A Cooperative Interaction Method for Distribution Networks Based on the Integration of Sensing and Computing

Abstract

The present invention provides a coordinated interaction method for a distribution network based on integrated communication, sensing, and computing. The method includes obtaining an end-side data queue; transmitting data through a channel connection to a terminal; generating a data calculation queue for an edge server; obtaining various delay information during the edge-end coordination process; taking meeting the set delay requirement as the optimization goal, and continuously optimizing the edge data transmission process and the computing resource allocation process based on the channel selection principle and the computing resource allocation principle to achieve coordinated interaction of the distribution network based on integrated communication, sensing, and computing. When terminal access confrontation occurs on a certain channel in the present invention, terminals with a larger SINR are preferentially allowed to access the channel, thereby resolving the confrontation and realizing edge-end coordination optimization of channel selection and efficient transmission of low-delay and low-power data. At the same time, the difference between the total computing resources and the data calculation allocation resources is used to replace the resources allocated for block authentication calculation, thereby realizing decoupling between data calculation and block authentication calculation.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communication, and particularly relates to a coordinated interaction method for a distribution network based on integrated communication, sensing and computing. Background Art

[0002] With the wide application of technologies such as 5G (the fifth-generation mobile communication technology), cognitive IoT (cognitive Internet of Things), and Web3.0 (the third-generation Internet) in the distribution network, its operation mode and service innovation have undergone revolutionary changes. At the same time, it has also put forward new requirements for the integration and coordinated interaction of multi-dimensional heterogeneous resources such as distribution network sensing, transmission, and computing. Through the integrated communication, sensing and computing resource security allocation technology, the distribution network can achieve rapid transmission and processing of large-capacity data, monitor the real-time operation state data of electrical equipment, ensure the security of data transactions, and support low-latency and high-security data transmission and interaction in the distribution network, which is a great help for the digitalization and intelligentization of the power grid.

[0003] However, in the prior art, the situation where multiple terminals select the same channel during the process of uploading terminal data to the edge server is ignored, and the state information for solving adversarial access cannot be effectively mined from the feedback of the edge server, resulting in poor learning performance, slow convergence speed, unreasonable resource allocation, and high data transmission queuing delay. The prior art also ignores the coupling of data calculation delay and block authentication calculation delay involved in computing resource allocation, and does not consider the relationship between dynamically balancing delay and security, resulting in the inability to ensure the balance and compromise between delay and security. Summary of the Invention

[0004] In view of this, the present invention aims to propose a coordinated interaction method for a distribution network based on integrated communication, sensing and computing to solve the above problems existing in the prior art.

[0005] To solve the above technical problems, the present invention provides the following technical solutions:

[0006] A coordinated interaction method for a distribution network based on integrated communication, sensing and computing includes the following steps:

[0007] Obtain the end-side data queue, where the end-side data queue is the total amount of data that all terminals are ready to transmit to the edge server;

[0008] Transmit data through the channel to connect the terminals;

[0009] Generate the data calculation queue of the edge server, where the data calculation queue is the amount of data to be processed transmitted by the terminals to the edge server;

[0010] Obtain various delay information in the edge - terminal collaboration process. The various delay information includes at least the transmission - related delay information generated by the total data volume of the terminal - side data queue during data transmission and the calculation - related delay information generated by the data volume to be processed in the data calculation queue after computing resource allocation.

[0011] Taking meeting the set delay requirement as the optimization goal, continuously optimize the edge data transmission process and the computing resource allocation process based on the channel selection principle and the computing resource allocation principle, so as to achieve the collaborative interaction of the distribution network based on integrated communication, sensing and computing.

[0012] Among them, the channel selection principle is that when there is terminal access confrontation in the channel, the terminal access channels are sequentially selected in descending order according to the edge - terminal signal - to - interference - plus - noise ratio.

[0013] The computing resource allocation principle is to use the difference between the total computing resources and the computing resources allocated for data calculation to replace the resources allocated for block authentication calculation.

[0014] Furthermore, taking meeting the set delay requirement as the optimization goal, continuously optimizing the edge data transmission process and the computing resource allocation process based on the channel selection principle and the computing resource allocation principle is constructed as a first joint optimization problem model for solution. The first joint optimization problem model is as follows:

[0015]

[0016] In the formula, is the channel sensing time allocation ratio of the terminal , is the channel selection indication variable of the terminal in the th time slot, is the th time slot, and the terminal transmits data through the channel with the transmit power of is the computing resource allocated by the edge server for processing data from the terminal , is the set of time slots, is the set of terminals, is the transmission queuing delay of the terminal - side queue, is the queuing delay weight of the edge - side calculation queue, is the th calculation queuing delay of the data calculation queue, is the blockchain authentication calculation delay weight, is the blockchain authentication calculation delay; is the time allocation ratio constraint, is the set of terminals; is the channel availability constraint, is the Slot terminal Channel availability indication variable; , is the channel selection constraint, indicating that each terminal can select at most one channel for transmission within each time slot, and each channel can be multiplexed by at most terminals within a time slot, is the total number of channels, is the th channel, is the channel set; is the power allocation constraint, indicating that the terminal transmission power is discretized from to levels, where the th level is , , and are the minimum and maximum values of the transmission power respectively; is the computing resource allocation constraint, indicating that the total computing resources used to process the data of terminals must be less than or equal to the available computing resources of the edge server ; represents the terminal energy consumption constraint, that is, the sum of the sensing and transmission energy consumption of the terminal in time slots must be less than or equal to the maximum energy consumption constraint ; represents the channel selection reliability constraint, that is, the SINR of the channel selected by the terminal must be greater than or equal to the threshold ; represents the data volume of the end-side data queue in the th time slot and the data volume of the data calculation queue is stable at an average rate.

[0017] Furthermore, based on Lyapunov optimization, the first joint optimization problem model is converted into the second joint optimization problem model, as follows:

[0018]

[0019] In the formula, represents the weights of data transmission delay, data calculation delay and blockchain authentication calculation delay, is the data volume collected by the end-side data queue, is the data volume transmitted by the end-side data queue, is the data volume processed by the edge server, is up to the At a time slot, the terminal accumulates the magnitude of the energy consumption exceeding the expectation, which is the sensing energy consumption, and

[0020] Further, the second joint optimization problem model is transformed into a first sub-optimization problem model for solution to optimize the edge data transmission process. The first sub-optimization problem model is specifically as follows:

[0021]

[0022] In the formula, represents the optimization objective of the first sub-optimization problem model.

