CN-115509759-B - Battlefield dynamic data acquisition timeliness optimization method and device and computer equipment

CN115509759BCN 115509759 BCN115509759 BCN 115509759BCN-115509759-B

Abstract

The embodiment of the invention discloses a battlefield dynamic data acquisition timeliness optimization method, a battlefield dynamic data acquisition timeliness optimization device and computer equipment. The method comprises the steps of obtaining an environment data dynamic acquisition signal, converting average information age of mobile terminal data into a mathematical problem, inputting the mathematical problem into an Actor-Critic network model of a fused attention mechanism to perform bandwidth allocation proportion set calculation to obtain a calculation result, and allocating an edge server by using the calculation result to perform unloading calculation of the environment data. The Actor-Critic network model integrating the attention mechanism is deployed on the mobile terminal and the edge server. By implementing the method provided by the embodiment of the invention, the timeliness of dynamic data acquisition can be improved.

Inventors

ZHENG HUIJI
CUI LILONG
ZHANG YOUSHAN
QIU XINYUAN

Assignees

中国人民武装警察部队工程大学

Dates

Publication Date: 20260505
Application Date: 20221011

Claims (4)

1. The time-efficiency optimization method for battlefield dynamic data acquisition is characterized by comprising the following steps: Acquiring an environment data dynamic acquisition signal; Converting the average information age of the mobile terminal data into a mathematical problem, wherein the information age is defined as the difference between two time slices and comprises the difference value between the data generation time stamp and the current time slice; Inputting the mathematical problem into an Actor-Critic network model integrating an attention mechanism to perform bandwidth allocation proportion set calculation so as to obtain a calculation result, specifically, disposing ActorCritic network models on all intelligent agents, and firstly, the basic elements of the Actor-Critic network model are as follows: The state S of the mobile terminal agent comprises a position for collecting data, a position of an edge server, available resources, data information to be unloaded, two queue states of the mobile terminal agent, allocated bandwidth and channel states, which can be expressed as s= [ l, I, B and w ]; The method comprises the actions of preprocessing strategies and unloading strategies, wherein the preprocessing strategies and the unloading strategies can be expressed as a n (t)＝[cac n (t),off n (t) ], meanwhile, an edge server allocates bandwidth for each mobile terminal, and the actions of the edge server are a bandwidth allocation proportion set and can be expressed as: Wherein, the method comprises the steps of, The bandwidth allocation rate is the bandwidth allocation rate of the mobile terminal and the edge server, m is the edge server, and n is the mobile terminal; Rewarding of As rewards for agents, expressed as: The value function taking into account other agent information is expressed as: ; Wherein, the Representing a neural network and, Representing an integrated function; The coding value representing the weight sum of other agents represents the contribution of other agents, and the calculation formula is as follows: ; ; Wherein, the Is obtained by substituting actions and states of other agents into the integrated function code, and the actions and states of substituted agents n can be obtained by the same way Normalization of And After similarity of (2), weights are generated through softmax layer In this process, the process is performed, Will be Converted into a key, Will be Converting into query, firstly, converting Performing linear transformation with a common matrix Y, and substituting a nonlinear function L to obtain the code value ; Distributing an edge server by using the calculation result so as to perform unloading calculation of the environment data; the converting the average information age of the mobile terminal data into a mathematical problem includes: Modeling the optimization problem of the average information age of the mobile terminal data as a Markov process for minimizing the average information age of all the mobile terminal data to obtain mathematical problems, specifically, introducing AoI to analyze the timeliness of the data in the battlefield dynamic data acquisition system, minimizing the optimization problem of the average AoI of all the mobile terminals as a target, modeling the problem as a Markov process, adopting a deep reinforcement learning multi-agent collaborative algorithm for adding an attention mechanism for solving the optimization problem, training agents through continuous interaction of the agents with the environment and adopting a federal method, and carrying out calculation unloading and resource allocation.
2. A battlefield dynamic data acquisition timeliness optimization apparatus operable to perform the battlefield dynamic data acquisition timeliness optimization method of claim 1, comprising: The signal acquisition unit is used for acquiring environment data dynamic acquisition signals; the system comprises a conversion unit, a data generation unit and a data generation unit, wherein the conversion unit is used for converting the average information age of the mobile terminal data into a mathematical problem, the information age is defined as the difference between two time slices and comprises the difference value between a data generation time stamp and the current time slice, and the optimization problem of the average information age of the mobile terminal data is particularly modeled as a Markov process for minimizing the average information age of all the mobile terminal data so as to obtain the mathematical problem; the calculation unit is used for inputting the mathematical problem into an Actor-Critic network model of a fused attention mechanism to perform bandwidth allocation proportion set calculation so as to obtain a calculation result; And the distribution unit is used for distributing an edge server by using the calculation result so as to carry out unloading calculation of the environment data.
3. A computer device comprising a memory and a processor, the memory having stored thereon a computer program, the processor implementing the method of claim 1 when executing the computer program.
4. A storage medium storing a computer program which, when executed by a processor, performs the method of claim 1.

