unit, and the flow control unit adds a dummy PAD area to the data packet.

  FIG. 9 shows the configuration of the parent apparatus according to the third embodiment. The parent apparatus includes a flow control unit 30, a PON control unit 33, and an FEC encoding unit 36. The flow control unit 30 includes a buffer 31 and a speed adjustment unit 32. The PON control unit 33 includes a multiplexing unit 34 and a TS insertion unit 35. The FEC encoding unit 36 includes a packet identification unit 37, a buffer 38, and an encoding unit 39.

  The buffer 31 of the flow control unit 30 accumulates the data packets input from the preceding apparatus, and outputs the accumulated data packets to the multiplexing unit 34 of the PON control unit 33 according to the control of the speed adjustment unit 32. The flow control unit 30 monitors the input flow rate of the data packet, and makes a data stop request to the upstream device when there is an input exceeding the allowable range. In addition, when a back pressure signal (to be described later) is input from the FEC encoding unit 36, the speed adjustment unit 32 stops sending the data packet to the multiplexing unit 11, and when the back pressure signal is not input, the buffer 31 The data packets stored in the PON control unit 33 are read out and output to the multiplexing unit 11 of the PON control unit 33, thereby executing control for reducing the output rate of the data packets.

  Here, when the speed adjustment unit 32 of the flow control unit 30 reads the data packet from the buffer 31, a dummy PAD as dummy data having a length corresponding to the FEC code area in the FEC encoding unit 36 is added to the end of the data packet. (Shown in black in the figure) is added, and the data packet is output as a new packet including the dummy PAD portion. For this reason, the speed adjustment unit 32 recalculates and adds a frame check sequence code (FCS) at the end of each data packet. Therefore, precisely, the dummy PAD part is added after the FCS code of each data packet.

  The multiplexing unit 34 receives the data packet with the dummy PAD added from the flow control unit 30 and the internally generated control packet and OAM packet. The multiplexing unit 34 adds the dummy PAD. The processed data packet is processed as a normal data packet, and the data packet, the control packet, and the OAM packet are time-division multiplexed and output. The TS insertion unit 35 inserts the current time information (time stamp) into the control packet from the multiplexing unit 11, performs transmission path coding on all packets, and then sends all packets to the FEC coding unit 36. Output.

  The packet identification unit 37 of the FEC encoding unit 36 performs a process of identifying the control packet and the OAM packet from the input packets as data packets. In this case, the packet identification unit 37 performs processing for identifying the control packet and the OAM packet from the data packet by inspecting the code of the type area (TYPE area) of the packet. The packet identification result is input to the encoding unit 39. The buffer 38 accumulates each packet input from the TS insertion unit 35 and outputs the accumulated data packet to the encoding unit 39 in accordance with a request from the encoding unit 39. The buffer 38 outputs a back pressure signal to the speed adjustment unit 32 of the flow control unit 30 when the packet accumulation amount increases to a certain value or more.

  The encoding unit 39 of the FEC encoding unit 36 newly adds an FEC code area for the control packet and the OAM packet based on the identification result of the packet identification unit 37, and the flow control unit 30 for the data packet. Error correction processing is performed by overwriting the FEC code area on the added dummy PAD portion. Each packet to which the FEC code is added is output to the transmission path. In the configuration of the third embodiment, when adding the FEC code area of the control packet and the OAM packet, if there is a subsequent packet, it is necessary to make the packet wait, so that the FEC encoding unit 36 receives data. An accumulation buffer 38 is provided. Then, when packets continuously arrive at the FEC encoder 36, congestion occurs, so that a back pressure signal is output to the flow controller.

  FIG. 10-1 illustrates a packet arrangement when the multiplexing unit 11 is transmitting a data packet when a control packet arrives at the multiplexing unit 34 (when there is a multiplex waiting), and FIG. When the control packet arrives at the multiplexing unit 34, the packet arrangement when the multiplexing unit 34 is in an idle state (when there is no multiplexing waiting) is shown.

