JP5404972B2 - Optical communication system, communication apparatus, and bandwidth control method - Google Patents

Description

  The present invention relates to an optical communication system configured by one-to-many connection of a station side device and a subscriber side device (subscriber termination device).

  A PON (Passive Optical Network), which is one form of an optical communication system, includes a station side device (OLT: Optical Line Terminal) disposed at a central office and a subscriber termination device (ONU: Optical) disposed at a plurality of subscriber homes. Network Unit) is connected one-to-many with star couplers and optical fibers, and time division multiplexing communication is performed in the upstream direction (direction from the subscriber to the central office), and continuous communication is performed in the downstream direction. This PON is widely used as an economical communication system in which a plurality of ONUs are accommodated by one OLT by having a function of filtering a downstream signal with an ONU identifier given to a frame on the ONU side (for example, Non-patent document 1). In recent years, standardization of 10 Gbps class PON for the purpose of further increasing the communication capacity of the access network has been completed (for example, see Non-Patent Document 2). This standard is a 10 Gbps version of the 1 Gbps class PON defined in the conventional IEEE 802.3, and is accommodated in the OLT that can accommodate both the conventional 1 Gbps EPON and the 10 Gbps 10 G-EPON ONU, and this OLT. It defines the functions that the ONU should realize. Similarly, ITU-T is also standardizing 10 Gbps class PON.

  When these international standards are applied, the communication speed in the PON section becomes 10 times. In this case, two merits can be considered. The first merit is that the communication carrier operates with the same number of subscribers per OLT as the number of subscribers of a 1 Gbps class PON, so the communication capacity per subscriber increases and the subscribers It is to be able to benefit from the increased capacity of the section. The second advantage is that the communication carrier accommodates more subscribers per OLT than the 1 Gbps class PON, thereby lowering the operation cost and allowing the subscriber to be equal to or more than the 1 Gbps class PON. Service (communication capacity).

  In general, 1-Gbps class PON star couplers use a maximum of 32 to 64 branches, and the number of branches beyond that is almost realized due to the small communication capacity and loss of the optical transmission line due to the branching of the star coupler. Although not done, the number of ONUs accommodated by multi-branching has been studied conventionally. For example, in Patent Document 1, a plurality of PON signals with different wavelengths handled by a plurality of station side devices installed in a central office are coupled to a WDM coupler that multiplexes or demultiplexes each wavelength, and is multiplexed by a WDM coupler. The wavelength-division multiplexed signal is connected to a plurality of transponders capable of transmitting and receiving a PON signal having a wavelength corresponding to a specific station-side device using one optical fiber and a star coupler, and the transponder and the plurality of ONUs are connected to one light. There is disclosed a multi-branching method in which a fiber and a star coupler are connected and a PON signal is transmitted and received by time division multiplexing in the upstream direction.

  Also, when considering the effective use of a 10 Gbps class PON device, it is possible to increase the number of subscribers per OLT and reduce the operating cost of the communication carrier, rather than just considering the improvement of the communication speed per subscriber. It is desirable to have an operational form that improves satisfaction.

  Applying this type of operation and sharing the OLT with a larger number of subscribers results in lowering the OLT power consumption (total power consumption of all OLTs on the station side) and spreading it all over the world. This is in line with the trend toward lower power consumption.

  However, only with the provisions stipulated in the international standard of 10 Gbps class (Non-Patent Document 2 above), the transmission quality deteriorates due to multi-branching, so that the distance between the currently served central office and the subscriber's home is not reduced. It is difficult to realize branching. For example, it is possible to achieve both service distance and multi-branching by increasing the power of OLT and ONU transmitters or inserting an amplifier in the fiber, but the transmission power of OLT and ONU is high. Then, operation and technical problems occur, such as a higher hazard level for access fiber laying workers and difficulty in amplifying an upstream burst signal with an amplifier.

  Further, in the method disclosed in Patent Document 1, the configuration of the station-side device is not changed from the conventional configuration, and a transponder is newly inserted, so that multi-branching is possible, but low power consumption is achieved. The problem that is difficult. Further, as a problem common to the conventional system, when a service is provided to a larger number of subscribers with one OLT, the central office device and the subscriber device are connected by a single fiber. When a fiber (especially a trunk line) breaks down, the problem that a failure area expands arises.

  Therefore, in order to solve the above-mentioned problems, in a PON apparatus of 10 Gbps class or higher, a subscription that allows power saving and a redundant configuration while allowing an increase in the number of subscribers per OLT (multiple branching) is allowed. For the purpose of providing a person accommodation device, a method of consolidating more ONUs into one OLT than before is disclosed by the fusion of OTN technology and PON technology (for example, Non-Patent Document 3).

JP 2008-206008 A

IEEE802.3 IEEE802.3av IEICE Technical Report CS2010-9 (2010-7)

  In Non-Patent Document 3 described above, the PON function is integrated into the OLT, and the PON signal is directly accommodated in the OTN (Optical Transport Network) signal to be transmitted and received, thereby increasing the number of branches, lengthening, and network reliability. A method for realizing low power consumption is shown. However, in this conventional technique, there is a problem relating to a method, method, and means for linking bandwidth control of an optical transport network and an access network with an OLT (hereinafter referred to as an aggregate OLT) arranged on the optical transport network. In other words, because bandwidth control was performed separately for the optical transport network (metro network) and the access network, the upstream traffic from the access network to the optical transport network has not increased so much over a long period of time. There has been a problem that efficient use of the upstream bandwidth has not been realized, for example, the upstream traffic from the access network is instantaneously concentrated, congestion occurs, and some traffic is discarded.

  The present invention has been made in view of the above, and obtains an optical communication system, a communication device, and a bandwidth control method that efficiently use an upstream bandwidth by linking bandwidth control of a metro network and an access network with an aggregation OLT. For the purpose.

  In order to solve the above-described problems and achieve the object, the present invention constitutes a ring network, extracts signal light of a specific wavelength from the ring network, and does not constitute the ring network. A plurality of ring nodes that output to the ring network the signal light input from the other nodes, a remote node that operates as the other nodes, and the remote node via an optical coupler An optical communication system configured to include a connected ONU, wherein a part of the ring nodes operates as an OLT that executes a process of allocating an upstream band to the ONU, and the OLT The operating node treats ONUs under the same remote node as one ONU group and assigns them to its own system. A first bandwidth allocation function for distributing the assigned upstream bandwidth to each of the ONU groups based on an upstream bandwidth allocation request amount from each ONU and the number of ONUs belonging to each ONU group; and the ONU group And a second bandwidth allocation function for redistributing the upstream bandwidth distributed by the first bandwidth allocation function to each ONU in the group based on an upstream bandwidth allocation request amount from each ONU. Features.

  According to the present invention, it is possible to dynamically allocate a band in a ring to a plurality of remote nodes so that a plurality of remote nodes share the same wavelength and interlock with a bandwidth amount allocated to the ONU. Metro access fusion type bandwidth allocation that is more flexible than the transport network that allocated bandwidth quasi-statically. As a result, there is an effect that it is possible to realize an optical communication system capable of efficiently using the uplink band by preventing congestion from occurring in the transmission path to the upper network.

