WO2018149102A1

WO2018149102A1 - Method and device for reducing transmission latency of high-priority data, and storage medium

Info

Publication number: WO2018149102A1
Application number: PCT/CN2017/097325
Authority: WO
Inventors: 孙月; 胡达; 钱晓东; 肖洁
Original assignee: 深圳市中兴微电子技术有限公司
Priority date: 2017-02-20
Filing date: 2017-08-14
Publication date: 2018-08-23
Also published as: CN108462649B; CN108462649A

Abstract

Provided in embodiments of the present invention are a method and device for reducing a transmission latency of high-priority data, and storage medium. The method comprises: storing, according to packet descriptor information obtained by parsing a packet, data of the packet correspondingly into a buffer; computing, according to current bandwidth information sent by an optical line terminal (OLT), a predetermined priority-based scheduling rule, and a buffer state corresponding to the bandwidth information, a buffer index of a current scheduling operation; and reading the data from the buffer corresponding to the buffer index, and transmitting the same to the OLT.

Description

Method and device for reducing high priority data transmission delay, storage medium

Cross-reference to related applications

The present application is based on a Chinese patent application filed on Jan. 20, 2017, the entire disclosure of which is hereby incorporated by reference.

Technical field

The present invention relates to the field of network communication technologies, and in particular, to a method and apparatus, and a storage medium for reducing a high priority data transmission delay in a congested state in an optical network unit (ONU).

Background technique

With the development of network application technology, more and more data traffic is transmitted in the network. In order to guarantee the quality of service of network applications, the guarantee of priority and delay in network equipment is very important.

For Optical Line Terminals (OLTs), bandwidth control is based on Static Bandwidth Allocation (SBA) and Dynamic Bandwidth Allocation (DBA) mechanisms. For example, Gigabit Passive Optical Network (GPON) devices are based on Transmission Containers (Transmission Containers). TCONT) is used to allocate bandwidth. The TCONT is a scheduling time slot of GPON time division multiplexing, in which data belonging to the time slot can be transmitted, Logical Link Identifier (LLID) A scheduling time slot for EPON time division multiplexing, the principle is the same as GPON. In the case of multiple TCONT/LLID, the bandwidth is polled between each TCONT/LLID. In order to ensure the line speed performance in this case, the traffic management (TM) egress module usually needs a larger cache. The space stores the entire data packet to ensure that the overhead of the MAC framing is reduced, the bandwidth utilization is improved, and the transmission of a large number of IDLE frames is prevented due to waiting for the whole packet.

It can be seen that the bandwidth allocation method of the OLT is based on TCONT\LLID, not In the priority queue, the internal scheduling of the TM egress module is mostly based on the priority queue. The intermediate cache stores the data after the internal scheduling of the TM egress module. For the end-to-end scheduling, the following problems may occur: First, the packets are sequentially written into the cache. Once the low-priority data (messages) is scheduled, in order to ensure that no error packets and out-of-orders occur, even then The arrival of data with high priority cannot be scheduled immediately, and it needs to wait and congestion occurs, as shown in Figure 1. On the other hand, the slot scheduling feature of the ONU causes the next scheduling of the message to wait for an interval above one slot. Therefore, the delay of high priority data will be amplified, affecting the quality of service, and the user experience is not good.

Summary of the invention

In order to solve the existing technical problems, embodiments of the present invention provide a method and apparatus, and a storage medium for reducing a high priority data transmission delay.

To achieve the above objective, the technical solution of the embodiment of the present invention is implemented as follows:

An embodiment of the present invention provides a method for reducing a high-priority data transmission delay, which is applied to an ONU, and the method includes:

Corresponding to storing the data in the packet in the cache according to the descriptor information of the packet parsed from the packet;

Calculating the cache index of the current schedule according to the current bandwidth information delivered by the OLT, the preset priority scheduling rule, and the state of the cache corresponding to the bandwidth information;

Data is read from a cache corresponding to the cache index and ultimately transmitted to the OLT.

In the above solution, when the data is read from the cache corresponding to the cache index, the method further includes:

Recording the current state of all the caches and the current scheduling information for subsequent data scheduling; the current scheduling information includes at least: the current scheduled data and the corresponding stored cache.

In the above solution, the storing, in the cache, the data in the packet according to the descriptor information includes:

Determining, according to the TCONT\LLID ID and the priority ID of the data in the descriptor information, the data in the packet, and the data of different priorities corresponding to the same TCONT\LLID ID are written into different caches; The higher the priority of the data, the larger the priority ID.

