A method and system for optimizing network-on-chip transmission based on data packet monitoring

By setting up monitoring points and data center modules in the on-chip network, recording and congestion analysis of data packet location information, and dynamically adjusting the packet transmission path, solving the problem of low transmission efficiency caused by congestion from routing nodes in the existing technology, and achieving more efficient network utilization and resource management.

CN119766652APending Publication Date: 2025-04-04SHANGHAI FANGYI WANQIANG MICROELECTRONICS CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411753861.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-02
Publication Date
2025-04-04

AI Technical Summary

Technical Problem

In the existing on-chip network design, congestion in routing nodes leads to increased packet transmission delays and decreased network utilization, and dynamic routing algorithms cannot effectively avoid network hotspots, resulting in the failure to achieve global optimal path selection.

Method used

By setting monitoring points on router nodes and links, data packet location information is recorded and updated, and this information is reported to the data center module for congestion analysis. The data center module broadcasts the congestion analysis results to the router node, so that it can dynamically adjust the transmission path of the data packets to avoid hotspot paths.

Benefits of technology

It realizes a global view of on-chip network congestion, dynamically adjusts packet routing paths, optimizes network utilization, improves transmission efficiency, and reduces resource waste.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119766652A_ABST
    Figure CN119766652A_ABST
Patent Text Reader

Abstract

The present invention provides a method and system for optimizing network-on-chip transmission based on data packet monitoring. In step A1, before transmitting a data packet, a router node records the location information carried by the data packet, and updates the location information carried by the data packet according to the current location information; in step A2, the router node reports the recorded location information to a data center module; in step A3, the data center module performs congestion analysis on the location information reported by all router nodes to obtain a congestion analysis result; in step A4, the data center module broadcasts the congestion analysis result to a first type of router node; in step A5, the first type of router node performs transmission path planning for a data packet newly entering the network-on-chip according to the congestion analysis result. The network utilization is optimized, the transmission efficiency is improved, and the waste of resources is reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of network-on-chip technology, and in particular to a method and system for optimizing network-on-chip transmission based on data packet monitoring. Background Art

[0002] Network on Chip (NoC) is a communication architecture implemented within a single chip to connect multiple functional modules (such as processor cores, memory, I / O interfaces, etc.). NoC is designed to solve the scalability, bandwidth and latency issues encountered by traditional bus and crossbar switch interconnects in high-performance, multi-core systems.

[0003] In existing network-on-chip (NoC) designs, static or dynamic routing algorithms are usually used to achieve data packet transmission in the network. However, with the increase in the number of on-chip processors, the congestion problem of routing nodes inside the NoC becomes more and more serious, resulting in increased transmission delays for data packets and decreased network utilization. Although conventional dynamic routing algorithms can select paths based on the current network conditions, they often rely on local congestion information and are insufficient to obtain global network traffic conditions. Data packets can only select routing nodes based on local information and cannot effectively avoid network hotspots (congested areas), resulting in the inability to achieve global optimal path selection. In addition, the adjustment speed of the data packet transmission path is slow and it is difficult to adapt to rapidly changing network traffic. It can be seen that the current routing algorithm cannot effectively improve the overall network utilization, resulting in a waste of network resources. Summary of the invention

[0004] Based on the above problems, the present invention provides a method and system for optimizing network-on-chip transmission based on data packet monitoring, aiming to solve the technical problems of low transmission efficiency in the prior art.

[0005] A method for optimizing network-on-chip transmission based on data packet monitoring, wherein the network-on-chip includes a plurality of router nodes, including:

[0006] Step A1, before transmitting a data packet, the router node records the location information carried by the data packet, and updates the location information carried by the data packet according to the current location information;

[0007] Step A2, the router node reports the recorded location information to the data center module;

[0008] Step A3, the data center module performs congestion analysis on the location information reported by all router nodes to obtain a congestion analysis result;

[0009] Step A4, the data center module broadcasts the congestion analysis result to the first type of router node, where the first type of router node is the first router node for the on-chip network to receive external data packets;

[0010] Step A5: the first type of router node plans a transmission path for a data packet newly entering the on-chip network according to the congestion analysis result.