  As shown in FIG. 10A, when there are multiple waits, the interval between the data packets is clogged in the preceding stage of the multiplexing unit 34, but the flow control unit 30 already has a dummy for the FEC code area for each packet. Since PAD is added, the newly added area in the output of the FEC encoding unit 36 is only the FEC code area of the control packet. On the other hand, as shown in FIG. 10-2, when no waiting occurs between the packets, similarly, a dummy PAD for the FEC code area is already added to each packet in the flow control unit 30, and The area newly added in the output of the FEC encoding unit 36 is only the FEC code area of the control packet.

  The upper four packets in FIG. 11 show the packet state change after multiplexing (a) before multiplexing, (b) after multiplexing, (c) after TS insertion, and (d) after addition of FEC code. 11 is a chart, and the lower three packets in FIG. 11 show packet state changes after (b) ′ multiplexing, (c) ′ TS insertion, and (d) ′ error correction code addition when there is no multiplex waiting. It is a time chart which shows. Reference numerals (a) to (d) and (b) ′ to (d) ′ in FIG. 7 correspond to reference numerals (a) to (d) and (b) ′ to (d) ′ in FIG.

  If the multiplexing unit 11 is in an idle state when the control packet arrives at the multiplexing unit 11 (when there is no multiplexing waiting), the control packet is multiplexed and output as it is. The time from when the control frame is output from the multiplexing unit 34 to when it is output from the TS insertion unit 35 is Tda1, and the delay time from when the control frame is output from the TS insertion unit 35 to when it is output from the encoding unit 39 Is Tdb1.

  On the other hand, when there is multiple waiting, this control packet is output after this data packet after waiting until the data packet being transmitted is transmitted. After the data packet output from the multiplexing unit 11, a dummy PAD having a length corresponding to the FEC code area is added by the flow control unit 30, and then the control packet is output from the multiplexing unit 11. Also in this case, the processing delay time from when the control frame is output from the multiplexing unit 34 to when the TS insertion unit 35 is output is Tda1 as is the case with the processing delay time when there is no multiplexing waiting.

  Further, since the dummy PAD area for the FEC code area is secured after the data packet, the encoding unit 39 does not need to newly add a redundant code area as in the prior art, and is added in advance. Since only the FEC code created for the dummy PAD area is overwritten, the delay time does not increase. Accordingly, in this case as well, the delay time from when the control frame is output from the TS insertion unit 35 to when the FEC encoding unit 36 is output is Tdb1 as is the case with the processing delay time when there is no multiplex waiting. As described above, the delay time from the insertion of the control packet TS to the output of the FEC encoding unit 36 is constant regardless of whether there is waiting in the multiplexing unit 34.

  As described above, according to the third embodiment, the flow control unit in the parent apparatus performs the packet output rate control considering the FEC code region by adding the dummy PAD region to the data packet. Even when control packet multiplex waiting occurs at 34, the control packet processing delay does not occur in the FEC encoding unit 36, so that the control packet transfer delay regardless of whether there is packet waiting at the multiplex unit 34. Time fluctuation can be suppressed. Therefore, the control frame is always output with a constant transfer delay, the synchronization of time information can be reliably established between the parent device and the child device, and between the parent device and the child device, Appropriate and stable communication without interruption due to disconnection of the communication link can be achieved.

  Further, in the third embodiment, since the flow control unit 30 adds dummy data after each packet, the throughput of the data packet of the subsequent apparatus is reduced. However, the FEC encoding unit 36 performs the FEC encoding. When the unit 36 is in a congested state, a back pressure signal is output to the flow control unit 30, and the flow control unit 30 performs an operation of reducing the data output rate. Congestion can be reliably prevented. Furthermore, in the third embodiment, since it is not necessary to add a speed adjustment function or a Pause insertion function to the PON control unit, it is possible to use a PON control IC defined in the current IEEE Draft P802.3ah. There is an effect that the development volume of the apparatus can be suppressed.