In the foregoing solution, the current bandwidth information delivered by the OLT, the preset priority scheduling rule, and the cached state corresponding to the bandwidth information, and the cache of the current scheduling are calculated according to an absolute priority (SP) scheduling method. Index, including:

Determining a cache of the current schedule according to the TCONT\LLID ID in the current bandwidth information and the cached index;

The highest priority data ID corresponding to the TCONT\LLID ID in the current bandwidth information is selected from the mapping table based on the mapping table of the TCONT\LLID ID and the data priority in the preset priority scheduling rule;

Determining, according to the descriptor information of the packet, a priority ID corresponding to the data ID;

When it is determined that the state of the cache corresponding to the priority ID is a non-empty state, determining that the cached index is the cached index of the current schedule.

In the above solution, the current bandwidth information delivered by the OLT, the preset priority scheduling rule, and the cached state corresponding to the bandwidth information, and the cached index of the current scheduling are calculated according to the SP scheduling method, including:

Determining, in the determined cache, a state in which the state is non-empty;

And determining, according to the preset priority scheduling rule, a cache index with a maximum priority ID based on the cache index corresponding to the non-empty cache as a cache index of the schedule.

The embodiment of the present invention further provides an apparatus for reducing a high-priority data transmission delay, which is applied to an ONU, and the apparatus includes:

The cache module is configured to store the data in the packet in the cache according to the descriptor information of the packet parsed from the packet; and the cache index of the current schedule based on the scheduling module transmits from the cache index Reading data in the corresponding cache and finally transmitting to the OLT;

The scheduling module is configured to calculate, according to the current bandwidth information delivered by the OLT, a preset priority scheduling rule, and a cached state corresponding to the bandwidth information, and calculate a cache index of the current scheduling according to the SP scheduling method, and send the cached index to the Cache module.

In the above solution, when the scheduling module reads data from a cache corresponding to the cache index,

The scheduling module is further configured to record the current state of all the caches and the current scheduling information for subsequent data scheduling. The current scheduling information includes at least the current scheduled data and the corresponding stored cache.

In the above solution, the cache module includes:

The cache unit is configured to store the data in the packet in the cache according to the descriptor information of the packet parsed from the packet;

The reading unit is configured to read data from the cache corresponding to the cache index based on the currently scheduled cache index transmitted by the scheduling module, and finally transmit the data to the OLT.

In the above solution, the cache unit includes:

a descriptor cache unit configured to cache the descriptor information;

The data buffering unit is configured to store, according to the TCONT\LLID ID and the priority ID of the data in the descriptor information stored by the descriptor buffer unit, the data in the packet, corresponding to the same TCONT\LLID ID. Data of different priorities are written into different caches; wherein the higher the priority of the data, the larger the priority ID.

In the above solution, the scheduling module includes:

The first cache determining unit is configured to determine, according to the TCONT\LLID ID in the current bandwidth information and the cached index, the cache of the current scheduling;

The mapping unit is configured to: according to the mapping table of the TCONT\LLID ID and the data priority in the preset priority scheduling rule, select the highest priority data ID corresponding to the TCONT\LLID ID in the current bandwidth information from the mapping table;

a first index determining unit, configured to determine a priority ID corresponding to the data ID according to the descriptor information of the packet; and determine, when the state of the cache corresponding to the priority ID is a non-empty state, determining the cached The index is the cached index of this schedule.

In the above solution, the scheduling module includes:

a second cache determining unit, configured to determine a cache of the current scheduling according to the TCONT\LLID ID in the current bandwidth information and the cached index;

a state determining unit, configured to determine a cache in the determined cache that is non-empty;

The second index determining unit is configured to determine, according to the preset priority scheduling rule, that the cache index with the highest priority ID is the cache index of the current scheduling based on the cache index corresponding to the non-empty cache.

Embodiments of the present invention also provide a computer storage medium storing a computer program configured to perform the above method of reducing a high priority data transmission delay.

The method and device for reducing the high-priority data transmission delay and the storage medium provided by the embodiment of the present invention, according to the descriptor information of the packet parsed from the packet, correspondingly storing the data in the packet in the cache; Calculating the cache index of the current schedule according to the current bandwidth information delivered by the OLT, the preset priority scheduling rule, and the state of the cache corresponding to the bandwidth information; and reading data from the cache corresponding to the cache index, And finally transmitted to the OLT. In the embodiment of the present invention, the data with the highest priority is preferentially transmitted based on the preset priority scheduling rule, and the packet with the lower priority is not passed or passed, thereby effectively reducing the delay of the high priority data and improving the user. At the same time, because the delay of high-priority data is significantly reduced (after the measured, the delay is reduced from 8ms to about 500us), the cache resource is reduced accordingly (no need to be large) Cache stores congested data), reducing system operating costs.

