A time-sensitive network scheduling method based on a priority queuing strategy

Abstract

The application relates to a time-sensitive network scheduling method based on a priority queuing strategy, which comprises the following steps: obtaining a specific transmission link of a data packet at a data link layer based on a classification-based link selection algorithm; obtaining a flow scheduling scheme based on a priority queuing strategy; and obtaining a frame scheduling scheme based on a priority queue. The application solves the problems of uncertain time delay and large transmission jitter of existing time-sensitive data transmission, solves the problem of unbalanced load of an existing data transmission link, further manages and optimizes network resources, realizes interconnection and intercommunication between devices, improves data end-to-end transmission certainty, and improves the interconnection efficiency of an energy internet field requiring real-time monitoring or feedback.

Description

Technical Field

[0001] The present application relates to the technical field of end-to-end high-reliability and low-latency data transmission for energy internet, and in particular to a time-sensitive network scheduling method based on a priority queuing strategy. Background Art

[0002] The Energy Internet uses advanced sensors, controls, and software applications to connect hundreds of millions of devices, machines, and systems at the energy production, transmission, and consumption ends, forming the "Internet of Things Foundation" of the Energy Internet. The Energy Internet integrates operational data, weather data, meteorological data, power grid data, and power market data to conduct big data analysis, load forecasting, and power generation forecasting, thereby improving and optimizing the operational efficiency of energy production and consumption. Demand and supply will be able to be dynamically adjusted at any time, which will increase the demand for communication in the Internet of Everything. Multiple control applications need to cooperate with each other to complete automated production tasks, and the precise positioning and operation of individual control processes must be guaranteed in real time. At this time, the corresponding real-time data used for system control must be transmitted in the network with extremely low transmission latency and jitter. Therefore, in many application scenarios, the bandwidth carrying requirements for low latency, low jitter, and high reliability are very strict. Traditional Ethernet cannot meet these requirements. Time-Sensitive Networking (TSN) is considered to be the next-generation local area network technology that can meet these transmission requirements. TSN is a new technology that gives traditional Ethernet real-time and deterministic features. It can isolate the transmission of critical and non-critical data in the network and achieve efficient transmission of different types of data through traffic shaping and scheduling. It breaks through the complexity and real-time barriers of bus communication and addresses some shortcomings of existing networks. The Energy Internet contains a large amount of time-sensitive data that must be transmitted reliably with minimal latency and jitter. Failure to ensure the real-time and deterministic transmission of this data will significantly impact the development of the Energy Internet. Applying Time-Sensitive Networking (TSN) technology in the Energy Internet ensures low-latency, low-jitter transmission of time-sensitive data, enabling interoperability between devices and providing an optimal path to improving interconnection efficiency in areas of the Energy Internet that require real-time monitoring or feedback. Summary of the Invention

[0003] The purpose of the embodiments of the present application is to provide a time-sensitive network scheduling method based on a priority queuing strategy to manage and optimize network resources, optimize the transmission delay and jitter of time-sensitive data, and improve the end-to-end transmission certainty of data.

[0004] To achieve the above objectives, this application provides the following technical solutions:

[0005] The present invention provides a time-sensitive network scheduling method based on a priority queuing strategy, comprising the following steps:

[0006] The classification-based link selection algorithm obtains the specific transmission link of the data packet at the data link layer;

[0007] Obtain a flow scheduling solution based on the priority queuing strategy;

[0008] A frame scheduling scheme is obtained based on the priority queue.

[0009] The classification-based link selection algorithm obtains the specific transmission link of the data packet at the data link layer, comprising the following steps:

[0010] According to the IEEE 802.1Q standard, the priority of data packets received from upstream applications is determined and divided into three data categories: data with a priority of 7 is separately classified as Class A data, data with priorities of 1 to 6 is separately classified as Class B data, and data with priority of 0 is separately classified as Class C data. First, link selection is performed for data with priority of 7 using the IEEE 802.1Qci standard. After link selection for data with priority of 7 is completed, a link is selected for data with priorities of 1 to 6 based on link compatibility. Link compatibility uses data priority as a coefficient and comprehensively considers link delay response and link load. After link selection for data with priorities of 1 to 6 is completed, a link is selected for data with priority of 0 based on link compatibility, thus obtaining a link selection strategy for data flows of different priority levels.