Embodiment 4 FIG.
Next, a fourth embodiment of the present invention will be described with reference to FIGS. FIG. 12 shows the configuration of the parent device according to the fourth embodiment. In the third embodiment, the packet identification unit 37 recognizes the control packet and the OAM packet based on the TYPE code of the packet. In the fourth embodiment, the packet identification unit 37 identifies the control packet and the OAM packet with respect to the data packet. An identifier adding unit 40 that adds a special identifier that can be given to the control packet and the OAM packet is added. The configuration and operation of the other parts are the same as those in the third embodiment, and a duplicate description is omitted.

  In FIG. 12, the identifier assigning unit 40 assigns a special identifier that can distinguish the control packet and the OAM packet from other packets among the packets output from the multiplexing unit 34 to the control packet and the OAM packet. The packet identification unit 37 of the FEC encoding unit 36 detects the identifier, and identifies the control packet and the OAM packet as a data packet.

  FIG. 13 shows an example of the accommodation location in the packet of the identifier assigned by the identifier assigning unit 40. As shown in FIG. 13, the identifier assigning unit 40 is 1 for a control packet or an OAM packet with respect to a specific bit in a reserve part (RES part) that is an unused part of a preamble area at the head of each packet. If the data packet, 0 is written. The packet identification unit 37 can identify the control packet and the OAM packet as the data packet by checking this bit to see if it is 1 or 0.

  The preamble portion has an area called LLID, which is a link identifier when performing PON control between the parent device and the child device. In PON, since downlink data is in a broadcast format, it is often encrypted to prevent information leakage, but the preamble is not encrypted for link identification. Accordingly, even when the PON control unit 33 performs encryption, the packet identification unit 37 can recognize the control packet and the OAM packet.

  As described above, according to the fourth embodiment, in addition to the effects of the third embodiment, the present invention can be applied to an encryption environment and has an effect of improving security.

Embodiment 5 FIG.
Next, Embodiment 5 of the present invention will be described with reference to FIG. 9 and FIG. In the fifth embodiment, a packet identification unit 50 having the configuration shown in FIG. 14 is used in place of the packet identification unit 37 in the third embodiment shown in FIG. The configuration and operation of the other parts are the same as those in the third embodiment, and a duplicate description is omitted. The packet identification unit 37 of the third embodiment recognizes the control packet and the OAM packet based on the TYPE code of the packet, but the packet identification unit 50 shown in FIG. 14 performs control by checking the packet length. Packets and OAM packets are identified as data packets.

  In order to identify with such a packet length, the following method is used. First, since the control packet is used with the shortest packet length of 72 bytes, that is, a predetermined fixed length, it can be easily identified. Even when the data packet is originally the shortest packet length, since the dummy PAD corresponding to the FEC code area is added in the flow control unit 30, it is the same length as the control packet when it is input to the packet identification unit 50. There is nothing. That is, if the length of the data packet not including the dummy pad is set to be equal to or longer than the control packet, the length of the data packet to which the dummy pad is added does not match the control packet. Therefore, if 72 bytes, which is the length of the shortest packet, is detected, the control packet can be identified as a data packet.

  Further, regarding the distinction between the OAM packet and the data packet, when the dummy PAD is added to the data packet in the flow control unit 30, the length of the data packet including the dummy PAD is not long enough to be used in the OAM packet. Limit. For example, when an OAM packet is used with a length of 136 bytes, a dummy PAD is added to the data packet so that it does not become 136 bytes. Similarly, when the OAM packet has a plurality of different lengths, a dummy PAD length that does not become the same length is selected.

  Thus, the control packet is set to a fixed length, and the dummy PAD length added to the data packet is controlled so that the total length of the data packet including the dummy PAD is different from the OAM packet length. That is, the length of the data packet to which the dummy pad is added is different from the control packet length and the OAM packet length. The packet length monitor 51 inspects the packet lengths of control packets, OAM packets, and data packets input to the packet identification unit 50. The identification unit 52 distinguishes the control packet and the OAM packet from the data packet based on the detection length.