DRAWINGS

1 is a schematic diagram of a high priority data generation delay in the prior art;

2 is a schematic structural diagram of an apparatus for reducing a high priority data transmission delay in a congestion state in an ONU according to an embodiment of the present invention;

3 is a schematic flowchart of a method for reducing a high priority data transmission delay in a congestion state in an ONU according to an embodiment of the present invention;

4 is a schematic structural diagram 1 of the egress data cache management module in the embodiment of the present invention;

FIG. 5 is a schematic structural diagram of an egress data buffer scheduling module in an embodiment of the present invention;

6 is a schematic structural diagram 2 of the egress data cache management module in the embodiment of the present invention;

FIG. 7 is a flowchart of a method for reducing a high priority data transmission delay in a congestion state in an ONU according to an embodiment of the present disclosure;

FIG. 8 is a schematic structural diagram of an apparatus for reducing a high priority data transmission delay in a congestion state in an ONU according to an embodiment of the present disclosure;

FIG. 9 is a schematic structural diagram of a cache module according to an embodiment of the present invention;

FIG. 10 is a schematic structural diagram of a cache unit according to an embodiment of the present invention;

FIG. 11 is a schematic structural diagram of a scheduling module according to an embodiment of the present invention;

FIG. 12 is a schematic structural diagram of a scheduling module according to another embodiment of the present invention.

detailed description

The invention is described in detail below with reference to specific embodiments.

Embodiments of the present invention provide a method and apparatus for reducing a high-priority data transmission delay. Here, high-priority data refers to high-priority data in a congestion state in an ONU.

An embodiment of the present invention provides a method for reducing a high priority data transmission delay. As shown in FIG. 7, the method includes:

Step 701: Corresponding to storing the data in the packet in the cache according to the descriptor information of the packet parsed from the packet;

Step 702: Calculate the cache index of the current scheduling according to the current bandwidth information delivered by the OLT, the preset priority scheduling rule, and the cache status corresponding to the bandwidth information.

Step 703: Read data from the cache corresponding to the cache index, and finally transmit the data to the OLT.

Here, the preset priority scheduling rule may be: a priority scheduling rule that is pre-configured according to the needs of the user service, that is, the data that needs to be preferentially scheduled may be arbitrarily set; or the preset priority scheduling rule is : determining the highest priority data from the priority of storing data in the cache as the priority scheduled data. The descriptor information may include: data priority obtained from parsing in the message, bandwidth information (such as TCONT\LLID ID, etc.), data ID, and the like.

In the embodiment of the present invention, the data with the highest priority is preferentially transmitted based on the preset priority scheduling rule, and the packet with the lower priority is not passed or passed, thereby effectively reducing the delay of the high priority data and improving the user. At the same time, because the delay of high-priority data is significantly reduced, the consumption of cache resources is reduced correspondingly (no need to cache data in a large cache), and the system operation cost is reduced.

In the embodiment of the present invention, when the data is read from the cache corresponding to the cache index, the method further includes:

In the embodiment of the present invention, the storing, in the cache, the data in the packet according to the descriptor information includes:

According to the TCONT\LLID ID in the descriptor information and the priority ID of the data The data of the different priorities of the same TCONT\LLID ID is written into different caches, wherein the higher the priority of the data, the larger the priority ID.

In practical application, the descriptor information may be stored separately from the data, the descriptor information is stored in the descriptor cache, and the data is stored in the data cache. For the same data (queue), the data is cached and described. The cache index of the cache is the same.

In one embodiment, the current bandwidth information delivered by the OLT, the preset priority scheduling rule, and the cached state corresponding to the bandwidth information, and the cached index of the current scheduling are calculated according to the SP scheduling method, including:

In another embodiment, the current bandwidth information delivered by the OLT, the preset priority scheduling rule, and the cached state corresponding to the bandwidth information, and the cached index of the current scheduling are calculated according to the SP scheduling method, including :

Determining, in the determined cache, a state in which the state is non-empty;

It can be seen that the embodiment of the present invention can flexibly set the data to be preferentially transmitted according to the service requirement, and can also preferentially send the high priority data according to the priority of the data in the packet, thereby implementing the method flexible and effectively reducing the data delay, especially in the case of congestion. The delay time of high priority data ensures the quality of service and improves the user experience.

Embodiments of the present invention also provide a storage medium having stored therein executable instructions that, when executed, are configured to perform any of the steps of reducing a high priority data transmission delay .

