Topology Optimization Method for Wireless Sensor Networks Based on Asynchronous Deep Reinforcement Learning

Abstract

The invention discloses a wireless sensor network topology optimization method based on asynchronous deep reinforcement learning, which utilizes the rules of a scale-free network model to generate an initialized wireless sensor network topology structure; compresses the wireless sensor network topology structure; initializes an asynchronous deep reinforcement learning model; trains and testing stage; in the training stage, first serialize the wireless sensor network topology, and use a row vector to represent the wireless sensor network topology; then, the network topology row vectors are respectively input into different local network training models; secondly , the local network training model contains two neural network models, namely the action selection strategy network and the strategy evaluation network; in the testing phase, the global network training model evaluates the test data set; repeat steps 1, 2, 3 and 4; until The maximum number of iterations.

Description

technical field [0001] The invention mainly relates to the technical field of wireless sensor networks, in particular to a wireless sensor network topology optimization method based on asynchronous deep reinforcement learning. Background technology [0002] As an important part of the smart city Internet of things, wireless sensor network plays an important role in obtaining information in real time, which can make people get more information they want. By analyzing these data, we can improve the service quality of smart city and help people deal with some emergencies. Wireless sensor networks are widely used, such as smart home, smart wearable devices, intelligent transportation, environmental monitoring, homeland security, border detection and so on. The premise of the above application is that the network has high robustness, can send the sensed data to the server data center through the network, and relevant personnel or systems can carry out subsequent emergency handling str...

AI Technical Summary

Problems solved by technology

For example, the journal "Deep Actor-Critic Learning-basedRobustness Enhancement of Internet of Things" proposes a strategy for intelligent wireless sensor network topology based on deep reinforcement learning, but in the process of running the model, it is very dependent on GPU computing resources. Moreover, the isomorphism of the data will lead to the oscillation of the optimization results.

Images