[0023] Further, the first sub-optimization problem model is modeled as a Markov decision process for solution. The Markov decision process includes the following elements:

[0024] State space: , where is the th state of the terminal at the

[0025] Action space: , where is the action taken by the terminal at the th time slot;

[0026] Reward: .

[0027] Further, the first joint optimization method based on the adversarial DQN network is used to solve the Markov decision process. The first joint optimization method includes a main network and a target network, and the corresponding parameter vectors are and respectively;

[0028] The main network is used to generate a policy using the state of the current time slot and the main network parameters . According to the policy, the algorithm is used to determine the action. The policy is parameterized as , that is, , which is used to represent the probability of taking the action under the policy with parameters when the system is in the state ; when terminal access confrontation occurs, the channel selection principle is used to solve the confrontation problem;

[0029] The target network is used for rewards based on environmental feedback and combines the terminal Current state and the next state Calculate the TD error and loss function, and the main network adjusts the strategy using gradient descent based on the loss function.

[0030] Furthermore, transform the second joint optimization problem model into a second sub-optimization problem model for solution to optimize the computing resource allocation process. The second sub-optimization problem model is specifically as follows:

[0031]

[0032] In the formula, is the data computing complexity of the terminal , and

[0033] Furthermore, based on the results of the first sub-optimization problem model, transform the second sub-optimization problem model into a convex optimization problem, and obtain the optimal conditions for computing resource allocation based on dual decomposition and Karush-Kuhn-Tucker conditions, specifically as follows:

[0034] Use to represent the computing resources for blockchain authentication calculation, and transform the second sub-optimization problem model into a convex optimization problem as follows:

[0035]

[0036] In the formula, represents the optimization objective of the convex optimization problem;

[0037] Construct the Lagrangian function of as:

[0038]

[0039] In the formula, and respectively represent and Lagrange multipliers;

[0040] Using the dual decomposition method, transform into:

[0041]

[0042] According to the Karush-Kuhn-Tucker conditions, the necessary condition for the Lagrangian function to take an extreme value is The first-order partial derivative of

[0043]

[0044] Solve the above formula to obtain the optimal computing resource allocation decision.

[0045] Furthermore, the solution process for obtaining the optimal computing resource allocation decision specifically includes the following steps:

[0046] Initialize the number of iterations , the iteration step size and as well as the convergence accuracy ;

[0047] Judge the iteration condition. If or then update the optimal value of to be

[0048]

[0049] Update the Lagrange multipliers and to be

[0050]

[0051]

[0052] Let Repeat the above steps until the convergence accuracy is satisfied. The obtained is the result of the optimal computing resource allocation, that is, let .

[0053] Furthermore, the delay information at least includes:

[0054] The computing queuing delay of the data computing queue:

[0055]

[0056] The blockchain authentication computing delay:

[0057] .

[0058] In summary, the present invention provides a collaborative interaction method for a distribution network based on integrated sensing, communication, and computing. The method includes obtaining an end-side data queue; transmitting data through a channel-connected terminal; generating a data calculation queue for an edge server; obtaining various delay information during the edge-end collaboration process; and taking meeting the set delay requirement as the optimization goal, continuously optimizing the edge data transmission process and the computing resource allocation process based on the channel selection principle and the computing resource allocation principle, so as to realize the collaborative interaction of the distribution network based on integrated sensing, communication, and computing. Among them, the channel selection principle is that when terminal access confrontation occurs in a channel, the terminals are sorted in descending order according to the edge-end signal-to-interference-plus-noise ratio (SINR), and the terminals with larger SINR are preferentially allowed to access the channel to resolve the confrontation and achieve edge-end collaborative optimization of channel selection and efficient transmission of low-delay and low-power data. At the same time, the difference between the total computing resources and the data calculation allocation resources is used to replace the resources allocated for block authentication calculation, so as to decouple the data calculation and the block authentication calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0060] Figure 1 It is a flowchart of a collaborative interaction method for a distribution network based on integrated sensing, communication, and computing provided by an embodiment of the present invention;

[0061] Figure 2 It is a process framework diagram of a collaborative interaction method for a distribution network based on integrated sensing, communication, and computing provided by an embodiment of the present invention;

[0062] Figure 3 It is a system diagram of a collaborative interaction system for a distribution network based on integrated sensing, communication, and computing provided by an embodiment of the present invention;

[0063] Figure 4 It is a weighted delay performance diagram provided by an embodiment of the present invention;

[0064] Figure 5 It is a schematic diagram of the energy consumption performance of a 5G cognitive IoT terminal device provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0065] To make the objectives, features, and advantages of the present invention more obvious and understandable, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the embodiments described below are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0066] Please refer to Figure 1 , an embodiment of the present invention provides a coordinated interaction method for a distribution network based on integrated communication, sensing, and computing, including the following steps:

[0067] S1: Obtain the end-side data queue, where the end-side data queue is the total amount of data that all terminals are ready to transmit to the edge server;

[0068] S2: Transmit data through a channel to connect the terminals;

[0069] S3: Generate a data calculation queue for the edge server, where the data calculation queue is the amount of data to be processed transmitted by the terminals to the edge server;

[0070] S4: Obtain various delay information during the edge-end coordination process. The various delay information at least includes transmission-related delay information generated by the total amount of data in the end-side data queue during data transmission and calculation-related delay information generated by the amount of data to be processed in the data calculation queue after computing resource allocation;

[0071] S5: Take meeting the set delay requirement as the optimization objective, and continuously optimize the edge data transmission process and the computing resource allocation process based on the channel selection principle and the computing resource allocation principle, so as to achieve coordinated interaction of the distribution network based on integrated communication, sensing, and computing;

[0072] Among them, the channel selection principle is that when terminal access confrontation occurs in the channel, the terminal access channels are sequentially selected in descending order according to the edge-end signal-to-interference-plus-noise ratio;

[0073] The computing resource allocation principle is to use the difference between the total computing resources and the computing resources allocated for data calculation to replace the resources allocated for block authentication calculation.