FIG. 1 is a diagram showing a configuration example of an optical communication system according to the present invention. FIG. 2 is a diagram illustrating a configuration example of the OLT, the ring node, and the RN and an outline of the control method. FIG. 3 is a diagram illustrating the relationship between the metro bandwidth update period and the access bandwidth update period. FIG. 4 is a flowchart showing an example of the overall operation of bandwidth allocation control. FIG. 5 is a flowchart showing an example of bandwidth allocation control for RN3. FIG. 6 is a flowchart showing an example of bandwidth allocation control for RN3. FIG. 7 is a flowchart showing an example of bandwidth allocation control for the transponder. FIG. 8 is a diagram showing an example of a ranging procedure in the optical communication system according to the present invention. FIG. 9 is a diagram illustrating an example of a bandwidth allocation operation. FIG. 10 is a flowchart showing a procedure for updating the GST and a procedure for updating the reference time of the band update cycle. FIG. 11 is a flowchart showing a procedure for updating the reference time of the band update cycle.

  Embodiments of an optical communication system, a communication apparatus, and a bandwidth control method 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.
<System configuration>
First, the overall configuration of the system will be described. FIG. 1 is a diagram showing a configuration example of an optical communication system according to the present invention. As shown in the figure, an optical communication system includes a ring network having a configuration in which a plurality of nodes (OLT1 and ring nodes 2 ₁ to 2 _n ) are connected by a duplexed optical fiber (trunk optical fiber), and a duplex network. Remote nodes (RN) 3 _{11 to} 3 _1p , 3 _{21 to} 3 _2q ,..., 3 _{n1 to} 3 _nr connected to the ring nodes 2 ₁ to 2 _n through the optical fiber, an optical fiber and an optical coupler ONUs (10G-ONU5 ₁₀ , 1G-ONU5 ₁ , 10G · 1G-ONU5 ₁₀₁ ) connected to the RN via a (32 branch coupler) 4. Incidentally, 10G-ONU 5 ₁₀ is capable of uplink and downlink bidirectional 10Gbps communication ONU (e.g., those defined in IEEE802.3av), 1G-ONU5 ₁ is capable of uplink and downlink bidirectional 1Gbps communication ONU 10G · 1G-ONU5 ₁₀₁ is an ONU (for example, specified in IEEE802.3av) that can communicate at an uplink of 1 Gbps and a downlink of 10 Gbps. In the following description, when “ONU” is simply described, 10G-ONU5 ₁₀ , 1G-ONU5 ₁ and 10G · 1G-ONU5 ₁₀₁ are indicated. Further, an RN may be connected to the OLT 1 corresponding to the above-described aggregate OLT, and various ONUs may be connected to the RN. In this embodiment, the case where the optical coupler 4 is a 32-branch coupler will be described, but other branch numbers may be used. In FIG. 1, the communication means of the metro network exemplifies the transfer of an OTU (Optical-channel Transport Unit) frame (frame used in OTN) by continuous optical WDM. This is an understanding of the principle of the present invention. It is merely an example for helping, and does not limit the physical communication means in the metro network. If the communication method has the same effect, the physical layer may be, for example, OFDMA communication or digital coherent communication.

  In the optical communication system shown in FIG. 1, an OLT 1 is a communication device that connects an upper network and a ring network, and includes a demultiplexing unit, two redundant OLT-IFs, a monitoring control unit, a ROADM ( Reconfigurable Optical Add / Drop Multiplexer) section. Each ring node is a communication device having the same configuration, and includes a ROADM unit, an OTN-XC (OTN cross connect) unit, and a transponder. Each RN is a communication device having the same configuration, and includes a transponder and a monitoring control unit. The OLT 1 is a communication device having a function of operating as a PON OLT in addition to the function of each ring node. Conventionally, the optical communication system of the present embodiment is a communication device under each ring node (FIG. 1). The OLT functions (functions that operate as OLTs) possessed by the communication device at the location where each RN is located in are integrated into one ring node to form OLT 1. In the optical communication system shown in FIG. 1, each ring node includes an OTN-XC unit, so that a plurality of ring nodes can share the same wavelength.

  The demultiplexing unit of OLT 1 outputs the downstream signal light (frame) received from the upper network to one of the two OLT-IFs, and outputs the upstream signal light received from the OLT-IF to the upper network. To do. The OLT-IF executes various processes for the OLT to operate as a station-side device of the PON. One of the two OLT-IFs is set as the active system and the other is set as the standby system. The monitoring control unit performs monitoring (failure detection) in the communication system, path switching at the time of failure detection, and the like. The ROLT unit of the OLT 1 and each ring node extracts and adds a specific wavelength (output to the ring network) for a signal (light) flowing on the ring network. The transponder of each ring node performs wavelength conversion of the received signal light. Each RN transponder also converts the wavelength of the received signal light. In the upstream direction, a plurality of optical burst signals having different communication speeds are terminated, and light of at least one wavelength for RN and OLT 1 to communicate with each other. A process of converting to a continuous signal is also performed. On the other hand, in the downlink direction, a signal for each transponder is extracted from a signal having a predetermined communication wavelength predetermined for each RN, and a communication signal is transmitted in a predetermined frame format at a wavelength specific to the PON defined by the standard. To do.

OLT1 and RN3 ₁₁ ~3 _1p (p is the number RN which is connected under the ring node _{2 1), RN3 21 ~3 2q} (q is the number RN which is connected under the ring node _{2 2), RN3 n1 ~3 nr} ( r is the number of RNs connected under the ring node 2 _n ) is connected via one or more ring nodes 2 ₁ to 2 _n (n is the number of ring nodes constituting the ring network), and RN And the ONUs (10G-ONU5 ₁₀ , 1G-ONU5 ₁ , 10G · 1G-ONU5 ₁₀₁ ) under its control by performing OTU frame transfer by continuous optical WDM at a higher transmission speed than the communication speed. On the other hand, the transponders accommodated in each RN and the ONUs connected to the transponders are connected in a star topology like the conventional PON and communicate by the conventional TDMA-PON system.

  In the present embodiment, for simplification of description, as shown in FIG. 1, optical communication having a configuration in which the OLT function is integrated in one of the nodes (communication devices) constituting the ring network. The explanation is given assuming the system. However, it is not essential to consolidate into one node. The OLT function may be distributed to some nodes.

<Outline of configuration and control of OLT, ring node and remote node (RN)>
2 shows the configuration of the OLT 1, the ring node 2 (ring nodes 2 ₁ , 2 ₂ ,..., 2 _n ) and the RN 3 (RN 3 _{11 to} 3 _1p ,..., RN3 _{n1 to} 3 _nr ) and the entire system shown in FIG. It is a figure explaining the principal point of this control system. Here, as an example, in this communication system, it is assumed that the OLT-IF performs all band allocation control in the system. In addition, the OLT-IF performs band allocation control using a PON control protocol that conforms to the IEEE standard between the RN and the ONU. In FIG. 2, one ring node 2 is used, but as shown in FIG. Since each RN3 has the same configuration, only one unit shows the internal configuration, and the rest are omitted.

  The OLT 1 includes an OLT-IF 11, a monitoring control unit 12, and a ROADM unit 13 as main components. In order to simplify the description, only one of the two redundant OLT-IFs (the side set as the active system) is described as the OLT-IF 11. Further, the description of the demultiplexing unit is omitted. The OLT-IF 11 includes a highly integrated PON control unit including an IF unit 111, an OTN mapper / demapper unit 112, a SER / DES unit 113, and an uplink bandwidth allocation control unit 114, and an OptTRx unit 115. The IF unit 111 is an interface with a demultiplexing unit (not shown). The OTN mapper / demapper unit 112 multiplexes the main signal input from the IF unit 111 and the control signal input from the uplink bandwidth allocation control unit 114, and is multiplexed from the SER / DES unit 113. A process for separating the main signal and the control signal in a separated state is performed. The SER / DES unit 113 performs processing for converting the serial data input from the OTN mapper / demapper unit 112 into parallel data, and processing for converting the parallel data input from the OptTRx unit 115 into serial data. The uplink bandwidth allocation control unit 114 performs bandwidth allocation control in the metro network and the access network. The OptTRx unit 115 transmits and receives optical signals.