[0011] Furthermore, a monitoring point is provided on the link formed between two adjacent router nodes;

[0012] In step A1, when the monitoring point detects a data packet, the location information carried by the data packet is recorded, and the location information carried by the data packet is updated according to the current location information;

[0013] In step A2, the monitoring point reports the recorded location information to the data center module;

[0014] In step A3, the data center module performs congestion analysis on the location information reported by each router node and each monitoring point to obtain a congestion analysis result.

[0015] Furthermore, in step A4, the data center module also broadcasts the congestion analysis result to the second type of router nodes, where the second type of router nodes are router nodes of the on-chip network other than the first type of router nodes;

[0016] In step A5, the second type of router node dynamically updates the transmission path of the data packets that have been transmitted in the on-chip network according to the congestion analysis result.

[0017] Furthermore, in step A5, planning the transmission path of the data packet includes planning the router nodes and links that need to be passed through.

[0018] Further, in step A1, the position information is written into the initial position of the payload portion of the data packet.

[0019] A network-on-chip transmission optimization system based on data packet monitoring, used to execute the aforementioned network-on-chip transmission optimization method based on data packet monitoring, comprising connected router nodes and data center modules:

[0020] The router node is used to: before transmitting the data packet, record the location information carried by the data packet, update the location information carried by the data packet according to the current location information, and report the recorded location information to the data center module;

[0021] The data center module is used to: perform congestion analysis on the location information reported by all router nodes to obtain a congestion analysis result, and broadcast the congestion analysis result to the first type of router nodes, where the first type of routing nodes is the first routing node for the on-chip network to receive external data packets;

[0022] The first type of router nodes is used to plan the transmission path for the data packets newly entering the on-chip network according to the congestion analysis results.

[0023] Further, a monitoring point is provided on the link formed between two adjacent router nodes, and the monitoring point is connected to the data center module;

[0024] The monitoring point is used to: when a data packet is monitored, record the location information carried by the data packet, update the location information carried by the data packet according to the current location information, and report the recorded location information to the data center module;

[0025] The data center module is used to perform congestion analysis on the location information reported by each router node and each monitoring point to obtain the congestion analysis results.

[0026] Furthermore, the data center module also broadcasts the congestion analysis result to the second type of router nodes, where the second type of router nodes are router nodes of the on-chip network other than the first type of router nodes;

[0027] The second type of router nodes dynamically updates the transmission paths of data packets that have been transmitted in the on-chip network according to the congestion analysis results.

[0028] Furthermore, planning the transmission path of the data packet includes planning the router nodes and links that need to be passed through.

[0029] Furthermore, the position information is written to the initial position of the payload portion of the data packet.

[0030] The beneficial technical effect of the present invention is that: by comprehensively analyzing the congestion situation of each routing node, the present invention grasps the global view of the congestion situation of the entire on-chip network, so that the routing path of the data packet can be dynamically adjusted to avoid hot spots, optimize network utilization, improve transmission efficiency, and reduce resource waste. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 A flowchart of a method for optimizing network-on-chip transmission based on data packet monitoring according to the present invention;

[0032] Figure 2 Provides an optimized schematic diagram of a transmission path for the present invention;

[0033] Figure 3A schematic diagram of the location information carried by the data packet of the present invention;

[0034] Figure 4 The module diagram of a network-on-chip transmission optimization system based on data packet monitoring of the present invention is shown in FIG. DETAILED DESCRIPTION

[0035] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0036] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0037] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, but they are not intended to limit the present invention.

[0038] See also Figure 1 The present invention provides a method for optimizing network-on-chip transmission based on data packet monitoring, wherein the network-on-chip includes a plurality of router nodes, including:

[0039] Step A1, before transmitting a data packet, the router node records the location information carried by the data packet, and updates the location information carried by the data packet according to the current location information;

[0040] Step A2, the router node reports the recorded location information to the data center module;

[0041] Step A3, the data center module performs congestion analysis on the location information reported by all router nodes to obtain a congestion analysis result;

[0042] Step A4, the data center module broadcasts the congestion analysis result to the first type of router node, where the first type of router node is the first router node for the on-chip network to receive external data packets;

[0043] Step A5: the first type of router node plans a transmission path for a data packet newly entering the on-chip network according to the congestion analysis result.

