Q学习博弈论的WSNs混合覆盖漏洞恢复

Hybrid Coverage Vulnerability Recovery in WSNs Based on Q Learning Game Theory

针对恶劣环境下分布式无线传感器网络,为了降低成本与恢复能力,提出了一种Q学习博弈论的无线传感器网络混合覆盖漏洞恢复方法。首先设计了一种能够以分散、动态和自治的方式缩小覆盖差距的混合算法,该方法利用基于Q学习算法的博弈论概念,融合了节点重新定位和功率传输调整两种覆盖控制方案。对于所制定的潜在博弈论,传感器节点可以仅使用局部熟悉来恢复覆盖漏洞,从而减小覆盖间隙,每个传感器节点选择节点重新定位和调整感知范围。最后仿真结果表明,这里的提出的方法能够在存在连续随机覆盖漏洞条件下保持网络的整体覆盖。

Aiming at distributed wireless sensor networks in harsh environment,in order to reduce cost and recovery ability,a hybrid coverage vulnerability recovery method of wireless sensor networks based on Q-Learning game theory was proposed.Firstly,a hybrid algorithm which could narrow the coverage gap in a decentralized,dynamic and autonomous way was designed.The concept of game theory based on Q-learning algorithm and integrates two coverage control schemes:node relocation and power transmission adjustment were used.For the developed potential game theory,sensor nodes could only use local familiarity to recover coverage vulnerabilities,so as to reduce the coverage gap.Each sensor node selects nodes to place and adjust the sensing range.Finally,the simulation results show that the proposed method can maintain the overall coverage of the network under the condition of continuous random coverage vulnerabilities.