[0074] It should be noted that edge computing is an information transmission and processing technology derived from cloud computing and is an extension of cloud computing at the user side. Edge computing sinks part or all of the data computing and data storage to the network edge side close to the user, thereby reducing data transmission latency, shortening service response time, reducing network bandwidth, and improving computing efficiency. Distributed edge computing meets the requirements of a large number of energy terminals in the energy Internet for response speed and QoS service quality. Blockchain technology implemented based on consensus algorithms is essentially a distributed ledger technology. Commonly used consensus algorithms include Proof of Work (PoW) and Practical Byzantine Fault Tolerance (PBFT) algorithms. In the communication architecture and application scenarios of the energy Internet, blockchain technology and edge computing have extensive applications. Integrating blockchain and edge computing into a system can achieve reliable access and control of the network, with storage and computing distributed at the edge, thereby providing large-scale network servers, data storage, and effective computing close to the end in a secure manner.

[0075] In a distribution network collaborative interaction system based on blockchain technology and edge computing technology, for the problems of terminal access confrontation in the channel and unreasonable allocation of computing resources, this embodiment proposes a distribution network collaborative interaction method based on channel selection principles and computing resource allocation principles.

[0076] In this method, for the problem of terminal access confrontation in the channel during the distribution network collaborative interaction process, the channel selection principle is adopted to solve it. When there is terminal access confrontation in the channel, the terminal access channels are selected in descending order according to the edge signal-to-interference-plus-noise ratio (SINR). That is, when there is terminal access confrontation in a certain channel, that is, the number of terminals selecting the same channel for data transmission exceeds the maximum allowable access terminal number threshold of the channel, the edge server will collect the SINR (Signal to Interference plus Noise Ratio) information and sort the terminals in descending order based on the SINR, and preferentially allow the terminals with larger SINR to access the channel, thereby solving the confrontation.

[0077] For the problem that the computing resource allocation in the distribution network collaborative interaction process involves the coupling of data computing delay and block authentication computing delay and does not consider the relationship between dynamic balance delay and security, the computing resource allocation principle is adopted to solve it. The difference between the total computing resources and the data computing allocated resources is used to replace the resources allocated for block authentication computing, thereby achieving decoupling between data computing and block authentication computing. Combining the delay information in the collaborative interaction process to optimize the entire interactive collaboration process, thereby achieving a balance and compromise between data processing delay and blockchain consensus security.

[0078] The edge server obtains the end - side data queue. After the end - side data queue starts transmitting data according to the channel selection principle, a data calculation queue is generated. According to the data volume situation in the queue, such as latency, etc., it can reflect the execution optimization situation of the collaborative interaction method proposed in this embodiment.

[0079] This embodiment provides a collaborative interaction method for a distribution network based on integrated communication, sensing, and computing. When terminal access confrontation occurs on a certain channel, the terminals are sorted in descending order based on SINR, and the terminals with larger SINR are preferentially allowed to access this channel, thus solving the confrontation and realizing edge - side collaborative optimization of channel selection and efficient transmission of low - latency and low - power data. At the same time, the difference between the total computing resources and the data - calculation - allocated resources is used to replace the resources allocated for block authentication calculation, thereby realizing the decoupling between data calculation and block authentication calculation.

[0080] In a preferred embodiment of the present invention, with the goal of meeting the set latency requirement, the edge - data - transmission process and the computing - resource - allocation process are continuously optimized based on the channel selection principle and the computing - resource - allocation principle, and are constructed as a first joint - optimization problem model for solution. The first joint - optimization problem model is specifically as follows:

[0081] (1)

[0082] In the formula, is the channel - sensing time - allocation ratio of terminal , is the channel - selection indication variable of terminal in the th time slot, is the th time slot, the transmit power of terminal transmitting data through channel , is the computing resource allocated by the edge server for processing data from terminal , is the set of time slots, is the end - side queue transmission queuing latency, is the weight of the side - side computing queue queuing latency, is the th computing queuing latency of the data - calculation queue, is the weight of the blockchain authentication calculation latency, is the blockchain authentication calculation latency; is the time - allocation ratio constraint, is the set of terminals; is the channel availability constraint, is the th time slot, Channel available indication variable; , is the channel selection constraint, indicating that each terminal can select at most one channel for transmission in each time slot, and each channel is multiplexed by at most terminals in a time slot. is the total number of channels. is the th channel. is the channel set. is the power allocation constraint, indicating that the terminal transmission power is discretized into to levels. Among them, the th level is , , and are the minimum and maximum values of the transmission power respectively. is the computing resource allocation constraint, indicating that the total computing resources used to process terminal data must be less than or equal to the available computing resources of the edge server. represents the terminal energy consumption constraint, that is, the sum of the sensing and transmission energy consumption of the terminal in time slots must be less than or equal to the maximum energy consumption constraint . represents the channel selection reliability constraint, that is, the SINR of the channel selected by the terminal must be greater than or equal to the threshold . represents the data volume of the end-side data queue and the data volume of the data calculation queue in the th time slot is average rate stable.

[0083] In the process of uploading terminal data to the edge server, there are situations where multiple terminals select the same channel, which is likely to cause channel conflicts and serious interference effects, thus affecting the data transmission quality and system performance. Existing methods cannot effectively mine the status information for solving anti-access from the edge server feedback, resulting in poor learning performance, slow convergence speed, unreasonable resource allocation, and high data transmission queuing delay. Therefore, in this embodiment, a joint optimization problem of multi-dimensional resource allocation for communication, computing, and authentication is constructed. The optimization objective is to minimize the weighted sum of the queuing delay of the end-side transmission queue, the queuing delay of the edge-side calculation queue, and the block authentication calculation delay through joint optimization of the time allocation ratio, channel selection, power allocation, and computing resource allocation.