  As already described, the monitoring control unit 12 performs monitoring (failure detection) in the communication system, path switching at the time of failure detection (optical path switching control), and the like. As already described, the ROADM unit 13 extracts and adds a specific wavelength (output to the ring network) for a signal (light) flowing on the ring network.

  The ring node 2 includes a ROADM unit 21 and an OTN-XC unit 22 as main components. In addition, description of the transponder is omitted. Similar to the ROADM unit 13 of the OLT 1, the ROADM unit 21 extracts and adds an optical signal having a specific wavelength for a signal flowing on the ring network. The OTN-XC unit 22 has a function of converting a signal of a specific wavelength extracted by the ROADM unit 21 into an electrical signal, extracting a signal to be transferred to the RN 3, reconverting it to an optical signal, and connecting And a function of placing received signals from the respective RNs 3 on the light of the same wavelength based on an instruction from the OLT-IF and transferring the received signals to the optical transport network side.

  The RN 3 includes an OptTRx unit 31, an OTN mapper / demapper unit 32, and a plurality of burst transponders 33 (hereinafter simply referred to as transponders 33). Note that the description of the monitoring control unit is omitted. The configuration of each transponder 33 is the same. Therefore, the internal configuration is shown only for one transponder 33, and the description of the rest is omitted.

  The OptTRx unit 31 transmits and receives an optical signal to and from the ring node 2. In the upstream direction (direction toward the ring node 2), the OTN mapper / demapper unit 32 encapsulates the input signal from each transponder 33, generates a frame (OTU frame) handled by the OTN, and outputs the frame to the OptTRx unit 31. . On the other hand, in the downstream direction, the decapsulation of the OTU frame input from the OptTRx unit 31 is performed, and the signal for each transponder 33 obtained as a result is output to the corresponding transponder 33.

  The transponder 33 includes an RTT management unit 331, a DOB CDR unit 332, a DOB Rx unit 333, a time stamp processing unit 334, a 1G / 10G OptTx unit 335, a WDM coupler 336, and a local timer 337. The RTT management unit 331 measures the RTT of the PON section (between the own device (own transponder) and each subordinate ONU) and manages the measurement result. In addition, the managed RTT is notified to the OLT 1 as necessary. The DOB CDR unit 332 and the DOB Rx unit 333 receive the upstream signal light having a communication speed of 1 Gbps and 10 Gbps, specifically, a burst signal of two wavelengths transmitted from the ONU via the WDM coupler 336, and receive 1 Gbps and 10 Gbps. As a continuous signal. The time stamp processing unit 334 adds a time stamp indicating the time managed by the local timer 337 to the PON control frame generated by the OTN mapper / demapper unit 32 based on the control signal from the OLT. The 1G / 10G OptTx unit 335 converts the wavelength of the frame (1 Gbps and 10 Gbps PON frame) input from the time stamp processing unit 334 and is connected to the star coupler (optical coupler 4) via the WDM coupler 336. Transmit to the optical fiber side. In the PON control frame input to the 1G / 10G OptTx unit 335, a time stamp indicating the time managed by the local timer 337 is embedded. The local timer 337 is configured by a 16-ns granular 32-bit counter that is used as standard in PON control, and manages the local time for each transponder.

  In each RN3 configured as described above, in the frame processing operation in the downlink direction, each transponder 33 identifies the MPCP frame, and a time stamp indicating the time (local time) managed by the local timer 337 for each transponder 33 is given to this. For example, the information in the Gate frame is corrected. On the other hand, in the upstream frame processing operation, each transponder 33 uses the local timer value managed by the local timer 337 and the time stamp value stamped in the received MPCP frame to generate an RTT for each logical link. Is calculated and held. In FIG. 2, the configuration example in which each of the transponders 33 includes the local timer 337 is illustrated. However, the transponder 33 or the RN 3 includes the local timer, and the transponder operates according to the local time managed by the local timer. You may do it. By doing in this way, time synchronization peculiar to TDM-PON controls between ONU5 and RN3, and PON control represented by ONU Discovery and band allocation control is collected in OLT-IF11 of OLT1. In other words, the OLT 1 that is the aggregate OLT performs the control of the discovery of the ONU 5 and the calculation of the upstream bandwidth allocation amount for the ONU 5, but the part related to the time correction depending on the time synchronization and the connection distance of the ONU 5 is terminated at the RN 3 Centralize the main functions of PON control. Even in such a control, the bandwidth request amount of the ONU 5 is transferred to the OLT-IF 11, so that the OLT-IF 11 can grasp the bandwidth request amount of the entire network. Access link bandwidth allocation control is possible.

<Mechanical zone dynamic bandwidth allocation method linked to access zone bandwidth control>
In the optical communication system according to the present embodiment, when the OLT-IF 11 of the OLT 1 allocates an upstream band to each ONU 5, first, the bandwidth is allocated in the metro section, and after this is completed, the bandwidth is allocated in the access section. In bandwidth allocation in the metro section, the ONUs 5 accommodated in the same RN 3 are treated as one group, and the upstream bandwidth request amount from each group (the sum of the upstream bandwidth requested from the ONU 5 in the same group) and each group belong to each group. Based on the number of ONUs 5 and the like, how to allocate the upstream bandwidth allocated to the system to each group is determined. Hereinafter, the bandwidth control method for the access section will be described in detail.

  First, a method for setting a bandwidth allocation update period in a metro section (section between OLT 1 and each RN 3) will be described. FIG. 3 is a diagram illustrating an example of a method for setting a bandwidth allocation update period in a metro section. This method focuses on the fact that the bandwidth allocation in the metro section can be expected to have a statistical multiplexing effect, so that dynamic allocation control with a short bandwidth update period is not required unlike PON control. Therefore, the OLT 1 monitors the ONU request amount (REPORT amount) under each RN 3 over a plurality of bandwidth update periods in the PON interval (the interval between each transponder 33 of the RN 3 and each ONU 5), and the bandwidth in the PON interval. An integrated value of the bandwidth request amount is measured in the metro bandwidth allocation cycle L set in a cycle longer than the update cycle (PON bandwidth allocation cycle 1 shown in the figure), and the bandwidth allocation amount in the metro network is based on the integrated value. Is calculated. In FIG. 3, the bandwidth allocation period (L) in the metro network is a fixed period, but the period may be changed by monitoring the traffic fluctuation amount. In this case, the minimum cycle of the bandwidth update cycle (metro bandwidth allocation cycle) in the metro network is the same as the bandwidth update cycle (PON bandwidth allocation cycle) of the PON section. As an example, in a transitional state where the metro band transitions from a non-congested state to a congested state (or vice versa), the metro band allocation period L is shortened so that the bandwidth allocation control of the metro section with respect to the traffic increase amount of the access section is reduced. In a state where the followability is improved and the traffic fluctuation is small, the metro band allocation period L may be set to be long, and control may be performed to reduce the processing amount of the OLT 1.