Optionally, in the embodiment, the foregoing storage medium may include, but is not limited to, a USB flash drive, a Read-Only Memory (ROM), and a Random Access Memory (RAM). A variety of media that can store program code, such as a hard disk, a disk, or an optical disk.

The embodiment of the present invention further provides a device for reducing the high-priority data transmission delay, which is used to implement the foregoing embodiments and preferred embodiments, and has not been described again. As used below, the term "module" "unit" may implement a combination of software and/or hardware of a predetermined function. As shown in Figure 8, the device includes:

The cache module 81 is configured to: according to the descriptor information of the packet parsed from the packet, store the data in the packet in the cache; and use the cache index of the current schedule transmitted by the scheduling module from the cache Reading data in the cache corresponding to the index, and finally transmitting to the OLT;

The scheduling module 82 is configured to calculate the current cached index according to the current bandwidth information delivered by the OLT, the preset priority scheduling rule, and the cached state corresponding to the bandwidth information, and send the cached index of the current scheduling according to the SP scheduling method. The cache module.

Here, the preset priority scheduling rule may be: a priority scheduling rule that is pre-configured according to the needs of the user service, that is, the data that needs to be preferentially scheduled may be arbitrarily set; or the preset priority scheduling rule is : determining the highest priority data from the priority of storing data in the cache as the priority scheduled data. The descriptor information may include: solving from a message Analysis of the resulting data priority, bandwidth information (such as TCONT\LLID ID, etc.), data ID, and so on.

In the embodiment of the present invention, when the scheduling module 82 reads data from the cache corresponding to the cache index,

The scheduling module 82 is further configured to record the current state of all the caches and the current scheduling information for subsequent data scheduling; the current scheduling information includes at least: the current scheduled data and the corresponding stored cache. .

In the embodiment of the present invention, as shown in FIG. 9, the cache module 81 includes:

The buffer unit 811 is configured to store, according to the descriptor information of the packet parsed from the packet, the data in the packet in the cache;

The reading unit 812 is configured to read data from the cache corresponding to the cache index based on the currently scheduled cache index transmitted by the scheduling module, and finally transmit the data to the OLT.

In the embodiment of the present invention, as shown in FIG. 10, the cache unit 811 includes:

a descriptor buffer unit 8111, configured to cache the descriptor information;

The data buffering unit 8112 is configured to store the data in the packet according to the TCONT\LLID ID and the priority ID of the data in the descriptor information stored by the descriptor buffer unit, and correspond to the same TCONT\LLID ID. Different priority data is written into different caches; wherein the higher the priority of the data, the larger the priority ID.

In one embodiment, as described in FIG. 11, the scheduling module 82 can include:

The first cache determining unit 821 is configured to determine, according to the TCONT\LLID ID in the current bandwidth information and the cached index, the cache of the current scheduling;

The mapping unit 822 is configured to: according to the mapping table of the TCONT\LLID ID and the data priority in the preset priority scheduling rule, select the highest priority data ID corresponding to the TCONT\LLID ID in the current bandwidth information from the mapping table;

The first index determining unit 823 is configured to determine a priority ID corresponding to the data ID according to the descriptor information of the packet, and determine the cache when determining that the cached state corresponding to the priority ID is a non-empty state. The index of this is the cached index of this schedule.

In another embodiment, as illustrated in FIG. 12, the scheduling module 82 can include:

The second cache determining unit 824 is configured to determine, according to the TCONT\LLID ID in the current bandwidth information and the cached index, the cache of the current scheduling;

a state determining unit 825, configured to determine that the cached state in the cache is non-empty;

The second index determining unit 826 is configured to determine, according to the preset priority scheduling rule, that the cache index with the highest priority ID is the cache index of the current scheduling based on the cache index corresponding to the non-empty cache.

The present invention will be described below in conjunction with specific embodiments.

This embodiment provides another apparatus for reducing the high priority data transmission delay in a congestion state in an ONU, including the following modules:

The internal traffic management module A, the data cache management module B (corresponding to the cache module 81), the data scheduling management module C (corresponding to the scheduling module 82), and the bandwidth information management module D (the four modules are referred to as A , B, C, D); among them,

The internal traffic management module A is configured to generate descriptor information of the internal message of the ONU. According to the cache management module B, the data cache management module B is configured to store the corresponding data content according to the descriptor information, and sequentially write the cache; the data scheduling management module C is configured to receive the bandwidth information sent by the bandwidth information management module D, and simultaneously According to the software configuration rule (preset priority scheduling rule), SP scheduling is performed, and the current scheduling cache index is calculated and provided to the data cache management module B; the data cache management module B is based on the information (the current scheduling cache index) The data is read from the cache and provided to the bandwidth information management module D to complete a complete scheduling process.