Description

Battlefield dynamic data acquisition timeliness optimization method and device and computer equipment Technical Field The invention relates to a computer, in particular to a battlefield dynamic data acquisition timeliness optimization method, a device and computer equipment. Background When executing tasks, fighters can acquire environmental data by using a mobile terminal, such as an unmanned plane, a robot and the like, so as to update and master environmental security situations, the acquired data needs to be uploaded to a cloud center in time to update the data, and the timeliness of the system and the freshness of the data are required to be high. The dynamic data acquisition process is very sensitive to timeliness, the traditional time delay calculation is generally the difference between the receiving time and the transmitting time, and the method has the defect that the freshness of the data cannot be reflected, and the data generated firstly and then transmitted possibly are not suitable for a system needing to update the state in real time due to the reasons of channel quality and the like. Therefore, it is necessary to design a new method to achieve improved timeliness of dynamic data acquisition. Disclosure of Invention The invention aims to overcome the defects of the prior art and provides a battlefield dynamic data acquisition timeliness optimization method, a battlefield dynamic data acquisition timeliness optimization device and computer equipment. In order to achieve the purpose, the technical scheme adopted by the invention is that the time-efficiency optimization method for battlefield dynamic data acquisition comprises the following steps: Acquiring an environment data dynamic acquisition signal; Converting the average information age of the mobile terminal data into a mathematical problem; Inputting the mathematical problem into an Actor-Critic network model integrating an attention mechanism to perform bandwidth allocation proportion set calculation so as to obtain a calculation result; distributing an edge server by using the calculation result so as to perform unloading calculation of the environment data; the Actor-Critic network model integrating the attention mechanism is deployed on the mobile terminal and the edge server. The further technical scheme is that the method for converting the average information age of the mobile terminal data into the mathematical problem comprises the following steps: the optimization problem of the average information age of the mobile terminal data is modeled as a markov process to obtain a mathematical problem. The method comprises the following steps that in the training process of an Actor-Critic network model integrating an attention mechanism, after each Ef round of learning is set, all the agents share respective model parameters and update, and the Actor and Critic networks of the agents of each mobile terminal respectively reserve the parameters with weight w and update by combining the parameters of other agents. The method comprises the further technical scheme that an Actor-Critic network model fusing an attention mechanism is trained by adopting a federal learning algorithm, and an Actor-Critic network model fusing the attention mechanism is trained by taking a process of interaction of an intelligent agent and an environment as a training sample in an experience cache set. The method comprises the following steps that an Actor-Critic network model integrating attention mechanisms outputs a bandwidth allocation proportion set through the Actor-Critic network model, and after the attention mechanisms are integrated, data sampled from an experience cache set are combined to serve as a sample set to train the Actor-Critic network model. The invention also provides a battlefield dynamic data acquisition timeliness optimizing device, which comprises: The signal acquisition unit is used for acquiring environment data dynamic acquisition signals; The conversion unit is used for converting the average information age of the mobile terminal data into mathematical problems; the calculation unit is used for inputting the mathematical problem into an Actor-Critic network model of a fused attention mechanism to perform bandwidth allocation proportion set calculation so as to obtain a calculation result; And the distribution unit is used for distributing an edge server by using the calculation result so as to carry out unloading calculation of the environment data. The conversion unit is used for modeling the optimization problem of the average information age of the mobile terminal data into a Markov process so as to obtain a mathematical problem. The invention also provides a computer device which comprises a memory and a processor, wherein the memory stores a computer program, and the processor realizes the method when executing the computer program. The present invention also provides a storage medium storing a computer program which, when exe