  Next, a bandwidth allocation method for the metro section (the OTN-XC unit 22 of each ring node 2) and the access section (each ONU 5) will be described with reference to FIGS. In the following, a case where one RN 3 is connected to the OTN-XC unit 2 will be described in order to facilitate understanding of the principle of the bandwidth allocation method of the present embodiment. However, even when a plurality of RNs 3 are connected to the OTN-XC unit 22, the bandwidth allocation amount can be calculated by the same method. Although the description will be made assuming that there is no RN3 connected to the OLT 1, the bandwidth allocation amount can be calculated by the same method even when the RN3 is connected to the OLT 1.

  FIG. 4 is a flowchart showing an example of the overall operation of bandwidth allocation control, FIGS. 5 and 6 are flowcharts showing an example of bandwidth allocation control for each RN 3, and FIG. 7 shows an example of bandwidth allocation control for each transponder 33. It is a flowchart.

As shown in FIG. 4, the bandwidth allocation control performed by the OLT 1 in the optical communication system according to the present embodiment includes the RN _i bandwidth allocation processing (BW _i calculation) that is the bandwidth allocation processing for each RN 3 and the transponders 33. It includes transponder bandwidth allocation processing (bw _ij calculation) that is bandwidth allocation processing, and ONU bandwidth allocation processing (nu_bw _ijk calculation) that is bandwidth allocation processing for each ONU 5. In the OLT 1, the uplink bandwidth allocation control unit 114 performs bandwidth allocation control.

In the bandwidth allocation control shown in FIG. 4, when the process is started, it is confirmed whether or not the metro bandwidth allocation cycle timer has expired, that is, whether or not the metro bandwidth allocation cycle L shown in FIG. 3 has passed (step S1). ). If the metro bandwidth allocation cycle timer has not expired (step S1: No), it is confirmed whether the access bandwidth allocation cycle timer has expired, that is, whether the PON bandwidth allocation cycle 1 shown in FIG. Step S3). On the other hand, if the metro bandwidth assignment period timer has expired (step S1: Yes), executes the RN _i bandwidth allocation process to be described later access bandwidth allocation cycle timer after it is confirmed whether the expired (step S2 → S3). If the access bandwidth allocation cycle timer has not expired (step S3: No), the process returns to step S1 to check whether the metro bandwidth allocation cycle timer has expired. On the other hand, when the access band allocation cycle timer expires (step S3: Yes), a transponder band allocation process and an ONU band allocation process described later are executed (step S4 → S5), and the process returns to step S1. The OLT 1 performs band allocation control according to such a procedure, and realizes efficient use of the uplink band by linking the bandwidth control of the metro network and the access network.

Next, the band allocation method of Metro section (RN _i bandwidth allocation processing in step S2) will be described with reference to FIGS. Note that FIG. 5 and FIG. 6 both show a method of allocating an upstream bandwidth for RN3, exclusively use them as RN _i bandwidth allocation processing in step S2. First, the process shown in FIG. 5 will be described.

In FIG. 5, “B” indicates the maximum bandwidth of the optical path, “Req (i, k)” indicates the upstream bandwidth request amount from ONU _k of RN _i , and “N _i ” is connected to RN _i. It is assumed that “BW _i ” indicates the allocated bandwidth for RN _i , and “L” indicates the bandwidth update period of the metro section. The number of RN3 in the system is n + 1, OLT 1, compared n + 1 single RN3, determines the upstream band BW _i to be allocated to RN3 each executes step S11~S16 individually. The maximum bandwidth B of the optical path is the maximum bandwidth of the optical path between the OLT 1 and the upper network.

In the band allocation process for RN3, first, whether Σ _i Σ _k Req (i, k) [bit] / L [sec] ≦ B [bps] (where i = 0, 1,..., N), that is, Whether the time average value (hereinafter referred to as “total required amount average value”) of the sum of the requested upstream bandwidths requested by all ONUs 5 in the system in the metro bandwidth allocation period L is equal to or less than the maximum bandwidth B of the optical path. Confirm (step S11).

If the total required amount average value is equal to or less than the maximum band B (step S11: Yes), i-th allocation band BW _i the sigma _k Req for RN3 (i, k) and determines the [bit] / L [sec] A process of incrementing the value of i is executed for a range of i = 0 to n (steps S12 and S17). That is, BW _i is obtained by averaging the sum of the requested upstream bandwidths requested from each ONU 5 under the i-th RN 3 in the metro bandwidth allocation period L. This indicates that the bandwidth is allocated according to the upstream bandwidth request amount requested from all ONUs 5 under the i-th RN3.

On the other hand, if the total required amount average value exceeds the maximum band B (Step S11: No), the allocated bandwidth BW _i for the i-th _{RN3 Σ k Req (i, k} ) in [bit] / L [sec] A provisional decision is made (step S13). Whether or not BW _i > N _i × B / Σ _i N _i , that is, the temporarily determined BWi is larger than the band when the maximum band B is proportionally distributed to each RN 3 according to the number of ONUs under each RN 3 Whether or not (step S14).

When BW _i > N _i × B / Σ _i N _i is satisfied (step S14: Yes), N _i × which is a bandwidth that is proportionally distributed to each RN 3 according to the number of ONUs under each RN 3 B / Σ _i N _{i is set} to BW _i (step S15). That is, N _i × B / Σ _i N _i is assigned to the i-th RN 3. On the other hand, if BW _i > N _i × B / Σ _i N _i is not satisfied (step S14: No), BW _i temporarily determined in step S13 is set as the final BW _i ( Step S16). That is, a bandwidth is allocated to the i-th RN 3 in accordance with the uplink bandwidth request amount requested from all ONUs 5 under the i-th RN 3. i th After determining the allocated bandwidth BW _i for RN3 increments the value of i (step S18), and if there is RN3 not assign a band to repeat the process of steps S11 to S16.

As described above, in the bandwidth allocation for the metro section, it is determined whether or not there is a possibility of congestion in the OLT 1 (in the transmission path from the OLT 1 to the upper network). A bandwidth corresponding to the sum of bandwidth request amounts from each ONU 5 for each RN 3 (band request amount from a group of ONUs 5 under the same RN 3) is allocated to each RN 3. On the other hand, if there is a possibility of occurrence of congestion, it is further determined for each RN3 whether or not there is a possibility of congestion when the maximum bandwidth B is proportionally distributed to each RN3 according to the number of ONUs under each RN3. If there is no possibility of congestion occurrence, the same bandwidth as the total amount of bandwidth requests from each ONU 5 under RN3 is allocated to RN3, and if there is a possibility of congestion occurrence, maximum bandwidth B is assigned to the number of ONUs under each RN3. Accordingly, a bandwidth corresponding to a value when proportionally distributed to each RN3 is allocated. That is, when there is a possibility of congestion in the OLT 1, the bandwidth required when the maximum bandwidth B is proportionally distributed to each RN 3 according to the number of ONUs under each RN 3 is set as the upper limit for each bandwidth request amount RN 3 from the ONU 5. The amount of bandwidth allocated to each RN 3 is determined according to the sum of. Therefore, for all RN3, the total bandwidth request amount from the subordinate ONU 5 is N _i × B / Σ _i N _i (= the maximum bandwidth B is proportionally distributed to each RN 3 according to the number of ONUs under each RN 3 Even if it is in a state exceeding the value, the allocated bandwidth BW _i for each RN 3 is N _i × B / Σ _i N _i , and the sum of the allocated bandwidth BW _i for each RN 3 is the maximum bandwidth B of the optical path. The following can be suppressed. As a result, it is possible to prevent the occurrence of congestion in the OLT 1 and the discard of upstream traffic.