The scheduling method based on the above apparatus may include the following steps:

The first step: A is responsible for the internal traffic management of the ONU, including QOS functions such as speed limit, shaping, scheduling, etc., and finally provides queue descriptor information of different priorities to B;

In order to enable the ONU to reach the line rate, the pre-read descriptor needs to be used for buffering. That is, if the OLT has no bandwidth to send, A pre-schedules a certain number of data packets to be written to B, and keeps the buffer in B. The continuity of the readout, the generation of the IDLE frame is simulated. At this time, the index of the internal cache of B is TCONT\LLID ID+priority ID, that is, the packets of different priorities corresponding to the same TCONT\LLID ID are written into different caches. space.

The second step: B reports the status of D to the current cache, indicating which TCONT\LLID ID corresponding buffers have valid data (non-empty), which can be read. D is based on the bandwidth information delivered by the current OLT and the buffer status reported by B. , notify C of the TCONT\LLID ID and other information that can be scheduled this time;

The third step: C according to the TCONT\LLID ID, the pre-configured TCONT\LLID ID and the data priority mapping relationship, using the SP method combined with the current TCONT\LLID ID available for reading the queue status (cache status), calculate The highest priority queue ID is scheduled to be transmitted to B as a cached index, and B reads the corresponding data to D according to the scheduling information index to complete a scheduling process.

Among them, C also records the status and current scheduling information of all priority queue caches under all TCONT\LLID IDs to maintain the integrity of packet scheduling to prevent mis-packets and out-of-order situations.

2 is a schematic structural diagram of an apparatus for reducing a high priority data transmission delay in a congestion state in an ONU, as shown in FIG. 2, the apparatus includes: a descriptor scheduling module 201; an exit descriptor cache 202; The data cache management module 203, the egress data cache scheduling module 204, the bandwidth information processing module 205; wherein the descriptor scheduling module 201 corresponds to the above A, the egress descriptor cache 202 and the egress data cache management module 203 form a B, and the egress data cache The scheduling module 204 corresponds to C, and the bandwidth information processing module 205 corresponds to D;

The descriptor scheduling module 201 provides a descriptor that can be scheduled to the egress descriptor cache 202 for buffering the descriptor;

The bandwidth information processing module 205 provides the currently scheduled TCONT\LLID information (TCONT\LLID ID) to the egress data cache scheduling module 204;

The classification configuration module 207 provides pre-configured priority information to the egress data cache scheduling module 204;

The egress data cache scheduling module 204, along with the cache size and backpressure information (ie, the cache state) provided by the egress data cache management module 203, and the descriptor cache size and corresponding backpressure information provided by the egress descriptor cache 202 ( That is, the cache state), combined with the TCONT\LLID information and the pre-configured priority information, internally outputs the currently scheduled descriptor address to the egress descriptor cache 202 and the egress data cache management module 203 in an absolute priority manner, and from the exit. The descriptor cache 202 and the egress data cache management module 203 read out the corresponding descriptors and data and send them to the bandwidth information processing module 205.

FIG. 3 is a schematic flowchart of a method for reducing a high-priority data transmission delay in a congestion state in an ONU according to the embodiment of the present invention. As shown in FIG. 3, the method includes:

S301: The packet enters the ONU.

S302: Determine, according to the empty indication signal of the egress descriptor cache 202, whether the descriptor can be written, and determine whether there is a descriptor that can be pre-scheduled in the packet (there is a non-empty queue in the packet), if there is a non-empty queue , proceed to S303; otherwise, continue to judge (execute S302);

S303: Write the descriptor of the queue that can be pre-scheduled to the corresponding egress descriptor cache 202. The index of the cache is: TCONT\LLID ID+PRI ID, that is, the priorities corresponding to different TCONT\LLID IDs are stored separately. Look up the table to indicate which one of the TCONT\LLID IDs corresponds to which priority queue. The maximum number of TCONTs is 32, and the maximum number of priority queues per TCONT is 8. The cached resources can be flexibly configured. You can flexibly choose several fixed mapping relationships, or you can configure them by upper layer software.