[0011] The flow scheduling scheme obtained based on the priority queuing strategy includes: the remote center first obtains the transmission cycle, frame length, frame offset and priority information of the flow on the link, then calculates the length of the TSN network scheduling cycle according to the transmission cycle of all flows, calculates the frame transmission time of the flow according to the data frame length of the flow, calculates the number of frames sent by the flow within the scheduling cycle according to the TSN network scheduling cycle and the transmission cycle of the flow, obtains the sending sequence of the flow within the scheduling cycle according to the frame offset and the transmission cycle of the flow, and finally sorts all flows according to the flow priority, frame offset and frame transmission time to obtain the priority queue of the flow.

[0012] The frame scheduling scheme based on the priority queue is specifically as follows: taking out the streams from the priority queue in sequence, obtaining the frame time slot window according to the frame offset after stream scheduling and the frame transmission time, if the frame time slot window status is empty, then using the frame time slot window as the transmission time slot of the corresponding frame; if the frame time slot window status is occupied, searching for unoccupied time slot points and adding the frame transmission time to obtain the frame time slot window, until the corresponding frame time slot window is empty and used as the transmission time slot of the corresponding frame.

[0013] Compared with the prior art, the present invention has the following advantages:

[0014] It solves the existing problems of uncertain delay and large transmission jitter in time-sensitive data transmission, solves the existing problem of unbalanced load in data transmission links, further manages and optimizes network resources, realizes interconnection between devices, improves the certainty of end-to-end data transmission, and improves the interconnection efficiency in the energy Internet field that requires real-time monitoring or feedback. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0016] Figure 1 This is a block diagram of the proposed time-sensitive network scheduling method based on priority queuing strategy;

[0017] Figure 2 Select a strategy flow chart for data flow links;

[0018] Figure 3 This is a mapping diagram between data priority and switch gate queue;

[0019] Figure 4 Schematic diagram of flow scheduling strategy;

[0020] Figure 5 Schematic diagram of frame scheduling strategy.

[0021] Figure 6 Schematic diagram of traffic scheduling success rate. DETAILED DESCRIPTION

[0022] The technical solutions in the embodiments of the present application will be described below in conjunction with the accompanying drawings. It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings.

[0023] The terms "comprises," "comprising," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not preclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0024] Reference Figures 1 to 5 ,A time-sensitive network scheduling method based on priority queuing strategy,selects the link with the smallest fit according to the link fit,obtains the data stream transmission link, performs flow scheduling based on priority queuing strategy,and performs frame scheduling based on priority queue.

[0025] In the implementation method of the present invention, the priority of the data packets received from the uplink application is determined according to the IEEE 802.1Q standard and divided into three categories of data: A, B, and C. First, the IEEE 802.1Qci standard is used to select the link for the Class A data. Then, the link with the lowest link compatibility is selected for the Class B and Class C data according to the link compatibility. The link compatibility is obtained by taking the data priority as a coefficient and comprehensively considering the link delay response and the link load. Secondly, the flow scheduling is performed based on the priority queuing strategy. The TSN network scheduling period, frame transmission time, and sending sequence are calculated based on the obtained flow transmission period, frame length, frame offset, and priority information. The flow is sorted to obtain the flow priority queue. Finally, frame scheduling is performed based on the priority queue. The flow is sequentially taken out from the priority queue. The frame time slot window is obtained based on the frame offset and frame transmission time after the flow scheduling. The frame scheduling is performed according to the state of the frame time slot window.

[0026] In this example, the classification-based link selection algorithm is as follows:

[0027] For data packets received from upstream applications, the IEEE 802.1Q standard determines the priority of each packet and classifies the received data into three categories based on priority: data with a priority of 7 is classified as Class A data, data with priorities of 1 to 6 is classified as Class B data, and data with a priority of 0 is classified as Class C data. First, link selection is performed for Class A data using the IEEE 802.1Qci standard. When all Class A data has completed link selection (i.e., the Class A data queue is empty), link selection begins for Class B data. The data priority coefficient, which represents the latency requirement of the data stream, is defined as follows:

[0028]

[0029] Among them H i (0≤H i ≤6) is the priority of the i-th data stream, priority H i The larger the value, the corresponding β i The smaller the value, the higher the data flow's requirements for deterministic low latency and the higher the forwarding priority. When designing a link selection algorithm, it is necessary to evaluate the real-time link status. The real-time link status is mainly related to the link delay response and link load. A link status evaluation factor is designed to evaluate the real-time link status. The calculation formula for the link status evaluation factor is as follows:

[0030]