[0044] In step A1, for example, routing node 1 receives a data packet, and the location information of the data packet is routing node 1. Routing node 1 writes the location information into the data packet. Routing node 1 sends the data packet to routing node 2. Routing node 2 receives the data packet, records the location information carried by the data packet, i.e. routing node 1, and then writes the current location information, i.e. routing node 2, into the data packet, and updates the location information of routing node 1.

[0045] Specifically, the first type of routing node is the first routing node for the on-chip network to receive an external data packet, that is, the routing node for receiving a data packet newly entering the on-chip network.

[0046] The present invention comprehensively analyzes the congestion of each routing node and grasps the global view of the congestion of the entire on-chip network, so that the routing path of the data packet can be dynamically adjusted to avoid hot spots, optimize network utilization, improve transmission efficiency, and reduce resource waste.

[0047] Furthermore, a monitoring point is provided on the link formed between two adjacent router nodes;

[0048] In step A1, when the monitoring point detects a data packet, the location information carried by the data packet is recorded, and the location information carried by the data packet is updated according to the current location information;

[0049] In step A2, the monitoring point reports the recorded location information to the data center module;

[0050] In step A3, the data center module performs congestion analysis on the location information reported by each router node and each monitoring point to obtain a congestion analysis result.

[0051] The link formed between two adjacent router nodes is called an edge. Monitoring points are set on the edge to monitor the passage of data packets and report the recorded location information. The data center module can analyze which routing node is a congested node and which link is a congested link based on the location information reported by the routing nodes and monitoring points, thereby adjusting the transmission path of the data packet.

[0052] In step A1, for example, routing node 1 receives a data packet, and at this time the location information of the data packet is routing node 1, and routing node 1 writes the location information to the data packet. Routing node 1 sends the data packet to routing node 2, and link 1 formed between routing node 1 and routing node 2, that is, the monitoring point on the edge, detects that the data packet passes through, and the monitoring point records the location information carried by the data packet, that is, routing node 1, and then writes the current location information, that is, the current link 1, into the data packet to update the location information of the data packet. Routing node 2 receives the data packet, records the location information carried by the data packet, that is, link 1, and then writes the current location information, that is, routing node 2, into the data packet, and updates the location information of link 1. When the data packet is at the routing node, the location information is the routing node location, and when the data packet is at the link, the location information is the link location.

[0053] like Figure 2 As shown in the figure, the data packet needs to go from M1 to S1. The data center module analyzes that Node3 (routing node 3) and Edge5 (link 5) are congested. Then, the path of Edge1 (link 1) -> Node1 (routing node 1) -> Edge2 (link 2) can be selected as the transmission path.

[0054] Furthermore, in step A4, the data center module also broadcasts the congestion analysis result to the second type of router nodes, where the second type of router nodes are router nodes of the on-chip network other than the first type of router nodes;

[0055] In step A5, the second type of router node dynamically updates the transmission path of the data packets that have been transmitted in the on-chip network according to the congestion analysis result.

[0056] The second type of routing node is a non-first routing node on the on-chip network. Before receiving a data packet, the data packet has been transmitted by the previous routing node.

[0057] During the transmission of data packets, the routing nodes are dynamically adjusted according to the congestion status of the on-chip network, which improves the adaptability of the NoC network and reduces the delay of data packet transmission.

[0058] The first type of routing nodes and the second type of routing nodes constitute all routing nodes of the on-chip network. Each routing node in the on-chip network can grasp the congestion of the global network in real time, avoiding making undesirable routing decisions based on local information.

[0059] Furthermore, in step A5, planning the transmission path of the data packet includes planning the router nodes and links that need to be passed through.

[0060] Further, in step A1, the position information is written into the initial position of the payload portion of the data packet.

[0061] like Figure 3 As shown, the initial position of the payload portion of the data packet includes a first position portion and a second position portion, the first position portion is used to carry the routing node position information, namely, Node Position, and the second position portion is used to carry the link position information, namely, Edge Position. The link and routing node position information are stored separately, so that the monitoring point and routing node can distinguish the type of position information carried when reading, that is, whether it is the link position type or the routing node position type.