[0084] Specifically, considering the low-latency and security requirements of the distribution network service for data collection, transmission, and processing, a joint optimization problem of multi-dimensional resource allocation for communication, sensing, and computing is constructed. The optimization objective is to minimize the weighted sum of the queuing delay of the transmission queue at the edge side, the queuing delay of the computing queue at the edge side, and the block authentication computing delay by jointly optimizing the time allocation ratio, channel selection, power allocation, and computing resource allocation. The joint optimization problem is constructed as shown in Equation (1).

[0085] It is necessary to decouple the optimization decision of multi-dimensional resources for communication, sensing, and computing in a single time slot from the long-term terminal energy consumption constraint guarantee in the time slot. For this problem, in a preferred embodiment of the present invention, a virtual queue in Lyapunov optimization is introduced to convert the long-term terminal energy consumption constraint into a queue stability constraint to achieve inter-slot decoupling. Define the virtual energy deficit queue corresponding to as , and its queue backlog evolution formula is

[0086] (2)

[0087] In the formula, represents the cumulative amount by which the terminal exceeds the expected energy consumption up to the th time slot. The larger is, the more the terminal exceeds the expected energy consumption, and the more difficult it is to meet the long-term terminal energy consumption constraint ; conversely, it means that the terminal

[0088] exceeds the expected energy consumption less. Define

[0089] (3)

[0090] The Lyapunov drift is defined as the conditional expectation change in two consecutive time slots. By minimizing the upper bound of the Lyapunov drift plus penalty, the queue backlog value can be effectively guaranteed to be low. Therefore, the optimization problem is transformed into

[0091] (4)

[0092] In the formula, represents the weights of the data transmission delay, data computing delay, and blockchain authentication computing delay, is the amount of data collected by the data queue at the edge side, is the amount of data transmitted by the data queue at the edge side, ​​is the amount of data processed by the edge server, is the sensing energy consumption, is the transmission energy consumption.

[0093] In a specific implementation manner of this embodiment, a collaborative interaction model for a distribution network based on integrated communication, sensing, and computing is proposed. The joint optimization problems P1 and P2 are solved for this model, and the model gives a feasible calculation method for each parameter in the above joint optimization problem, so as to facilitate understanding how the constructed joint optimization problem specifically realizes the collaborative interaction method of the distribution network in conjunction with other embodiments.

[0094] This model includes two parts: a channel collaborative interaction sensing and data transmission model and an edge collaborative interaction computing model. The following is a specific introduction to these two parts.

[0095] (1) Channel collaborative interaction sensing and data transmission model

[0096] Assume there are 5G cognitive IoT terminals, and their set is . The sensing layer contains channels, and their set is . The time slot set is , and the length of each time slot is . In the th time slot, each terminal senses the availability of the channel based on cognitive radio technology and performs data transmission. Define the channel sensing time allocation ratio of terminal as , where is used for channel sensing, and is used for data transmission. Assume that the energy consumption required to sense one channel is , and the required time is , then the sensing energy consumption of terminal is

[0097] (5)

[0098] Among them, represents rounding down, and represents the number of channels that the terminal can sense.

[0099] Define the channel selection indication variable of terminal in the th time slot as , when it means that in the th time slot, terminal selects to access channel . Define the channel availability indication variable of terminal in the th time slot as . Channel availability is obtained based on channel collaborative interaction perception, where represents the channel is available for the terminal , otherwise represents unavailable. Only available channels can be selected for data transmission. Therefore, .

[0100] The data collected by the terminal is stored in the local cache and can be modeled as an edge-side data queue. The input of the queue is the amount of collected data , and the output of the queue is the amount of transmitted data . Therefore, the backlog evolution of the edge-side data queue is

[0101] (6)

[0102] In the formula, represents taking the maximum value.

[0103] Based on Little's Law, the transmission queuing delay of the edge-side queue is

[0104] (7)

[0105] In the formula, is the average data arrival rate up to time slots, and its calculation formula is .

[0106] At the time slot, the transmission data rate of the terminal through the channel is

[0107] (8)

[0108] In the formula, represents the signal-to-interference-plus-noise ratio of the terminal selecting the channel , represents the co-channel interference suffered by the terminal when transmitting data through the channel , represents the Gaussian white noise power, represents the th time slot of the terminal selected channel bandwidth, represents the th time slot of the terminal when transmitting data through the channel transmission power, represents the Slot terminal Selected channel Channel gain of

[0109] Terminal The amount of data transmitted by the terminal to the edge server through the 5G channel is the minimum of the queue backlog and the theoretical transmission capacity, and the expression is

[0110] (9)

[0111] Terminal At the slot, the transmission energy consumption of the terminal is

[0112] (10)

[0113] (2) Edge collaborative interactive computing model

[0114] At each time slot, the edge server optimizes the computing resource allocation for processing data from 5G cognitive IoT terminals and completing the blockchain authentication computing required for consensus. Define as the computing resources allocated by the edge server for processing data from the terminal . Define as the total available computing resources of the edge server, then the computing resources allocated for completing block authentication computing are .

[0115] There are data queues on the edge server, corresponding to terminals. Among them, the data computing queue of the terminal on the edge server is

[0116] (11)

[0117] In the formula, is the amount of data transmitted by the terminal to the edge server at the th time slot, is the amount of data processed by the edge server from the terminal at the th time slot, expressed as

[0118] (12)

[0119] In the formula, is the data computing complexity of the terminal .

[0120] On the edge server, the computing queuing delay of the th data computing queue is

[0121] (13)

[0122] In the formula, is the average data arrival rate on the edge side up to time slots, and its calculation formula is .

[0123] Considering the consensus of the consortium blockchain based on the Practical Byzantine Fault Tolerance (PBFT) algorithm, during the consensus process, including the primary node and the secondary nodes, the primary node is responsible for proposing new transactions and blocks, and broadcasting the transactions and blocks to other secondary nodes, playing the role of interacting and coordinating the operations of other nodes. The secondary nodes are responsible for receiving the transactions and blocks broadcast by the primary node and verifying their legality. PBFT includes five steps: request, pre-prepare, prepare, commit, and reply.

[0124] Define the computing resources required to complete the consistency authentication calculation as , then the blockchain authentication calculation delay is

[0125] (14)

[0126] In the formula, is the computing resource for blockchain authentication.