Note that, as a result of assigning the bandwidth to all RNs 3 by the above method, when a residual bandwidth is generated, the residual bandwidth may be divided by the number of RN 3 and assigned to each RN 3 equally. Alternatively, all the remaining bandwidths may be given to a certain RN, and may be given to each RN 3 in turn for each bandwidth update cycle (metro bandwidth allocation cycle). Further, it may be given preferentially to RN3 are congested residual bandwidth (RN3 that the _{BW i = N i × B /} Σ i N i). Further, the remaining bandwidth may be allocated to the RN 3 having the smallest bandwidth allocation amount. When the remaining bandwidth is allocated to the congested RN 3, further improvement in bandwidth utilization efficiency can be realized.

In the above-described control shown in FIG. 5, the control is performed such that BW _i = N _i × B / ΣN _i (bps) is uniformly given to the congested RN 3, but as shown in FIG. It is good also as a control. In addition to the control shown in FIG. 5, the control shown in FIG. 6 adds step S21 for calculating a control amount x _i (t) at the time of congestion, which is a bandwidth allocated to the congested RN 3, and the control shown in FIG. as the processing for determining the allocated bandwidth BW _i for it has RN3, it is obtained so as to perform the step S22 instead of step S15. According to this control, the OLT 1 calculates the control amount x _i (t) at the time of congestion of each RN in advance, and performs bandwidth allocation based on the control amount at the time of congestion.

In step S21 shown in FIG. 6, the control amount x _i (t) at the time of congestion is calculated according to the following equation (1). “X _i (t) [bps]” represents the control amount during congestion of the current cycle for RN _i , and “y _i (t) [bps]” represents the average bandwidth allocation amount for RN _{i up} to the previous cycle. , “Y0 _i [bps]” indicates a bandwidth allocation target value for RN _i . Further, Δy _i (t) = y _i (t) −y0 _i , y0 _i = N _i × B / ΣN _i . K _P represents a proportional gain, K _I represents an integral gain, and K _D represents a differential gain.

In the control shown in FIG. 6, when BW _i > N _i × B / Σ _i N _i is satisfied (step S14: Yes), x _i (t) calculated in step S21 is set to BW _i. (Step S22).

That is, the process shown in FIG. 6 can be expressed by the following equation (2).
If {Σ _i Σ _k Req (i, k)> B and BW _i > y0 _i }
then BW _i = x _i (t)
else
BW _i = Σ _k Req (i, k) / L (2)

In this process, control is performed so that the metro allocated bandwidth approaches the target bandwidth y0 _i when RN _{i is} congested. In bandwidth allocation at the time of congestion, not only a proportional term but also a control having an integral term and a differential term is performed, so that it is possible to eliminate residual deviation and improve response speed. Note that the control amount during congestion may be allocated while shuffling the RN number (i) every cycle if fairness of the allocated amount cannot be maintained if the allocation is performed in the same order each time (BW _i is determined). The order may be changed every time). Further, the derivative term may not be applied to the control.

<Access section bandwidth allocation method>
Next, a bandwidth allocation method for the access section will be described.

The OLT 1 performs bandwidth allocation for the access section when the bandwidth allocation for the metro section according to the procedure shown in FIG. 5 or FIG. 6 is completed. In the bandwidth allocation of the access section, the ONU 5 connected to each transponder included in the RN 3 determines how the bandwidth distributed to each RN 3 (group of ONUs 5 under the same RN 3) is allocated to each ONU 5 in the group. The number is determined based on the number, the requested upstream bandwidth from the ONU 5, and the like. That is, it is determined how the bandwidth (BW _i ) assigned to each RN3 (group of ONUs 5 under the same RN3) is distributed to each transponder 33 in RN3 by bandwidth assignment in the metro section (step S4 in FIG. 4). : Transponder band allocation process), and further, how to distribute the band distributed to each transponder 33 to each ONU 5 accommodated in the transponder 33 is determined (step S5 in FIG. 4: ONU band allocation process). Thus, by performing the bandwidth allocation of access period not exceeding the bandwidth allocation amount BW _i metro section, the alignment of the band is taken of the bandwidth and access interval Metro interval allocated to each RN3, each PON It is possible to avoid a bottleneck in a transmission line (transmission line between each RN 3 and the ring node 2) between the PON and the device accommodating the PON. Details of the bandwidth allocation operation in the access section are as follows.

In the bandwidth allocation of the access section, the OLT 1 performs the processing shown in FIG. 7 for each RN 3 and determines the bandwidth allocated to each transponder 33. In the process shown in FIG. 7, for the i-th RN3, the band bw _ij to be allocated to the j-th (j = 0, 1,..., M) transponder 33 included in the RN3 is BW _i × n _ij / N _i is set (step S31). “N _ij ” indicates the number of users (number of ONUs) connected to the j-th transponder among the transponders 33 included in the i-th RN 3.

Further, by repeatedly executing the process of step S31 while changing the value of j in step S32, a band is allocated to all the transponders 33 included in the i-th RN3. That is, the bandwidth BW _i assigned to the i-th RN3, depending on the number of users (the number of ONU 5) contained in each of the transponders 33 that this RN3 is provided, proportional distributed to each transponder 33. In step S33, the process of steps S31 and S32 is repeatedly executed while changing the value of i, and the bandwidth allocated to each transponder 33 is determined for all RN3 in the system.

When the bandwidth to be allocated to all the transponders 33 in the system is determined according to the control of FIG. 7, the OLT 1 further accesses the access section for each ONU 5 based on the bandwidth request amounts from the plurality of ONUs 5 connected to each of the transponders 33. (The uplink bandwidth is assigned to each ONU 5). The bandwidth allocation amount in the access section may be bandwidth allocation control as performed in the conventional PON. As an example, when the sum of bandwidth request amounts from all ONUs 5 under a certain transponder 33 is less than bw _ij , the requested bandwidth amount is given to all ONUs 5, and a preset bandwidth is given to the ONU 5 that exceeds bw _ij. After giving the amount, there is a method of allocating the remaining bandwidth by an arbitrary allocation algorithm. Incidentally, in the band allocation for the access period, after allocating bandwidth all the ONU5 under RN _i as described above, if there is more residual bandwidth, the residual bandwidth in addition to ONU5 under transponders congested You may make it give.

  With such a control method and calculation algorithm, it is possible to link the bandwidth allocation between the metro section and the access section, and flexible bandwidth allocation is possible, so that efficient use of the uplink bandwidth can be realized.

<Outline of OLT bandwidth allocation result notification method and uplink transmission operation>
Of the bandwidth allocation amounts of the metro network and access network calculated in the above-described procedure, the OLT 1 stores the bandwidth allocation amount of the metro network in the header information of the OTU frame and stores it in the OTN-XC unit 22 of each ring node 2. Notice. For example, the OLT 1 acquires the ID of the OTN-XC unit 22 in the network by an arbitrary protocol, and obtains the band allocation result (band allocation amount) for each OTN-XC unit 22 and the notification destination ID (each OTN-XC unit 22 OTU frames each including the ID acquired from (1) are generated and transmitted to the ring network. At this time, an OTU frame is transmitted with an optical signal having a wavelength allocated to the ring node 2 provided with the OTN-XC unit 22 that is a notification destination of the bandwidth allocation amount. Further, the notified bandwidth allocation amount is the sum of the bandwidth allocation amounts (BW _i ) to each RN 3 accommodated in the OTN-XC unit 22. In each ring node 2, the ROADM unit 21 extracts an optical signal having a wavelength assigned to the own ring node, and outputs an OTU frame included in the extracted optical signal to the OTN-XC unit 22. When the received OTU frame includes its own ID and bandwidth allocation amount, the OTN-XC unit 22 determines that this bandwidth allocation amount is a bandwidth allocation result for itself. On the other hand, the amount of bandwidth allocated to the ONU 5 determined by the bandwidth allocation of the access network is notified to each ONU 5 using a Multipoint Control Protocol (MPCP) Gate frame. The Gate frame is generated by the OLT 1 and is encapsulated and transferred as an OTU frame in the section up to the RN 3. Alternatively, information indicating the bandwidth allocation amount is transmitted from the OLT 1 to the RN 3 by the OTU frame, and the RN 3 generates a Gate frame using the transmitted bandwidth allocation amount information and transmits it to the subordinate ONU 5.