S304: According to the current bandwidth information, that is, the TCONT\LLID ID that can be scheduled, the current egress descriptor buffer corresponding to the status information of all priority queues under the TCONT\LLID ID, the currently saved TCONT\LLID ID and corresponding The priority information, and the storage status information of the data cache (exit data cache management module) corresponding to the currently configurable TCONT\LLID ID, performs primary SP scheduling to determine the TCONT\LLID ID and corresponding priority of the current schedule. Level (PRI) ID;

ONU is based on fragment scheduling in GPON mode. That is, different fragments can be scheduled under different TCONTs within a period of time. It is not necessary to read the entire packet. In order to maintain the integrity and continuity of the packet, it is necessary to maintain a current The scheduling status information of all queues. If a priority packet has been scheduled in the current TCONT, packets of other priorities cannot be scheduled to prevent out-of-order. At this time, if the currently saved TCONT\LLID ID+PRI ID is obtained, S307 is performed; then, the remaining messages or new messages corresponding to the TCONT\LLID ID+PRI ID are dispatched through S305, and S306 is continued to be executed. Scheduling use.

S305: Write message data to the subsequent module (exit data cache management module 203) according to the descriptor information, and perform fragmentation processing;

S306: The message is stored in the corresponding egress buffer (FIFO, corresponding to the egress data cache management module 203) under the current TCONT\LLID+PRI ID, and the FIFO size is configurable, and the configuration of the maximum fragment length is considered at the same time, and the cache index is :TCONT\LLID ID+PRI ID, that is, the priority queues corresponding to different TCONT\LLID IDs are stored separately, and this needs to maintain a current slowdown. Store the mapping table corresponding to TCONT\LLID ID+PRI ID for judgment in S304; also convert all priority queue states under each TCONT\LLID ID to TCONT\LLID state (ie: put all All the priority queue states under the TCONT\LLID ID are aggregated and reflected in the form of the TCONT\LLID ID total schedulable state, and fed back to the passive optical network (PON) MAC;

S307: Read a data packet in the corresponding FIFO according to the TCONT\LLID ID;

At this time, it is determined by which the priority message under the TCONT\LLID ID is processed by the SP scheduling in the above S304. For the PON MAC, the interface timing is unchanged and is processed by the ONU internally.

S308: The read data packet is transmitted to the OLT, and the MAC module extracts corresponding bandwidth information to the egress data cache scheduling module.

Each module in Fig. 2 will be described in detail below.

The classification configuration module 207 provides a classification configuration corresponding to each descriptor. The configuration can support software configuration or automatic hardware identification. If the software configuration is valid, the software can specify which channel cache priority under which TCONT/LLID ID. The highest, can be arbitrarily specified. Each channel buffer is represented by 1 bit, and up to 32x8=256bit registers are used to implement the classification configuration corresponding to all caches; hardware can also be used to automatically identify the information carried in the message descriptor generated by the descriptor scheduling module 201, because TM When parsing the packet header, the descriptor scheduling module 201 obtains the difference service code point (dscp) value or related priority information corresponding to the packet, and can be used by the subsequent module for scheduling. In this way, the classification configuration module 207 can flexibly provide the priority information corresponding to each path cache, and flexibly perform scheduling.

A structure of the egress data cache management module 203 is shown in FIG. 4. The 401 is a minimum cache division basic unit, and includes an address generation module 403 (ADDR_GEN) configured to calculate an address for each update; a cache control module. 404 (FIFO_CTRL), configured to generate an empty full signal according to read and write enable; the division of the address generation module 403 is determined by cfg_mode Address and end address, and cfg_mode is the flexible configuration mode mentioned in the classification configuration module 207; followed by a dual port 1 write 1 read RAM 402 for storing data, using different 401 modules to generate different RAM read and write addresses , to achieve cache sharing and dynamic segmentation.

In the embodiment of the present invention, the implementation manners of the egress descriptor cache (202) and the egress data cache management module (203) are based on FIFO, and the principle of partitioning of the cache is based on TCONT\LLID ID+pri ID, that is, each TCONT\ The FIFO IDs of all pri IDs corresponding to the LLID ID are grouped into one group, and each pri ID corresponds to one FIFO. To save resources, in the case where the number of buffers remains unchanged, the number of pris under each TCONT\LLID can be flexibly allocated. According to the number of TCONT\LLID and PON mode, these caches are dynamically divided. In GPON mode, the maximum number of TCONTs is 32. In EPON mode, the maximum number of LLIDs is 8. In this case, if the number of caches is a, then The number of pris supported in GPON mode can vary according to the number of different TCONT configurations. The number of FIFOs per TCONT is a/t. For the same number of buffers, the number of FIFOs under each LLID is a/l. l is the number of LLIDs, that is, all TCONT and LLID shared caches to maximize cache utilization. Here, the design of pri can choose the supported pri according to the number of caches, TCONT and LLID. Number, can be flexibly configured {8, 4} way, this embodiment can be divided into the cache software to write configuration representation is achieved in that the dynamic partitioning logic to the buffer size according to the configuration mode.