[0127] A coordinated interaction method for power distribution networks integrating communication and computing based on adversarial learning proposed by the present invention. This method is divided into two stages. The first stage is a method proposed for terminal access confrontation in the channel, and the second stage is a method proposed for the coupling of data calculation delay and block authentication calculation delay involved in computing resource allocation.

[0128] In a preferred embodiment of the present invention, for the first stage, a joint optimization method for time allocation, channel allocation, and energy control based on the Deep Q-Network (DQN) (i.e., the first joint optimization method) is proposed. When terminal access confrontation occurs, adversarial perception is performed based on the edge-side SINR information, and terminals with larger SINR are preferentially allowed to access the channel, thereby solving the terminal confrontation problem. The terminal confrontation perception information is fed back to the adversarial DQN network in the form of rewards. With the assistance of the target network, the adversarial DQN main network realizes efficient learning of the adversarial perception information, effectively improving the learning rate of the adversarial DQN and reducing the terminal access confrontation conflict. The specific solution of this method is as follows:

[0129] (1) Joint optimization method for time allocation, channel allocation, and energy control based on adversarial DQN

[0130] The transformed problem Further decomposed into sub - problems, and the sub - problems are those of the joint optimization of time allocation, channel allocation, and power control.

[0131] The sub - problems are formulated as

[0132] (15)

[0133] In a further embodiment of the present invention, the sub - problems are modeled as a Markov Decision Process (MDP), which is specifically described as follows.

[0134] State space: The time slot in which the terminal has a state space that contains the information required for optimization , including the end - side queue backlog of the terminal , the terminal computing queue backlog at the edge server , the virtual energy consumption deficit queue , the weight , and and . Therefore, the state space can be represented as .

[0135] Action space: The action space consists of the channel sensing time allocation ratio , the channel selection indicator variable , and the control power , and is represented as .

[0136] Reward: Since is a minimization problem, the reward of the MDP is defined as the negative value of the optimization objective, that is .

[0137] In a further embodiment of the present invention, the proposed first - stage method uses an adversarial DQN network to solve the constructed MDP problem. The adversarial DQN network includes a main network and a target network, and their parameter vectors are respectively and . Among them, the main network uses the state of the current time slot and the main network parameter to generate a policy , and decides the action according to the policy using algorithm. The policy is parameterized as , that is , and is used to represent that when the system is in the state , with parameters Take actions under the strategy of Probability. When terminal access confrontation occurs, the edge server performs confrontation perception based on the collected edge - end SINR information, and preferentially permits terminals with larger SINR to access the channel, thus solving the terminal confrontation problem. The target network is based on the reward feedback from the environment and combines the terminal Current state And the next state Calculate the TD error and the loss function, and the main network adjusts the strategy using the gradient descent method based on the loss function.

[0138] The proposed joint optimization method for time allocation, channel allocation, and energy control based on adversarial DQN includes five stages: network initialization, action selection, adversarial perception, action execution, and network parameter update. The execution steps are specifically introduced as follows

[0139] 1) Initialize the parameters of the main network As random values, initialize the parameters of the target network .

[0140] 2) At the th time slot, the terminal Based on the strategy obtained from the adversarial DQN network , adopt The strategy to select an action . The strategy selects an action with As the exploration degree and As the exploitation degree. As the training progresses, the exploration degree Will continuously decay to a very small value, and finally the model will execute the current optimal strategy with a high probability .

[0141] 3) Define the set Indicates the set of terminals that select channel . For the terminals In the set , calculate Adopt Of . When , allow To select channel , and let . Conversely, when , do not allow the terminal To select channel , and let . Remove From the set . Determine whether terminal access confrontation occurs on any channel. If channel Has terminal access confrontation, that is , the terminals in the set are sorted in descending order based on , and the first terminals are selected to access the channel , and the remaining terminals are removed from the set .

[0142] 4) Each terminal, such as , according to the time allocation ratio in the action decision , power control , and the channel selection strategy , performs channel sensing and data transmission, observes the reward , and further updates the queue according to equations (2) and (11) , . .

[0143] Based on the current state and the next state, calculate the TD error, expressed as

[0144] (16)

[0145] In the formula, is the discount factor, indicating the degree to which the target network considers the future reward value; is the Q function of the main network and the target network. The Q value of the network is calculated through different network parameters . Introducing the Q value of the target network can improve the stability of learning; is the penalty function, expressed as

[0146] (17)

[0147] In the formula, is the number of times the channel is selected.

[0148] The TD error represents the prediction error of the environment in the current state, thereby guiding the agent to learn and adjust the strategy. Based on the TD error, calculate the loss function, expressed as

[0149] (18)

[0150] Update the main network based on the loss function

[0151] (19)

[0152] Among them, represents the update step size.

[0153] 6) , until The iteration ends at this time. Among them, the target network is updated every time slot .

[0154] In a preferred embodiment of the present invention, for the second stage, a calculation resource allocation method based on the trade-off between time delay and security (i.e., the second joint optimization method) is proposed. The resources allocated for block authentication calculation are replaced by the difference between the total calculation resources and the resources allocated for data calculation, so as to realize the decoupling between data calculation and block authentication calculation, and use the optimal time allocation, channel allocation and power control joint optimization results to transform the original problem into a convex optimization problem. Then, based on the dual decomposition and the Karush-Kuhn-Tucker (KKT) conditions, the optimality conditions for calculation resource allocation are obtained, and iterative solutions are carried out to achieve the balance between data processing time delay and blockchain consensus security. The specific scheme of this method is as follows:

[0155] (2) Calculation resource allocation method based on the trade-off between time delay and security

[0156] The transformed problem is further decomposed into sub-problems, and the sub-problem is the edge-side calculation resource allocation optimization sub-problem.