  Each OTN-XC unit 22 and each ONU 5 that have received the notification of the band allocation result transmit data to the OLT 1 according to the notification content. Since the transmission operation of the ONU 5 is the same as the ONU of a general PON system, the description is omitted. When the OTN-XC unit 22 receives an uplink signal from the ONU 5 via each subordinate RN 3, the OTN-XC unit 22 stores and transmits the uplink signal in an OTU frame having a size corresponding to the bandwidth allocation amount notified from the OLT 1.

  As described above, in the optical communication system according to the present embodiment, the OLT 1 collects the upstream bandwidth request amount from all the ONUs 5 in the system as the bandwidth allocation control operation, and the same period as the bandwidth update period in the PON section or Based on the sum of the collected uplink bandwidth request amounts, the sum of the collected uplink bandwidth request amounts for each RN3, and the number of users accommodated in each RN3 (number of ONUs 5) in an integer multiple cycle (the same) The uplink band to be allocated to the RN3 subordinate group) is determined, and then the uplink band allocated to each RN3 unit is determined in each PON section (each transponder 33 included in each RN3) in the band update period in the PON section. Distribution is determined based on the number of ONUs 5 connected to each transponder 33 and each PON section How to allocate uplink bandwidth that is distributed to each ONU 5, it was decided to determine based on the uplink bandwidth request amount from each ONU 5.

  As a result, the total sum of the uplink traffic transmitted from each RN 3 can be adjusted by the OLT 1, so that the sum of the uplink traffic can be suppressed to be equal to or less than the maximum bandwidth in the transmission path with the upper network to which the OLT 1 is connected. Thus, it is possible to prevent the occurrence of congestion in the OLT 1 and the discard of the upstream traffic. As a result, it is possible to prevent the occurrence of retransmission due to the occurrence of congestion and realize efficient use of the uplink bandwidth.

  In the present embodiment, when determining how to allocate the upstream band allocated to RN3 to each transponder 33, the determination is made simply based on the number of connected ONUs 5 (see FIG. 7). . However, there are ONUs 5 having an uplink communication speed of 1 Gbps and 10 Gbps, and fairness may not be maintained when only the number of connected ONUs 5 is determined. Therefore, the OLT 1 may collect the ONU 5 type (uplink communication speed) together with the upstream bandwidth request amount, and determine the bandwidth distribution amount to each transponder 33 in consideration of the ONU 5 type. Further, the bandwidth distribution amount may be determined according to the sum of the upstream bandwidth request amounts from the respective ONUs 5 connected to the same transponder.

Embodiment 2. FIG.
In the system having the configuration described in the first embodiment, the OLT 1 operates as an aggregate OLT and allocates an upstream band to all ONUs 5 in the system. Here, since the upstream communication of the PON is performed in a time-sharing manner, it is important to manage the time synchronization between the OLT 1 and each ONU 5 in order to prevent the upstream signals burst transmitted from each ONU 5 from colliding with each other. Become. Therefore, in this embodiment, a time synchronization management method between the OLT 1 and each ONU 5 in the optical communication system described in the first embodiment will be described. In the optical communication system according to the present embodiment, RTT (Round Trip Time) is managed by performing ranging (measurement of transmission line delay) according to the following procedure, and uplink burst signals transmitted from different ONUs 5 collide with each other. To prevent it.

  FIG. 8 is a diagram illustrating an example of a ranging procedure in the optical communication system having the configuration illustrated in FIG. Here, as an example, a ranging procedure when registering an ONU will be described. In the PON transfer section (between RN and ONU) shown in FIG. 8, PON control MPCP frames are transmitted and received. The MPCP frame is encapsulated in the OTN transfer section (between OLT and RN) and transmitted and received as an OTU frame. However, description of the encapsulation / decapsulation operation in the RN is omitted. In order to simplify the description, the description will be made assuming that the RN includes a single transponder. Accordingly, in the ranging method described below, the process executed by the RN corresponds to the process executed by the transponder in the RN. In the OTN transfer section, necessary information may be extracted from the MPCP frame, and the necessary information may be mapped and transferred to the header of the OTU frame.

  In the control shown in FIG. 8, first, Discovery Gate frame information storing a relative value of a transmission start time (Grant start time, expressed as GST in the figure) is transmitted from the OLT to the RN (step S41). .

  The RN impresses a TS (time stamp) indicating the local time managed by the local timer in its own device with respect to the frame received from the OLT, and corrects the GST to an absolute time value (step S42). A PON frame is generated as needed and transferred to the ONU (step S43).

  The ONU that has received the Discovery Gate frame confirms the TS value (absolute time) embedded in the received frame, performs processing (TS synchronization) to synchronize the value of the local timer with this (step S44), and thereafter When the transmission start time (GST) specified in the Discovery Gate frame is reached, a Register Request frame with a TS indicating the value of the local timer of the ONU is transmitted toward the OLT (steps S45 and S46). By executing step S44, the local time of the ONU coincides with the local time of the RN when these steps S45 and S46 are executed.

  When the RN receives the Register Request frame, the RN transmits the Discovery Gate frame (TS value assigned to the Discovery Gate frame), the TS value embedded in the received frame, and the frame reception time (local timer at the time of reception). RTT is calculated based on (value) (step S47). Subsequently, the RN transfers Register Request frame information including the calculated RTT to the OLT (step S48).

  When receiving the Register Request frame information including the RTT, the OLT updates the RTT of the corresponding logical link based on the received RTT (step S49). Also, a predetermined process (such as a process of assigning a logical link to the ONU) for registering the frame transmission source ONU is executed. Next, the OLT transmits Register frame information indicating that the ONU registration process has been completed (step S50), and the RN generates a PON frame as necessary and imprints the TS on the received frame information. And transmit to the ONU (steps S51 and S52). When the ONU receives the Register frame, the ONU performs TS synchronization as in step S44 (step S53). Also, the setting inside the device is changed according to the information contained in the received frame. For example, the logical link assigned to the own device is grasped, and the identification information (logical link ID) is stored.

  The OLT further allocates a band for the ONU to transmit a response frame to the Register frame information transmitted in Step S50 (Grant generation), and transmits Gate frame information for notifying the band allocation result (Steps S54 and S55). ). In this Gate frame information, the ID of the logical link assigned to the ONU is embedded.

  When receiving the Gate frame information, the RN seals the TS indicating the local time of its own device instead of the embedded TS, corrects the GST based on the local time, and generates a PON frame as necessary. To the ONU (steps S56 and S57).

  When receiving the Gate frame, the ONU performs TS synchronization in the same manner as steps S44 and S53 (step S58). Also, the TS is stamped and a Register Ack frame is transmitted (steps S59 and S60). The Register Ack frame is transmitted using the band notified by the received Gate frame.