  Thus, the fifth embodiment has the same effect as the third embodiment. Furthermore, since a specific packet pattern is not used to identify the packet type, it can be applied to an encryption environment, and security can be improved.

Embodiment 6 FIG.
In the previous third embodiment, the speed control unit 32 outputs each data packet by adding a dummy PAD as dummy data having a length corresponding to the FEC code area to the end of each data packet, and outputs an FEC encoding unit. Based on the identification result of the packet identification unit 37, the 36 encoding unit 39 newly adds an FEC code region for the control packet and the OAM packet, and the dummy packet added by the flow control unit 30 for the data packet. The FEC code area is overwritten on the PAD part.

  On the other hand, in the sixth embodiment, the speed control unit 32 sets the packet length (Pa), which is the longer of the control packet and the OAM packet, between the data packets to which the dummy PAD is added. Each data packet is transmitted with a free time corresponding to the length (Pa + Pb) obtained by adding the length (Pb) of the FEC code area added to the long packet length packet.

  Thereby, the throughput of the data packet is reduced, but when the FEC code is added by the encoding unit 39, the control packet and the OAM packet area and the FEC code area of the control packet and the OAM packet are preliminarily determined. It will be secured. Therefore, when the control unit and the OAM packet are encoded by the encoding unit 39, a buffer for waiting for the subsequent packet required in the third embodiment and a back pressure signal are output to the flow control unit when congestion occurs. Therefore, there is an effect that the control circuit for performing the operation is unnecessary, and the circuit scale and control are simplified.

  As described above, the communication apparatus according to the present invention is useful for use in a PON system.

It is a block diagram which shows the structure of the transmitter about Embodiment 1 of this invention. It is a block diagram which shows the structure of a receiver about Embodiment 1 of this invention. It is a block diagram which shows the internal structure of the speed adjustment part in the transmission apparatus of FIG. 4 is a time chart showing the state change of each part of a packet in Embodiment 1 when there is a multiple waiting and when there is no waiting. It is a block diagram which shows the structure of a parent | device and a child apparatus about Embodiment 2 of this invention. It is a figure which shows a packet arrangement | sequence when there exists multiple waiting in the parent apparatus of Embodiment 2. FIG. FIG. 10 is a diagram showing a packet arrangement when there is no multiple waiting in the parent device of the second embodiment. 6 is a time chart showing the state change of each part of a packet in Embodiment 2 when there is a multiple waiting and when there is no multiple waiting. FIG. 10 is a diagram illustrating a modification of the parent device according to the second embodiment. It is a block diagram which shows the structure of a parent apparatus about Embodiment 3 of this invention. FIG. 10 is a diagram showing a packet arrangement when there is multiple waiting in the parent apparatus of the third embodiment. FIG. 10 is a diagram showing a packet arrangement when there is no multiple waiting in the parent apparatus of the third embodiment. 14 is a time chart showing the state change of each part of a packet in Embodiment 3 when there is a multiple waiting and when there is no waiting. It is a block diagram which shows the structure of a parent apparatus about Embodiment 4 of this invention. It is a figure explaining the accommodation place in the packet for providing an identifier. It is a block diagram which shows the structure of the packet identification part in the parent apparatus used for Embodiment 5 of this invention. It is a figure which shows a prior art. It is a figure which shows a packet arrangement | sequence when there exists multiple waiting in a prior art. It is a figure which shows a packet arrangement | sequence when there is no multiple waiting in a prior art. It is a time chart which shows the state change of each part of a packet in the prior art when there is no multiplex waiting and when there is no queuing.