[0157] Based on the optimal time allocation, channel allocation and power control results of the sub-problem, solve the sub-problem, which is expressed as

[0158] (20)

[0159] In a further embodiment of the present invention, the data calculation and the blockchain authentication calculation are decoupled, and using to represent the calculation resources for blockchain authentication calculation, and further substituting , , , into it, and is transformed into a convex optimization problem

[0160] (21)

[0161] Construct the Lagrangian function of

[0162] (22)

[0163] In the formula, and respectively represent and Lagrange multipliers

[0164] Using the dual decomposition method, is transformed into

[0165] (23)

[0166] According to the KKT conditions, the necessary condition for the Lagrangian function to take an extreme value is The first-order partial derivative of is equal to zero, expressed as

[0167] (24)

[0168] The optimal computing resource allocation decision can be obtained by solving equation (24).

[0169] In a further embodiment of the present invention, a computing resource allocation method based on a compromise between delay and security is proposed to solve equation (24), and the specific steps are as follows:

[0170] 1) Initialize the number of iterations , the iteration step and as well as the convergence accuracy .

[0171] 2) Judge the iteration condition. If or then update The optimal value of is as

[0172] (25)

[0173] 3) Update the Lagrange multipliers and as

[0174] (26)

[0175] (27)

[0176] 4) Let Repeat steps 2)-3) until the convergence accuracy is satisfied. The obtained is the optimal computing resource allocation result, that is, let .

[0177] Compared with the prior art, it has the following advantages:

[0178] 1. The present invention provides a joint optimization method for time planning, channel allocation, and energy control based on adversarial DQN. When there is a terminal access confrontation on a certain channel, that is, the number of terminals selecting the same channel for data transmission exceeds the maximum allowable access terminal number threshold of the channel, the edge server will collect the edge SINR information and sort the terminals in descending order based on the SINR, and preferentially allow the terminals with larger SINR to access the channel, thereby resolving the confrontation and achieving edge collaboration optimization of channel selection and efficient transmission of low-latency and low-power data.

[0179] 2. The present invention provides a computing resource allocation method based on a trade-off between latency and security. The difference between the total computing resources and the data computing allocated resources is used to replace the resources allocated for block authentication computing, thereby achieving decoupling between data computing and block authentication computing and transforming the original problem into a convex optimization problem. Then, based on dual decomposition and KKT conditions, the optimal conditions for computing resource allocation are obtained, and a computing resource allocation method based on a trade-off between latency and security is proposed for iterative solution to achieve a balance between data processing latency and blockchain consensus security.

[0180] The above is a detailed introduction to an embodiment of a coordinated interaction method for a distribution network based on integrated communication, sensing, and computing of the present invention. The following is a detailed introduction to an embodiment of a coordinated interaction system for a distribution network based on integrated communication, sensing, and computing of the present invention.

[0181] Please refer to Figure 2 , this embodiment provides a coordinated interaction system for a distribution network based on integrated communication, sensing, and computing, which can be divided into two parts: an integrated communication, sensing, and computing edge computing device and a 5G cognitive IoT terminal device.

[0182] 1) Integrated communication, sensing, and computing edge computing device. The integrated communication, sensing, and computing edge computing device proposed by the present invention includes: a joint optimization module for time allocation, channel allocation, and energy control based on adversarial DQN, a computing resource allocation module based on a trade-off between latency and security, a communication module, a data storage module, and a power supply module.

[0183] Joint optimization module for time allocation, channel allocation, and energy control based on adversarial DQN: It performs five steps of network initialization, action selection, confrontation perception, action execution, and network parameter update based on the edge SINR information collected by the SINR information acquisition module, thereby achieving joint optimization of edge collaboration for time allocation, channel allocation, and energy control and resolving terminal access confrontation.

[0184] Computing Resource Allocation Module Based on the Trade-off between Latency and Security: The computing resource allocation module based on the trade-off between latency and security uses the optimal strategy generated by the joint optimization module of time allocation, channel allocation, and energy control based on adversarial DQN to decouple data computing and blockchain authentication computing, transform the edge server computing resource allocation optimization problem into a convex optimization problem, and obtain the optimal result of computing resource allocation based on dual decomposition and KKT conditions.

[0185] Communication Module: The communication module is responsible for communicating with 5G cognitive IoT terminals, receiving the electrical equipment data collected by the terminals and sending down the resource allocation strategy. At the same time, when terminal access conflicts occur, it collects the edge SINR information.

[0186] Data Storage Module: The data storage module is mainly composed of a flash memory unit (Flash) and a random access memory unit (RAM). Among them, the RAM unit serves as the memory module of the portable electronic positioning tag device, providing the basis for the high-speed operation of the sensing, communication, and computing integrated edge computing device; while the flash memory unit serves as the long-term storage module of device data, providing sufficient storage space for the received electrical equipment data, etc.

[0187] Power Supply Module: The power supply module is responsible for providing a stable voltage level to each module in the sensing, communication, and computing integrated edge computing device and ensuring the normal operation of the system circuit.

[0188] 2) 5G Cognitive IoT Terminal Device. The 5G cognitive IoT terminal device proposed by the present invention includes: a data acquisition module, a channel sensing module, a communication module, a data storage module, and a power supply module.

[0189] Data Acquisition Module: The data acquisition module is responsible for collecting the electrical equipment data connected to it.

[0190] Channel Sensing Module: The channel sensing module senses the availability of the channel based on cognitive radio technology.

[0191] Communication Module: The communication module is responsible for communicating with the sensing, communication, and computing integrated edge computing device, uploading the collected electrical equipment data and receiving the optimal resource allocation strategy sent down by the sensing, communication, and computing integrated edge computing device.

[0192] Data Storage Module: The data storage module is mainly composed of a flash memory unit (Flash) and a random access memory unit (RAM). Among them, the RAM unit serves as the memory module of the portable electronic positioning tag device, providing the basis for the high-speed operation of the 5G cognitive IoT terminal device; while the flash memory unit serves as the long-term storage module of device data, providing sufficient storage space for the received electrical equipment data, etc.

[0193] Power module: The power module is responsible for providing a stable voltage level to each module in the 5G cognitive Internet of Things terminal device and ensuring the normal operation of the system circuit.