  When receiving the Register Ack frame, the RN calculates the RTT in the same manner as in Step S47 (Step S61), and transfers the Register Ack frame information including the calculated RTT (Step S62). When the OLT receives the Register Ack frame information including the RTT, the OLT updates the RTT of the corresponding logical link (step S63).

  In addition, when RN calculates RTT, it is not necessary to notify a calculation result to OLT every time. It is only necessary to notify the OLT at least when the previously calculated (previous) RTT is different from the newly calculated RTT (including when the RTT has not been previously notified to the OLT).

  As described above, when the RN receives a frame from the OLT, the RN updates the time stamp value (TS value) in the frame to the local time (local timer value), and transfers it to the ONU. When the received frame includes uplink bandwidth allocation information (in the case of a Discovery Gate frame or a Gate frame), the transmission start time (GST) included in the bandwidth allocation information is also updated and then sent to the ONU. Forward. Further, when a frame is received from the ONU, RTT is calculated. When the RN includes a plurality of transponders, the RN performs ranging according to the above-described procedure for each transponder. Also, each ONU accommodated in the RN is whether the TS value or GST of each frame received via the RN is based on the local time of the RN, and whether time synchronization is performed with the OLT, or It is not necessary to recognize whether time synchronization is performed with the RN. That is, it is sufficient to perform the same operation as that of a conventional ONU, and no special operation is required.

  Here, the band allocation operation including the GST correction operation executed in step S42 and the like will be described. FIG. 9 is a diagram illustrating an example of a bandwidth allocation operation, and illustrates an example of a bandwidth allocation operation for a network (denoted as ODN # k in FIG. 9) including a k-th RN and each ONU under the RN. ing. The operation in the case of targeting a network including an RN other than the k-th and an ONU under its control is the same as this.

In the bandwidth allocation operation in the optical communication system of the present embodiment, first, as Step (A), the OLT calculates the bandwidth allocation amount for each logical link (indicated as LL in the figure), and allocates the bandwidth for each logical link. Inform the RN of the result. At this time, the sum of the bandwidth allocation amounts for each logical link does not exceed the uplink bandwidth allocation amount (BW _k ) for the RN determined according to the procedure described in the first embodiment. Note that the band allocation result is notified to the RN using Gate frame (including Discovery Gate frame) information as described above.

As shown in the figure, the band allocation result is reported as a transmission start time (GST) for each logical link and a length of a period during which transmission is permitted (Grant Length, expressed as GL in the figure). FIG. 9 shows an example in which GST and GL (GL _k (LL)) of each logical link are calculated by setting GST (GST_LT _k (LL)) of the first logical link (LL = 1) to 0. . In addition, the GL is set to a value obtained by subtracting 10 from the interval (difference) between the GSTs so that transmission signals on the logical links do not collide. Note that the OLT does not need to consider the RTT between terminals connected to each logical link when determining the GST. The GST is corrected in the RN in consideration of the RTT between the RN and each of the ONUs under the RN. Therefore, when the OLT is in normal operation, the timing (time) at which the signal from each ONU arrives at its own device (OLT) Only need to be considered.

Next, as Step (B), the RN that has received the frame (Gate frame, Discovery Gate frame) information including the band allocation result (GST and GL) from the OLT, the local time, the RTT for each logical link, Based on the calculated reference time (RN_GST_base _k ), the information in the received frame (GST indicating the band allocation result) is corrected. For example, the reference time is uniformly added to the GST of each logical link, and further, the corrected GST is obtained by subtracting the corresponding RTT (RTT for each logical link) from each GST after the addition of the reference time. Note that GL is not changed.

For the bandwidth update period K, LL is the logical link ID, GST for each logical link calculated by the OLT is GST_LT _k (LL), the reference time is RN_GST_base _k , and the RTT for each logical link held by RN is RTT _k (LL ), GST _k (LL), which is an absolute GST (GST after correction), is given by the following equation. By performing such a calculation, it is possible to perform uplink communication control scheduled so that uplink bursts do not collide.
GST _k (LL) = GST_LT _k (LL) + RN_GST_base _k − RTT _k (LL)

  The RN transmits a Gate frame including each corrected GST generated by such a procedure and a GL for each logical link notified from the OLT to each ONU under its control.

  In the operation shown in FIG. 9, the OLT determines the GST without considering the RTT between the RN and each ONU, and the RTT (RTT for each logical link) and the local time between the RN and each ONU. However, the OLT may determine the GST in consideration of the RTT, and the RN may correct the GST in consideration of only the local time.

  FIG. 10 and FIG. 11 are flowcharts showing a procedure for the RN to update (correct) GST and a procedure for updating the reference time of the band update cycle. FIG. 10 is a flowchart showing the GST correction value calculation and reference time update procedure, and FIG. 11 is a flowchart showing the reference time update procedure. In the optical communication system according to the present embodiment, as shown in FIG. 10, when the GST correction value is calculated, the reference time is also updated. Further, as shown in FIG. As described above, an operation of updating the reference time without calculating the GST correction value is also performed.

As shown in FIG. 10, in the GST correction operation, when RN # k (kth RN) receives Gate information (information inserted in the Gate frame and Discovery Gate frame), it is received for the first time after activation. Check if it is Gate information, and if it is received for the first time, initialize the reference time (RN_GST_base _k ). In this initialization operation, a predetermined offset value (OFFSET) is added to the value of the local timer (indicated as TS in FIG. 10) to obtain a reference time after initialization.

When the initialization of the reference time is completed, or when the received Gate information is not received for the first time after activation, each GST (GST for each logical link) included in the received Gate information is updated. In this update operation, first, the reference time RN_GST_base _k is added to each transmission start time GST_LT _k (LL) included in the received Gate information to calculate RN_GST _k (LL) for each logical link. Next, from the calculated RN_GST _k (LL), the corresponding RTT for each logical link (RTT _k (LL)) is subtracted, and the updated transmission start time GST _k (LL) and To do.

When all the transmission start times are updated, the reference time is updated. Specifically, the one with the largest value is selected from the transmission start times GST _k (LL) for each logical link calculated by the above-described transmission start time update operation, and set as a new reference time RN_GST_base _k . Each RN performs such GST update operation every time it receives Gate information.

Each RN updates the reference time according to the procedure shown in FIG. 11 in addition to the operation shown in FIG. That is, each RN monitors the local time (Local_Timer) and the reference time (RN_GST_base _k ) in a state where the Gate information has been received, and if it detects a state where the local time is equal to or greater than the reference time, A predetermined offset value (OFFSET) is added to the current time value TS to obtain the updated reference time. The offset value is determined based on the difference between the local time and the reference time. Specifically, the local time is set to a value smaller than the reference time.

  As described above, in the optical communication system according to the present embodiment, each RN calculates and holds an RTT indicating an individual transmission delay amount with each subordinate ONU, and the OLT is accommodated in each RN. When the calculation result of the uplink bandwidth to be assigned to the ONUs being notified is determined, the set value of the transmission start time (GST) is determined without considering the RTT between each RN. The GST setting value included in the received frame is updated based on the local time of the device itself and the retained RTT. That is, the RN controls to synchronize the local time of each subordinate ONU with its own local time, calculates the RTT with each subordinate ONU, and holds each calculated RTT. Each GST set in the received frame is updated based on the local time and the RTT. As a result, even if the communication path higher than the RN is switched, the RTT held by the RN is not affected, so that the upstream burst control of the PON is facilitated and the communication is restored when the communication path is switched. The time required until this can be shortened compared to the conventional method.

  As described above, the optical communication system according to the present invention is suitable for an optical communication system having a configuration including a plurality of one-to-many connected networks and using upstream communication as time division multiplex communication.