Explanation of symbols

1,11,34 Multiplexing unit 2,12,35 TS insertion unit 3,13 Redundant code adding unit 4,14,32 Rate adjusting unit 5 Redundant code removing unit 6,23 Separating unit 7,24 TS extracting unit 8 Packet length monitor DESCRIPTION OF SYMBOLS 9 Read control part 13,36 FEC encoding part 15,30 Flow control part 16,22,33 PON control part 17 Pause insertion part 21 FEC decoding part 31,38 Buffer 37 Packet identification part 39 Encoding part 40 Identifier provision part 50 Packet identification unit 51 Packet length monitor 52 Identification unit

Claims (11)

Multiplexer that time-division-multiplexes control packets into data packets and outputs, time information inserter that inserts time information into control packets output from the multiplexer, and data packet and time information output from the multiplexer In a communication device comprising a redundant encoding unit that outputs a redundant code added to the control packet,
A speed adjusting unit for adjusting a speed for adding a free time corresponding to a redundant code area added by the redundant encoding unit to the packet after the data packet or the control packet is output from the multiplexing unit; Prepared,
The redundant encoding unit inserts a redundant code into a free time formed after each packet. The speed adjustment unit has means for outputting a detection signal when detecting a congestion state in the multiplexing unit,
The communication apparatus according to claim 1, further comprising a flow control unit that executes control for stopping transmission of a data packet to the multiplexing unit when the detection signal is received, at a preceding stage of the multiplexing unit. Multiplexer that time-division-multiplexes control packets into data packets and outputs, time information inserter that inserts time information into control packets output from the multiplexer, and data packet and time information output from the multiplexer In a communication device comprising a redundant encoding unit that outputs a redundant code added to the control packet,
A speed adjustment unit is provided in the preceding stage of the multiplexing unit for adjusting the rate of adding dummy data for the redundant code area added by the redundant encoding unit after each input data packet and outputting it to the multiplexing unit. ,
The redundant encoding unit includes:
Packet identification means for identifying the control packet and the data packet;
An encoding unit that overwrites a redundant code in a dummy data area added after the data packet and adds a redundant code after the control packet in which the time information is inserted based on the identification result of the packet identification unit When,
A communication apparatus comprising: The redundant encoding unit has means for outputting a detection signal when detecting a congestion state in the own apparatus,
4. The communication apparatus according to claim 3 , wherein the speed adjustment unit has a flow control function for executing control for stopping transmission of a data packet to the multiplexing unit when the detection signal is received. The speed adjusting unit outputs a data packet to the multiplexing unit so that at least a free time corresponding to a redundant code area added to the control packet and the control packet is provided between the data packets. The communication device according to claim 3 or 4 . The communication according to any one of claims 3 to 5 , wherein the speed adjustment unit calculates and adds a frame check sequence code to the data packet to which the dummy data is added. apparatus. An identifier adding unit for adding identification information for identifying the control packet to the control packet output from the multiplexing unit;
The communication apparatus according to claim 3 , wherein the packet identification unit identifies the control packet and the data packet by detecting the identification information. The communication apparatus according to claim 3 , wherein the packet identification unit inspects the packet length of each packet and identifies a control packet and a data packet based on the result. As a control packet, a first control packet into which time information is inserted and a second control packet containing maintenance operation information are input to the multiplexing unit, and the length of the data packet not including dummy data is set to the first control packet. With the above,
The communication apparatus according to claim 8 , wherein the speed adjustment unit determines the dummy data length so that the length of the data packet to which the dummy data is added is different from the length of the second control packet. When a data packet or a control packet is time-division multiplexed and output, an empty time corresponding to the redundant code area added to the packet is added after the packet, and the control packet is time-division multiplexed to the data packet. A first step of outputting
A second step of inserting time information into the time-division multiplexed control packet;
A third step of adding a redundant code to the time-division multiplexed data packet and a control packet into which the time information is inserted by inserting a redundant code in the idle time formed after each packet,
A packet communication method comprising: A first step of performing speed adjustment control for adding and outputting dummy data corresponding to a redundant code area to be added thereafter after each input data packet;
A second step of time division multiplexing and outputting a control packet to the data packet to which the dummy data is added;
A third step of inserting time information into the time-division multiplexed control packet;
A control packet and the data packet are identified, and based on the identification result, a redundant code is overwritten in a dummy data area added after the data packet, and a redundancy is added after the control packet in which the time information is inserted. A fourth step of adding and outputting a code;
A packet communication method comprising:

Images