[0194] The present invention proposes a coordinated interaction system for a distribution network based on integrated sensing, communication, and computing, including an integrated sensing, communication, and computing edge computing device and a 5G cognitive Internet of Things terminal device. Among them, the integrated sensing, communication, and computing edge computing device includes a joint optimization module for time allocation, channel allocation, and energy control based on adversarial DQN, a computing resource allocation module based on the trade-off between delay and security, a communication module, a data storage module, and a power module; the 5G cognitive Internet of Things terminal device includes a data acquisition module, a channel sensing module, a communication module, a data storage module, and a power module. The coordinated interaction system for the distribution network based on integrated sensing, communication, and computing can, first, in terms of sensing, optimize the time allocation ratio of channel sensing and data transmission based on cognitive radio technology to achieve intelligent sensing; second, in terms of transmission, jointly optimize channel selection and power allocation, and solve the access confrontation problem through edge-side coordinated interaction to achieve low-latency and low-power data transmission; finally, in terms of computing, optimize the edge-side computing resource allocation to achieve low-latency data processing and blockchain consensus security, ensuring the balance and trade-off between delay and security, and providing a basis for the efficient and secure coordinated interaction of the distribution network.

[0195] The following uses a simulation analysis example to illustrate the advantages of the present invention compared with the prior art.

[0196] The simulation experiments of the method and system proposed in the present invention use MATLAB software and are based on an Intel Core i7-6900K CPU. The test site area is 100m×100m, including 8 channels, 25 5G cognitive Internet of Things terminal devices, and 1 integrated sensing, communication, and computing edge computing device. The 5G cognitive Internet of Things terminal devices and the integrated sensing, communication, and computing edge computing device are randomly distributed in the simulation area. In this part, the proposed method is compared with two different comparison methods. Comparison method 1 uses the simulated annealing algorithm (SAA), which is a stochastic optimization method based on the Monte Carlo iteration strategy. This method can search for a random global optimal solution within the feasible domain and can get rid of the local optimal solution in a probability jump manner, and finally converge to the global optimal solution. However, comparison method 1 cannot perform channel confrontation sensing and adopts a random access strategy when terminal access confrontation occurs. Comparison method 2 uses the traditional DQN algorithm to allocate resources based on queue backlog sensing to ensure the minimum queue delay. However, it lacks channel confrontation sensing, resulting in an excessive delay in blockchain authentication computing and a significant reduction in security.

[0197] By Figure 4It can be seen that compared with the two comparison methods, the weighted time delays of the proposed method are reduced by 21.23% and 14.39% respectively. This is because the proposed method is based on adversarial perception, enhances the learning ability of DQN, avoids terminal access conflicts, improves resource utilization efficiency, and reduces data transmission time delay. At the same time, the proposed algorithm realizes the trade-off allocation of data computing resources and blockchain authentication computing resources.

[0198] It can be seen from Figure 5 that compared with the two comparison methods, the energy consumption of the 5G cognitive IoT terminal device of the proposed method is reduced by 5.48% and 3.28% respectively. It can be seen from the figure that the average energy consumption fluctuation of the proposed method is relatively small and always remains within the energy consumption constraint range. The fundamental reason for the reduction in energy consumption is that the proposed method optimizes the channel sensing time allocation ratio of the 5G cognitive IoT terminal device channel allocation. This optimization realizes comprehensive channel sensing, reduces terminal access confrontation, and at the same time ensures reliable and stable data transmission.

[0199] Based on the same inventive concept, the embodiment of the present application also provides a power distribution network collaborative interaction system based on integrated communication, sensing and computing for implementing the above-mentioned power distribution network collaborative interaction method based on integrated communication, sensing and computing. The implementation solution provided by this system to solve the problem is similar to the implementation solution recorded in the above method. Therefore, the specific limitations in the embodiment of the power distribution network collaborative interaction system based on integrated communication, sensing and computing provided below can refer to the limitations on the power distribution network collaborative interaction method based on integrated communication, sensing and computing in the above text, and will not be repeated here.

[0200] The embodiment of the present invention provides a power distribution network collaborative interaction system based on integrated communication, sensing and computing, including:

[0201] An edge server and a terminal;

[0202] The edge server performs power distribution network collaborative interaction with the terminal. The collaborative interaction steps based on the edge server are as follows:

[0203] Obtain the end-side data queue, where the end-side data queue is the total data volume that all terminals are ready to transmit to the edge server;

[0204] Transmit data through the channel to connect the terminal;

[0205] Generate the data calculation queue of the edge server, where the data calculation queue is the amount of data to be processed transmitted by the terminal to the edge server;

[0206] Obtain various delay information in the edge-end collaboration process. The various delay information includes at least the transmission-related delay information generated during the data transmission of the total data volume of the end-side data queue and the calculation-related delay information generated after the calculation resources are allocated for the amount of data to be processed in the data calculation queue;

[0207] Taking the achievement of the set time delay requirement as the optimization goal, continuously optimize the edge data transmission process and the computing resource allocation process based on the channel selection principle and the computing resource allocation principle, so as to realize the collaborative interaction of the distribution network based on the integration of communication, sensing and computing;

[0208] Among them, the channel selection principle is that when terminal access confrontation occurs in the channel, the terminal access channels are selected in descending order according to the side-end signal-to-interference-plus-noise ratio;

[0209] The computing resource allocation principle is to use the difference between the total computing resources and the data computing allocation resources to replace the resources allocated for block authentication computing.

[0210] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A collaborative interaction method for a distribution network based on integrated communication and sensing computing, characterized in that, It includes the following steps: Obtain the edge-side data queue, where the edge-side data queue is the total amount of data that all terminals are ready to transmit to the edge server; Transmit data through the channel connecting the terminals; Generate the data calculation queue of the edge server, where the data calculation queue is the amount of data to be processed transmitted from the terminals to the edge server; Obtain various delay information in the edge-terminal cooperation process, and the various delay information at least includes the transmission-related delay information generated during the data transmission process of the total amount of data in the edge-side data queue and the calculation-related delay information generated after the calculation resources are allocated for the amount of data to be processed in the data calculation queue; Taking meeting the set delay requirement as the optimization goal, continuously optimize the edge data transmission process and the calculation resource allocation process based on the channel selection principle and the calculation resource allocation principle, so as to realize the coordinated interaction of the distribution network based on communication-sensing-computation integration; Among them, the channel selection principle is that when terminal access confrontation occurs in the channel, the terminal access channels are sequentially selected in descending order according to the edge-terminal signal-to-interference-plus-noise ratio; The calculation resource allocation principle is to use the difference between the total calculation resources and the data calculation allocated resources to replace the resources allocated for block authentication calculation.