1 OLT
2, 2 ₁ , 2 ₂ , 2 _n ring nodes 3, 3 ₁₁ , 3 _1p , 3 ₂₁ , 3 _2q , 3 _n1 , 3 _nr remote nodes (RN)
4 Optical coupler 5 ONU
5 ₁ 1G-ONU
5 ₁₀ 10G-ONU
5 ₁₀₁ 10G ・ 1G-ONU
11 OLT-IF
12 Monitoring control unit 13, 21 ROADM unit 22 OTN-XC unit 31, 115 OptTRx unit 32, 112 OTN mapper / demapper unit 33 Burst transponder 111 IF unit 113 SER / DES unit 114 Upstream bandwidth allocation control unit 331 RTT management unit 332 DOB CDR unit 333 DOB Rx unit 334 Time stamp processing unit 335 1G / 10G OptTx unit 336 WDM coupler 337 Local timer

Claims (15)

A ring network is configured, signal light of a specific wavelength is extracted from the ring network and output to other nodes that do not configure the ring network, and signal light input from the other nodes is An optical communication system comprising a plurality of ring nodes that output to a ring network, a remote node that operates as the other node, and an ONU that is connected to the remote node via an optical coupler,
A part of the ring nodes operates as an OLT that executes a process of allocating an upstream band to the ONU;
The node operating as the OLT is:
The ONUs under the same remote node are treated as one ONU group, and the upstream bandwidth allocated to the own system is determined based on the upstream bandwidth allocation request amount from each ONU and the number of ONUs belonging to each ONU group. A first bandwidth allocation function distributed to each of the ONU groups;
A second bandwidth allocation function for redistributing the upstream bandwidth distributed by the first bandwidth allocation function for each ONU group to each ONU in the group based on an upstream bandwidth allocation request amount from each ONU;
An optical communication system comprising: 2. The optical communication system according to claim 1, wherein the first band allocation function is executed at a period that is the same as or longer than a band update period that is a period for updating an uplink band allocated to each ONU. In the first bandwidth allocation function,
In the case of (total sum of uplink bandwidth allocation requests from all ONUs) ≦ (uplink bandwidth allocated to own system),
2. The optical communication system according to claim 1, wherein the same amount of bandwidth as the sum of the upstream bandwidth request amount from each ONU for each ONU group is allocated to each ONU group. In the first bandwidth allocation function,
In the case of (total sum of uplink bandwidth allocation requests from all ONUs)> (uplink bandwidth allocated to own system),
BW upstream band assigned to the own system, i-th (i = 0,1, ...) the number of ONU belonging to the ONU group N _i of the number of the entire system ONU is N,
2. The optical communication system according to claim 1, wherein an upstream band is allocated to each ONU group such that “(bandwidth allocation amount for i-th ONU group) ≦ (BW × N _i / N)”. Let R _i be the sum of upstream bandwidth requests from ONUs belonging to the i-th ONU group.
If “R _i ≦ (B × N _i / N)” holds, the same amount of bandwidth as R _i is allocated to the corresponding ONU group,
5. The optical device according to claim 4, wherein when “R _i > (B × N _i / N)” is established, the same amount of bandwidth as B × N _i / N is allocated to the corresponding ONU group. Communications system. R _i is the sum of uplink bandwidth requests from ONUs belonging to the i-th ONU group, y _i (t) is the average value of past uplink bandwidth allocations for the i-th ONU group, and Δy _i (t) = If y _i (t) −B × N _i / N,
If “R _i ≦ (B × N _i / N)” holds, the same amount of bandwidth as R _i is allocated to the corresponding ONU group,
When “R _i > (B × N _i / N)” is satisfied, the same amount of bandwidth as x _i (t) represented by the following expression is allocated to the corresponding ONU group.

The optical communication system according to claim 4. If there is a remaining bandwidth at the time when the upstream bandwidth distribution to each ONU group by the first bandwidth allocation function is completed, the remaining bandwidth is additionally distributed to each ONU group, and the second distribution is performed after the additional distribution is completed. The optical communication system according to claim 1, wherein a bandwidth is allocated to the ONU by a bandwidth allocation function. The remote node comprises one or more transponders, and the ONU is connected to the transponders via the optical coupler;
In the second bandwidth allocation function,
In the process of assigning the upstream bandwidth to each ONU in each ONU group, first, the distributed upstream bandwidth is assigned to each transponder according to the number of connected ONUs, and then the bandwidth assigned to each transponder is: 2. The optical communication system according to claim 1, wherein the optical communication system is assigned to each ONU based on an upstream bandwidth request amount from each ONU. The remote node has a function of measuring an RTT indicating a transmission delay time between each ONU connected to the remote node, and when receiving Gate information indicating an uplink bandwidth allocation result for each ONU, The optical communication system according to claim 1, wherein the gate information is corrected based on its own local time and the measurement result of the RTT and then transferred to each ONU. The remote node is
GST is corrected so that uplink signals transmitted from each of the ONUs do not collide with each other including GST indicating transmission start time of each of the ONUs among the information included in the Gate information. The optical communication system according to claim 9. The remote node is
When measuring the RTT indicating the transmission delay time between each ONU accommodated and holding the measurement result, and receiving the gate information indicating the uplink band allocation result for each ONU, the Gate information Is corrected based on its own local time and the measurement result of the RTT and then transferred to each ONU,
The optical communication system according to claim 9, comprising one or more of the following. The transponder is
GST is corrected so that uplink signals transmitted from each of the ONUs do not collide with each other including GST indicating transmission start time of each of the ONUs among the information included in the Gate information. The optical communication system according to claim 11. The ring node operating as the OLT is
The optical communication according to claim 9, wherein after determining an uplink band to be assigned to each ONU, the gate information is generated without considering a transmission delay time between each ONU to which the uplink band is assigned. system. A ring network is configured, signal light of a specific wavelength is extracted from the ring network and output to other nodes that do not configure the ring network, and signal light input from the other nodes is A plurality of ring nodes that output to a ring network, a remote node that operates as the other node, and an ONU that is connected to the remote node via an optical coupler, and a part of the nodes in the ring node In an optical communication system that allocates an upstream band to the ONU, a communication device that operates as a ring node that executes the process of allocating the upstream band,
The ONUs under the same remote node are treated as one ONU group, and the upstream bandwidth allocated to the own system is determined based on the upstream bandwidth allocation request amount from each ONU and the number of ONUs belonging to each ONU group. A communication apparatus that distributes to each ONU group, and further redistributes the distributed upstream bandwidth to each ONU in the group based on an upstream bandwidth allocation request amount from each ONU. . A ring network is configured, signal light of a specific wavelength is extracted from the ring network and output to other nodes that do not configure the ring network, and signal light input from the other nodes is A plurality of ring nodes that output to a ring network, a remote node that operates as the other node, and an ONU that is connected to the remote node via an optical coupler, and a part of the nodes in the ring node Is a bandwidth control method in an optical communication system that allocates an upstream bandwidth to the ONU,
The ONUs under the same remote node are treated as one ONU group, and the upstream bandwidth allocated to the own system is determined based on the upstream bandwidth allocation request amount from each ONU and the number of ONUs belonging to each ONU group. A first bandwidth allocation step for distributing to each of the ONU groups;
A second bandwidth allocation step for redistributing the upstream bandwidth distributed in the first bandwidth allocation step to each ONU in the group based on an upstream bandwidth allocation request amount from each ONU for each of the ONU groups;
A bandwidth control method comprising:

Images