[0031] Among them D i (t) represents the delayed response time of link i at time t, D i _max represents the maximum delay response tolerance time of link i. If D i (t)≥D i _max, then record therefore L i (t) represents the load of link i at time t, L i _max represents the maximum load of link i. If L i (t)≥L i _max, then record therefore Then we get 0≤γ i (t)≤1. γ i The value of (t) reflects the status of the current link, γ i The smaller the value of (t), the better the real-time status of the current link. The link compatibility is obtained based on the data priority coefficient and the link status evaluation factor. The link compatibility is defined as follows:

[0032] A i,j (t) = β i γ j (t) (3)

[0033] where β i is the data priority coefficient of the i-th data stream, γ j (t) is the link status evaluation factor of link j at time t. The link fit represents the degree of adaptation of the current data flow on the specified transmission link. i,j The smaller the value of (t), the higher the degree of adaptation between the current data stream and the specified transmission link, and the more suitable the data stream is for transmission on the specified transmission link. Therefore, the current available link is selected so that A i,j The link with the smallest value of (t) transmits the i-th data stream.

[0034] In this example, the flow scheduling method based on the priority queuing policy is as follows:

[0035] The remote center first obtains the transmission period T of the upstream link i , frame length FL i and priority PT i The length of the TSN network scheduling period, discycle, is calculated based on the transmission period of all flows. The calculation formula is as follows:

[0036] discycle=LCM(T1,T2,…,Ti ,…) (4)

[0037] Where T i Represents flow F i The transmission cycle is based on the frame length FL i Calculate the frame transmission time FTT of each flow i , the calculation formula is as follows:

[0038]

[0039] Among them FL fix is the fixed overhead of frame transmission, and R is the output link flow rate. The number of frames of a flow within the network scheduling period is calculated based on the flow transmission period and the network scheduling period. The calculation formula is as follows:

[0040]

[0041] where F_num i Represents flow F i The number of frames in the network scheduling cycle, [] indicates a rounding operation. The sending sequence of the flow in each network scheduling cycle is calculated based on the first frame offset of the flow and the transmission cycle of the flow. The calculation formula is as follows:

[0042] FTS i1 =F_offset i1 (7)

[0044] FTS ij =FTS ij-1 +T i (8)

[0046] where F_offset i1 For flow F i Offset of the first frame, FTS ij For flow F i The jth element in the sending sequence corresponds to the time slot point in the network scheduling cycle, flow F i The sending sequence is characterized as follows:

[0047]

[0048] According to the sorting rules, the flow (F1, F2, ..., F n ) priority queue

[0049] ″′

[0050] FPQ=[F1,F2,…,F n ]. The sorting rules are as follows:

[0051] For priority PT i Two different flows F i and F j , assuming F i Flow priority PT i Higher than F j Flow priority PT j , if F j The first frame offset of the stream and F i The first frame offset of the stream satisfies F_offset j1 +FTT j ≤F_offset i1 , then the priority queue FPQ will be F j Flow ranked F i Before the flow, if F_offset i1 <F_offset j1 +FTT j , then the priority queue FPQ will be F j Flow ranked F i After the flow, and update F j The first frame offset of the stream is F_offset j1 =F_offset i1 +FTT i , then update the flow F according to formula (8) j The sending sequence.

[0052] For priority PT i The same two flows F i and F j , if F j The first frame offset of the stream and F i The first frame offset of the stream satisfies F_offset j1 +FTT j ≤F_offset i1 , then the priority queue FPQ will be F j Flow ranked F i Before the flow; if F_offset j1 <F_offset i1 <F_offset j1 +FTT j , you need to further calculate the value of F_offset i1 with F_offset j1 +FTT j / 2 and frame transmission time, if F_offset i1 ≥F_offset j1 +FTT j / 2, then the priority queue FPQ will be Fj Flow ranked F i Before the flow, update F i The first frame offset of the stream is F_offset i1 =F_offset j1 +FTT j , then update the flow F according to formula (8) i On the contrary, when F_offset j1 <F_offset i1 <F_offset j1 +FTT j / 2, if F_offset is satisfied i1 +FTT i <F_offset j1 +FTT j When the priority queue FPQ i Flow ranked F j Before the flow, update F j The first frame offset of the stream is F_offset j1 =F_offset i1 +FTT i , then update the flow F according to formula (8) j The sending sequence, on the contrary, if F_offset is satisfied i1 +FTT i ≥F_offset j1 +FTT j When the priority queue FPQ will F i Flow ranked F j After the flow, update F i The first frame offset of the stream is F_offset i1 =F_offset j1 +FTT j , then update the flow F according to formula (8) i The sending sequence.