2. The method for collaborative interaction of a distribution network based on integrated communication and sensing calculation according to claim 1, wherein, Taking meeting the set delay requirement as the optimization goal, continuously optimizing the edge data transmission process and the calculation resource allocation process based on the channel selection principle and the calculation resource allocation principle is constructed as a first joint optimization problem model for solution, and the specific form of the first joint optimization problem model is as follows: wherein, is the channel sensing time allocation ratio of the terminal , is the channel selection indication variable of the terminal in the th time slot, is the th time slot, and the transmission power of the terminal for transmitting data through the channel is the computing resource allocated by the edge server for processing data from the terminal , is the time slot set, is the terminal set, is the queuing delay of the end - side queue transmission, is the queuing delay weight of the edge - side computing queue, is the computing queuing delay of the th data computing queue, is the blockchain authentication computing delay weight, is the blockchain authentication computing delay; is the time allocation ratio constraint, is the terminal set; is the channel availability constraint, is the th time slot, and the channel availability indication variable of the terminal , are the channel selection constraints, indicating that each terminal can select at most one channel for transmission in each time slot, and each channel can be multiplexed by at most terminals in a time slot, is the total number of channels, is the th channel, is the channel set; is the power allocation constraint, indicating that the terminal transmission power is discretized into to levels, where the th level is , , and are the minimum and maximum values of the transmission power respectively; is the computing resource allocation constraint, indicating that the total computing resources for processing data of terminals must be less than or equal to the available computing resources of the edge server; represents the terminal energy consumption constraint, that is, the terminal at The sensing and transmission energy consumption of each time slot must be less than or equal to the maximum energy consumption constraint ; represents the channel selection reliability constraint, that is, the terminal selects a channel whose SINR must be greater than or equal to the threshold ; represents the data volume of the end-side data queue and the data calculation queue in the th time slot and is stable at an average rate .

3. The method for collaborative interaction of a distribution network based on integrated communication and sensing as claimed in claim 2, wherein Based on Lyapunov optimization, convert the first joint optimization problem model into a second joint optimization problem model, and the specific form is as follows: Wherein, represents the weights of data transmission delay, data calculation delay, and blockchain authentication calculation delay, is the amount of data collected by the end-side data queue, is the amount of data transmitted by the end-side data queue, is the amount of data processed by the edge server, is up to the th time slot, the terminal cumulative amount by which the energy consumption exceeds the expected value, is the sensing energy consumption, is the transmission energy consumption.

4. The method for collaborative interaction of a distribution network based on integrated communication and sensing calculation according to claim 3, wherein, Convert the second joint optimization problem model into a first sub-optimization problem model for solution to realize the optimization of the edge data transmission process, and the specific form of the first sub-optimization problem model is as follows: In the formula, represents the optimization objective of the first sub-optimization problem model.

5. The method for collaborative interaction of a distribution network based on integrated communication and sensing calculation according to claim 4, wherein Model the first sub-optimization problem model as a Markov decision process for solution, and the Markov decision process includes the following elements: State space: , is the state of the terminal at the nth time slot; Action space: , is the action taken by the terminal in the nth time slot; Reward: .

6. The method for collaborative interaction of a distribution network based on integrated communication and sensing calculation according to claim 5, wherein The Markov decision process is solved by using a first joint optimization method based on an adversarial DQN network. The first joint optimization method includes a main network and a target network, and the corresponding parameter vectors are respectively and ; The main network is used to utilize the state of the current time slot and the main network parameters to generate a policy , and according to the policy, adopt an algorithm to determine an action. The policy is parameterized as , that is , which is used to represent the probability of taking an action when the system is in state under the policy with parameter ; when terminal access confrontation occurs, the described channel selection principle is adopted to solve the confrontation problem; The target network is used to calculate TD error and loss function based on the reward of environmental feedback and combine the current state and the next state to calculate TD error and loss function. The main network adjusts the policy using gradient descent based on the loss function. ​ 7. The method for collaborative interaction of a distribution network based on integrated communication and sensing calculation according to claim 4, wherein Convert the second joint optimization problem model into a second sub-optimization problem model for solution to realize the optimization of the calculation resource allocation process, and the specific form of the second sub-optimization problem model is as follows: Wherein, is the data calculation complexity of the terminal , and is the time slot length.

8. The method for collaborative interaction of a distribution network based on integrated communication and sensing calculation according to claim 7, wherein Based on the result of the first sub-optimization problem model, convert the second sub-optimization problem model into a convex optimization problem, and based on dual decomposition and the Karush-Kuhn-Tucker conditions, obtain the optimality condition for calculation resource allocation, and the specific form is as follows: Utilize indicating computing resources for blockchain authentication calculation, transform the second sub-optimization problem model into a convex optimization problem as follows: In the formula, represents the optimization objective of the convex optimization problem; Structure The Lagrangian function of In the formula, and respectively represent and Lagrange multipliers; Using the dual decomposition method, transform into: According to the Karush-Kuhn-Tucker conditions, the necessary condition for the Lagrangian function to take an extreme value is The first-order partial derivatives of are equal to zero, which is expressed as: Solve the above formula to obtain the optimal calculation resource allocation decision.

9. The method for collaborative interaction of a distribution network based on integrated communication and sensing calculation according to claim 8, wherein The solution process for obtaining the optimal calculation resource allocation decision specifically includes the following steps: Initial number of iterations , iteration step size and as well as convergence accuracy ; Perform iterative condition judgment. If or then update the optimal value of to be Updated Lagrange multiplier and be Let Repeat the above steps until the convergence accuracy is met , and the resulting is the optimal result of computing resource allocation, that is, let .

10. The method for collaborative interaction of a distribution network based on integrated communication and sensing and computing according to claim 1, characterized in that The delay information at least includes: The calculation queuing delay of the data calculation queue: The blockchain authentication calculation delay: 。

Images