[0062] Specifically, the first position portion is recorded using 16 bits, and the second position portion is recorded using 16 bits.

[0063] Specifically, the first position portion is located earlier in the payload portion Payload than the second position portion and is closer to the header of the data packet.

[0064] Specifically, as an implementation method, the routing nodes and monitoring points report the recorded location information within a predetermined time period to the data center module, and the data center module performs statistical analysis on the recorded location information within the predetermined time period to obtain congestion information. Since the predetermined time period is very small, at the ms level, the lag of the analysis result is not strong, and it can be regarded as the congestion situation at the current time. For example, within 2ms, routing node 2 records the location information of the previous routing node 1 carried by 100 data packets, and routing node 4 records the location information of the previous routing node 3 carried by 10 data packets. This means that routing node 1 is more congested than routing node 3.

[0065] In a specific application scenario, for example, in a video processing SoC, the video frame is transmitted from the video input unit to the video processing unit through the NOC network. Based on the method of the present invention, the video frame data packet is recorded and carried on each routing node. The data center module collects information to analyze the congestion and broadcasts it to all routing nodes in the system. In this way, the subsequent incoming video frames can avoid crowded routing paths and optimize the transmission efficiency of the entire video data stream.

[0066] Through the present invention, the congestion perception ability of the NoC network is greatly improved, and the NoC network can be maximized, hot spot congestion areas can be avoided, and transmission conflicts can be reduced, thereby improving the overall data transmission efficiency. This global congestion information collection and broadcasting mechanism is similar to the vehicle navigation system in real life, which can help data packets in the NoC network select the optimal transmission path. In addition, by optimizing the transmission path of data packets and reducing congestion, energy consumption is also reduced.

[0067] See also Figure 4The present invention also provides an on-chip network transmission optimization system based on data packet monitoring, which is used to execute the above-mentioned on-chip network transmission optimization method based on data packet monitoring, comprising connected router nodes (1) and data center modules (2), wherein the second type of router nodes are router nodes of the on-chip network other than the first type of router nodes;

[0068] The router node (1) is used to: before transmitting a data packet, record the location information carried by the data packet, update the location information carried by the data packet according to the current location information, and report the recorded location information to the data center module (2);

[0069] The data center module (2) is used to: perform congestion analysis on the location information reported by all router nodes to obtain a congestion analysis result, and broadcast the congestion analysis result to the first type of router nodes;

[0070] The first type of router nodes is used to plan the transmission path for the data packets newly entering the on-chip network according to the congestion analysis results.

[0071] The data center module (2) may be a hardware module added separately for the on-chip network, or may be a software module set in a processor existing in the existing on-chip network.

[0072] For example, routing node 1 receives a data packet. At this time, the location information of the data packet is routing node 1. Routing node 1 writes the location information to the data packet. Routing node 1 sends the data packet to routing node 2. Routing node 2 receives the data packet and records the location information carried by the data packet, i.e. routing node 1. Then, the current location information, i.e. routing node 2, is written into the data packet, and the location information of routing node 1 is updated.

[0073] Specifically, the first type of routing node is the first routing node for the on-chip network to receive external data, that is, the routing node that receives a data packet newly entering the on-chip network.

[0074] The present invention comprehensively analyzes the congestion of each routing node and grasps the global view of the congestion of the entire on-chip network, so that the routing path of the data packet can be dynamically adjusted to avoid hot spots, optimize network utilization, improve transmission efficiency, and reduce resource waste.

[0075] Furthermore, a monitoring point is provided on the link formed between two adjacent router nodes (1), and the monitoring point (3) is connected to the data center module (2);

[0076] The monitoring point (3) is used to: when a data packet is detected, record the location information carried by the data packet, update the location information carried by the data packet according to the current location information, and report the recorded location information to the data center module (2);

[0077] The data center module (2) is used to perform congestion analysis on the location information reported by each router node (1) and each monitoring point (3) to obtain a congestion analysis result.

[0078] The link formed between two adjacent router nodes is called an edge. Monitoring points are set on the edge to monitor the passage of data packets and report the recorded location information. The data center module can analyze which routing node is a congested node and which link is a congested link based on the location information reported by the routing nodes and monitoring points, thereby adjusting the transmission path of the data packet.

