Reasearch On Confident Information Coverage Extension Region Coverage Method For Wireless Sensor Networks Based On Reinforcement Learning

Wireless sensor networks have attracted much attention due to their wide application in health and safety monitoring,especially in potentially dangerous workplaces.When an accident occurs in the monitoring area,the measuring physical quantity such as the radionuclides in the uranium tailings pond,which will cause serious radioactive pollution to the surrounding area once it spreads and posing a major threat to surrounding citizens' health and safety.In order to cover the radionuclide diffusion extended region in time,constructing an efficient extended region link and designing a high-speed,energy-efficient node deployment method have become the core of the security monitoring problem.The purpose of this thesis is to relocation redundant sensors in the initial region of interest(ROI)to provide coverage in the new region of interest(NROI)and to maintain the connectivity between ROI and NROI.The main tasks include:(1)Based on the novel Confident Information Coverage(CIC)model,this thesis studies how to extend the network life by activating a small number of nodes to provide confident information coverage service in the network area when the sensor energy is limited.Therefore,when NROI needs coverage,ROI can provide enough redundant nodes for moving,which is called CIC based network life maximization(CICNLM)problem.To effectively handle the CICNLM problem,we use the variable action set learning automata(LA)as the node learning model,and propose a multi-chain node coverage set generation algorithm,which can satisfy the coverage requirement of network area while activating as few nodes as possible.In order to verify the effectiveness of the proposed approach,we conducted a series of experiments.The experimental results show that CICVLA performs better than other typical algorithms and can effectively extend the network life.(2)This thesis analyzes the CIC-based sensor relocation for NROI(CIC-ISR-NROI)problem,which aims to reposition redundant sensors in ROI from their initial location to NROI to maintain the connectivity between ROI and NROI.A link construction algorithm based on reinforcement learning for confident information coverage extension region is proposed to complete node relocation.This algorithm first adopts two efficient sensor relocation algorithms: RRTLEC and optimized RRTLEC(ORRTLEC).It also introduces a learning automata to achieve the matching between nodes and relocation targets,and to reduce the energy consumption of mobile nodes.To verify the effectiveness of the proposed schemes,we conduct a series of experiments.Experimental results show that the performance of our algorithms is better than the other typical algorithms by the metrics of the total moving energy consumption and the maximum relocation time.