[0053] In this example, the frame scheduling method based on the priority queue is as follows:

[0054] Extract the first flow F'1 from the head of the priority queue. Since the first flow has the highest priority, its corresponding frame time slot window state must be empty. According to the transmission sequence FTS'1 of F'1, the offset of each frame is obtained, that is, the offset of the jth frame is FTS' 1j , according to the frame offset, the frame slot window FSW′ of the stream F′1 is obtained 1j , the calculation formula of the frame time slot window is as follows:

[0055] FSW′ i1=[FTS′ 11 ,FTS′ 11 +FTT′ i ) (10)

[0056] FSW′ ij =[FTS′ 1j ,FTS′ 1j +FTT i ′) (11)

[0058] FTT i ' is the flow F' i The transmission time of the first flow is scheduled, and all the corresponding frame time slot windows are set to occupied state. The first flow is removed from the priority queue, and the first flow F′2 is extracted from the head end of the priority queue again. According to the frame offset, the frame time slot windows FSW′ of flow F′2 are obtained using formulas (10) and (11). 2j , judge the state of the frame slot window. If the frame slot window is empty, schedule the corresponding flow. If the frame slot window state is occupied, use the current frame offset FSW′ ij As the starting point, in [FSW′ ij ,discycle) and add it to the frame transmission time to obtain the frame time slot window and judge the status until the frame time slot window is empty and schedule the frame, and set the corresponding frame time slot window to the occupied state. When all frames of flow F′2 are scheduled, flow F′2 is removed from the priority queue and this operation is repeated to schedule the remaining frames.

[0059] In this example, the scheduling success rate based on the priority queuing policy is as follows:

[0060] Figure 6 The figure shows how the scheduling success rate changes with the number of time-triggered flows, when the number of periodic time-sensitive flows and periodic high-bandwidth rate-limited flows is 20. The figure shows that the scheduling success rate remains high. As the number of time-triggered flows increases, the scheduling success rate slowly decreases and remains stable. This demonstrates that the scheduling strategy proposed in this example can effectively schedule various flows and is applicable to high-concurrency scenarios with a large number of flows.

[0061] The above description is merely an embodiment of the present application and is not intended to limit the scope of protection of the present application. For those skilled in the art, various modifications and variations of the present application are possible. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A time-sensitive network scheduling method based on priority queuing strategy, characterized in that: The following steps are involved: The classification-based link selection algorithm obtains the specific transmission link of the data packet at the data link layer; Obtain a flow scheduling solution based on the priority queuing strategy; Get the frame scheduling scheme based on the priority queue; The flow scheduling scheme obtained based on the priority queuing strategy includes: the remote center first obtains the transmission cycle, frame length, frame offset and priority information of the flow on the link, then calculates the length of the TSN network scheduling cycle according to the transmission cycle of all flows, calculates the frame transmission time of the flow according to the data frame length of the flow, calculates the number of frames sent by the flow in the scheduling cycle according to the TSN network scheduling cycle and the transmission cycle of the flow, obtains the sending sequence of the flow in the scheduling cycle according to the frame offset and the transmission cycle of the flow, and finally sorts all flows according to the flow priority, frame offset and frame transmission time to obtain the priority queue of the flow; The frame scheduling scheme based on the priority queue is specifically as follows: taking out the streams from the priority queue in sequence, obtaining the frame time slot window according to the frame offset after stream scheduling and the frame transmission time, if the frame time slot window status is empty, then using the frame time slot window as the transmission time slot of the corresponding frame; if the frame time slot window status is occupied, searching for unoccupied time slot points and adding the frame transmission time to obtain the frame time slot window, until the corresponding frame time slot window is empty and used as the transmission time slot of the corresponding frame.

2. A time-sensitive network scheduling method based on priority queuing strategy according to claim 1, characterized in that: The classification-based link selection algorithm obtains the specific transmission link of the data packet at the data link layer, comprising the following steps: According to the IEEE 802.1Q standard, the priorities of data packets received from upstream applications are determined and divided into three data categories. Data with a priority of 7 is separately divided and recorded as Class A data, data with priorities of 1 to 6 is divided and recorded as Class B data, and data with priority of 0 is separately divided and recorded as Class C data. First, the IEEE 802.1Qci standard is used to select a link for data with priority of 7. After the link for data with priority of 7 is selected, a link is selected for data with priorities of 1 to 6 based on the link compatibility. The link compatibility uses the data priority as a coefficient and comprehensively considers the link delay response and link load. After the link for data with priorities of 1 to 6 is selected, a link is selected for data with priority of 0 based on the link compatibility, thereby obtaining the link selection strategy for data flows of different priorities.

Images