[0079] For example, routing node 1 receives a data packet. At this time, the location information of the data packet is routing node 1, and routing node 1 writes the location information to the data packet. Routing node 1 sends the data packet to routing node 2. Link 1 formed between routing node 1 and routing node 2, that is, the monitoring point on the edge, detects that the data packet passes through. Then, the monitoring point records the location information carried by the data packet, that is, routing node 1, and then writes the current location information, that is, current link 1, into the data packet to update the location information of the data packet. Routing node 2 receives the data packet, records the location information carried by the data packet, that is, link 1, and then writes the current location information, that is, routing node 2, into the data packet, and updates the location information of link 1. When the data packet is at the routing node, the location information is the routing node location, and when the data packet is at the link, the location information is the link location.

[0080] like Figure 2 As shown in the figure, the data packet needs to go from M1 to S1. The data center module analyzes that Node3 (routing node 3) and Edge5 (link 5) are congested. Then, the path of Edge1 (link 1) -> Node1 (routing node 1) -> Edge2 (link 2) can be selected as the transmission path.

[0081] Furthermore, the data center module (2) broadcasts the congestion analysis result to the second type of router nodes, where the second type of router nodes are router nodes of the on-chip network other than the first type of router nodes;

[0082] The second type of router nodes dynamically updates the transmission paths of data packets that have been transmitted in the on-chip network according to the congestion analysis results.

[0083] The second type of routing node is a non-first routing node on the on-chip network. Before receiving a data packet, the data packet has been transmitted by the previous routing node.

[0084] During the transmission of data packets, the routing nodes are dynamically adjusted according to the congestion status of the on-chip network, which improves the adaptability of the NoC network and reduces the delay of data packet transmission.

[0085] The first type of routing nodes and the second type of routing nodes constitute all routing nodes of the on-chip network. Each routing node in the on-chip network can grasp the congestion of the global network in real time, avoiding making undesirable routing decisions based on local information.

[0086] Furthermore, planning the transmission path of the data packet includes planning the router nodes and links that need to be passed through.

[0087] Furthermore, the position information is written to the initial position of the payload portion of the data packet.

[0088] like Figure 3 As shown, the initial position of the payload portion of the data packet includes a first position portion and a second position portion, the first position portion is used to carry the routing node position information, namely, Node Position, and the second position portion is used to carry the link position information, namely, Edge Position. The link and routing node position information are stored separately, so that the monitoring point and routing node can distinguish the type of position information carried when reading, that is, whether it is the link position type or the routing node position type.

[0089] Specifically, the first position portion is recorded using 16 bits, and the second position portion is recorded using 16 bits.

[0090] Specifically, the first position portion is located earlier in the payload portion Payload than the second position portion and is closer to the header of the data packet.

[0091] Specifically, as an implementation method, the routing nodes and monitoring points report the recorded location information within a predetermined time period to the data center module, and the data center module performs statistical analysis on the recorded location information within the predetermined time period to obtain congestion information. Since the predetermined time period is very small, at the ms level, the lag of the analysis result is not strong, and it can be regarded as the congestion situation at the current time. For example, within 2ms, routing node 2 records the location information of the previous routing node 1 carried by 100 data packets, and routing node 4 records the location information of the previous routing node 3 carried by 10 data packets. This means that routing node 1 is more congested than routing node 3.

[0092] In a specific application scenario, for example, in a video processing SoC, the video frame is transmitted from the video input unit to the video processing unit through the NOC network. Based on the method of the present invention, the video frame data packet is recorded and carried on each routing node. The data center module collects information to analyze the congestion and broadcasts it to all routing nodes in the system. In this way, the subsequent incoming video frames can avoid crowded routing paths and optimize the transmission efficiency of the entire video data stream.

[0093] Through the present invention, the congestion perception ability of the NoC network is greatly improved, and the NoC network can be maximized, hot spot congestion areas can be avoided, and transmission conflicts can be reduced, thereby improving the overall data transmission efficiency. This global congestion information collection and broadcasting mechanism is similar to the vehicle navigation system in real life, which can help data packets in the NoC network select the optimal transmission path. In addition, by optimizing the transmission path of data packets and reducing congestion, energy consumption is also reduced.