The structure of the egress data cache scheduling module 204 is shown in FIG. 5. The module is composed of a mapping module 501 and a scheduling module 502. In order to consider the scalability and compatibility, it is assumed that the N-way status information corresponds to the N-way cache. According to the above cache division, each bit of the N-way information represents a pri queue state under the TCONT\LLID. In order to support the maximum M pri queues, according to the configuration, different mapping relationships are selected, and the mapping module 501 sets the N-channel state. The information is mapped into N×M road information, ensuring that each TCONT\LLID has M-path pri state information, so that the following scheduling module can use the same module structure to select, and can be flexibly extended according to the configuration, thereby obtaining N*M path cache status information, N*M way entry descriptor Status information, current scheduling status information of the N*M path, and phase-by-phase and after, finally obtain the entry information of the participating scheduling module 502, and then the principle of scheduling each M-way entry information is based on the SP scheduling mode (ie, the highest priority queue) The first scheduling is until there is no queue data of the priority at the current moment, and then the next highest priority queue is scheduled, and so on, and the N-way sub-scheduler outputs the final scheduling information of the N-way, and then enters the MUX module, according to the PON MAC input. The current scheduling read unit determines which routing information is output to the cache device and reads the data.

In addition, the embodiment of the present disclosure provides a structure diagram of another egress data cache management module 203. The structure in FIG. 4 is based on a FIFO architecture. The implementation of the cache structure based on the linked list management in this embodiment is as shown in FIG. a linked list control module 601, an address management module 602, a linked list pointer storage module 603, and a data content storage module 604, wherein

601 receives the descriptor of the written message and raises the free address request to 602, and allocates a free address to 601 according to the idle condition of the internal address, and saves the current address and the next hop address to 603, 601 according to the address returned by 602, the current packet content is written to 604 for saving; when 601 receives the read message descriptor, the address request corresponding to the message is raised to 602, 602 according to the content of the head pointer, In 603, the saved next hop address is read, and after returning to 601, 601 to obtain the address, the message is read from 604.

Those skilled in the art will appreciate that embodiments of the present invention can be provided as a method, system, or computer program product. Accordingly, the present invention can take the form of a hardware embodiment, a software embodiment, or a combination of software and hardware. Moreover, the invention can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) including computer usable program code.

The present invention has been described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (system), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or FIG. These computer program instructions can be provided to a general purpose computer, a special purpose computer, An processor of an embedded processor or other programmable data processing device to generate a machine such that instructions executed by a processor of a computer or other programmable data processing device are generated for implementation in a flow or a flow of flowcharts and/or Or a block diagram of a device in a box or a function specified in a plurality of boxes.

The computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture comprising the instruction device. The apparatus implements the functions specified in one or more blocks of a flow or a flow and/or block diagram of the flowchart.

These computer program instructions can also be loaded onto a computer or other programmable data processing device such that a series of operational steps are performed on a computer or other programmable device to produce computer-implemented processing for execution on a computer or other programmable device. The instructions provide steps for implementing the functions specified in one or more of the flow or in a block or blocks of a flow diagram.

The above is only the preferred embodiment of the present invention and is not intended to limit the scope of the present invention.

Industrial applicability

The technical solution of the embodiment of the present invention can preferentially transmit data with a high priority based on a preset priority scheduling rule, and the packet with a lower priority does not pass or pass the second priority, thereby effectively reducing the delay of the high priority data. The user experience is improved. At the same time, the delay of high-priority data is significantly reduced (after the measurement, the delay is reduced from 8ms to about 500us), which reduces the occupation of cache resources (no need for large cache storage congestion). Data), reducing system operating costs.

Claims

A method for reducing a high-priority data transmission delay is applied to an optical network unit ONU, and the method includes:

Corresponding to storing the data in the packet in the cache according to the descriptor information of the packet parsed from the packet;

Calculating the cache index of the current schedule according to the current bandwidth information delivered by the optical line terminal OLT, the preset priority scheduling rule, and the cache status corresponding to the bandwidth information;

Data is read from a cache corresponding to the cache index and ultimately transmitted to the OLT.
The method of claim 1, wherein when the data is read from a cache corresponding to the cache index, the method further comprises:

Recording the current state of all the caches and the current scheduling information for subsequent data scheduling; the current scheduling information includes at least: the current scheduled data and the corresponding stored cache.
The method according to claim 1 or 2, wherein the storing the data in the packet in the cache according to the descriptor information comprises:

And according to the transmission container TCONT\logical link identifier LLID ID and the priority ID of the data in the descriptor information as an index, storing data in the packet, and writing data of different priorities corresponding to the same TCONT\LLID ID Entering into different caches; wherein the higher the priority of the data, the larger the priority ID.
The method according to claim 3, wherein the current bandwidth information delivered by the optical line terminal OLT, a preset priority scheduling rule, and a cached state corresponding to the bandwidth information are based on an absolute priority SP. The scheduling method calculates the cache index of the current schedule, including:

Determining a cache of the current schedule according to the TCONT\LLID ID in the current bandwidth information and the cached index;

The highest priority data ID corresponding to the TCONT\LLID ID in the current bandwidth information is selected from the mapping table based on the mapping table of the TCONT\LLID ID and the data priority in the preset priority scheduling rule;

Determining, according to the descriptor information of the packet, a priority ID corresponding to the data ID;

When it is determined that the state of the cache corresponding to the priority ID is a non-empty state, determining that the cached index is the cached index of the current schedule.
The method according to claim 3, wherein the current bandwidth information delivered by the optical line terminal OLT, a preset priority scheduling rule, and a cached state corresponding to the bandwidth information are based on an absolute priority SP. The scheduling method calculates the cache index of the current schedule, including:

Determining a cache of the current schedule according to the TCONT\LLID ID in the current bandwidth information and the cached index;

Determining, in the determined cache, a state in which the state is non-empty;

And determining, according to the preset priority scheduling rule, a cache index with a maximum priority ID based on the cache index corresponding to the non-empty cache as a cache index of the schedule.
A device for reducing a high-priority data transmission delay is applied to an optical network unit ONU, the device comprising:

The cache module is configured to store the data in the packet in the cache according to the descriptor information of the packet parsed from the packet; and the cache index of the current schedule based on the scheduling module transmits from the cache index Reading data in the corresponding cache and finally transmitting to the OLT;

The scheduling module is configured to calculate the current cache according to the current bandwidth information delivered by the optical line terminal OLT, the preset priority scheduling rule, and the cache status corresponding to the bandwidth information, and based on the absolute priority SP scheduling method. Indexed and sent to the cache module.
The apparatus of claim 6 wherein said scheduling module is from said cache When reading data in the cache corresponding to the index,

The scheduling module is further configured to record the current state of all the caches and the current scheduling information for subsequent data scheduling. The current scheduling information includes at least the current scheduled data and the corresponding stored cache.
The apparatus according to claim 6 or 7, wherein the cache module comprises:

The cache unit is configured to store the data in the packet in the cache according to the descriptor information of the packet parsed from the packet;

The reading unit is configured to read data from the cache corresponding to the cache index based on the currently scheduled cache index transmitted by the scheduling module, and finally transmit the data to the OLT.
The apparatus of claim 8, wherein the cache unit comprises:

a descriptor cache unit configured to cache the descriptor information;

a data buffering unit configured to store, according to the transport container TCONT\logical link identifier LLID ID and the priority ID of the data in the descriptor information stored by the descriptor cache unit, the data in the packet, the same The data of different priorities corresponding to the TCONT\LLID ID is written into different caches; wherein the higher the priority of the data, the larger the priority ID.
The apparatus of claim 9, wherein the scheduling module comprises:

The first cache determining unit is configured to determine, according to the TCONT\LLID ID in the current bandwidth information and the cached index, the cache of the current scheduling;

The mapping unit is configured to: according to the mapping table of the TCONT\LLID ID and the data priority in the preset priority scheduling rule, select the highest priority data ID corresponding to the TCONT\LLID ID in the current bandwidth information from the mapping table;

a first index determining unit, configured to determine a priority ID corresponding to the data ID according to the descriptor information of the packet; and determine, when the state of the cache corresponding to the priority ID is a non-empty state, determining the cached The index is the cached index of this schedule.
The apparatus of claim 9, wherein the scheduling module comprises:

a second cache determining unit, configured to determine a cache of the current scheduling according to the TCONT\LLID ID in the current bandwidth information and the cached index;

a state determining unit, configured to determine a cache in the determined cache that is non-empty;

The second index determining unit is configured to determine, according to the preset priority scheduling rule, that the cache index with the highest priority ID is the cache index of the current scheduling based on the cache index corresponding to the non-empty cache.
A storage medium having stored therein computer executable instructions configured to perform the method of reducing high priority data transmission delays of any of claims 1-5.