[0094] The above are only preferred embodiments of the present invention, and are not intended to limit the implementation methods and protection scope of the present invention. Those skilled in the art should be aware that all solutions obtained by equivalent substitutions and obvious changes made using the description and illustrations of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for optimizing network-on-chip transmission based on data packet monitoring, wherein the network-on-chip includes a plurality of router nodes, characterized in that: include: Step A1, before transmitting a data packet, the router node records the location information carried by the data packet, and updates the location information carried by the data packet according to the current location information; Step A2, the router node reports the recorded location information to the data center module; Step A3, the data center module performs congestion analysis on the location information reported by all the router nodes to obtain a congestion analysis result; Step A4, the data center module broadcasts the congestion analysis result to the first type of router node, where the first type of router node is the first router node for the on-chip network to receive external data packets; Step A5: the first type of router node plans a transmission path for a data packet newly entering the on-chip network according to the congestion analysis result.

2. A method for optimizing network-on-chip transmission based on packet monitoring as claimed in claim 1, characterized in that: A monitoring point is provided on the link formed between two adjacent router nodes; In the step A1, when the monitoring point detects the data packet, the monitoring point records the location information carried by the data packet, and updates the location information carried by the data packet according to the current location information; In the step A2, the monitoring point reports the recorded location information to the data center module; In step A3, the data center module performs congestion analysis on the location information reported by each of the router nodes and each of the monitoring points to obtain the congestion analysis result.

3. The method for optimizing network-on-chip transmission based on packet monitoring according to claim 1, characterized in that: In the step A4, the data center module further broadcasts the congestion analysis result to a second type of router node, where the second type of router node is a router node of the on-chip network other than the first type of router node; In the step A5, the second type of router node dynamically updates the transmission path of the data packets that have been transmitted in the on-chip network according to the congestion analysis result.

4. The on-chip network transmission optimization method based on data packet monitoring according to claim 1, characterized in that: In step A5, planning the transmission path of the data packet includes planning the router nodes and links that need to be passed.

5. The on-chip network transmission optimization method based on data packet monitoring according to claim 1, characterized in that: In the step A1, the position information is written to the initial position of the payload part of the data packet.

6. A network-on-chip transmission optimization system based on data packet monitoring, characterized in that: A method for optimizing network-on-chip transmission based on data packet monitoring according to any one of claims 1 to 5, comprising a connected router node and a data center module: The router node is used to: before transmitting the data packet, record the location information carried by the data packet, update the location information carried by the data packet according to the current location information, and report the recorded location information to the data center module; The data center module is used to: perform congestion analysis on the location information reported by all the router nodes to obtain a congestion analysis result, and broadcast the congestion analysis result to a first type of router node, where the first type of routing node is the first routing node for the on-chip network to receive an external data packet; The first type of router nodes is used to plan a transmission path for a data packet newly entering the on-chip network according to the congestion analysis result.

7. The on-chip network transmission optimization system based on data packet monitoring as claimed in claim 6, characterized in that: A monitoring point is provided on the link formed between two adjacent router nodes, and the monitoring point is connected to the data center module; The monitoring point is used to: when the data packet is monitored, record the location information carried by the data packet, update the location information carried by the data packet according to the current location information, and report the recorded location information to the data center module; The data center module is used to perform congestion analysis on the location information reported by each of the router nodes and each of the monitoring points to obtain the congestion analysis result.

8. The on-chip network transmission optimization system based on data packet monitoring as claimed in claim 6, characterized in that: The data center module further broadcasts the congestion analysis result to a second type of router node, where the second type of router node is a router node of the on-chip network other than the first type of router node; The second type of router node dynamically updates the transmission path of the data packet that has been transmitted in the on-chip network according to the congestion analysis result.

9. The on-chip network transmission optimization system based on data packet monitoring according to claim 6, characterized in that: The transmission path planning for the data packet includes planning the router nodes and links that need to be passed through.

10. The on-chip network transmission optimization system based on data packet monitoring according to claim 6, characterized in that: The position information is written to the initial position of the payload portion of the data packet.