能量获取无线传感器网络中基于博弈论的分布式估计算法研究

Distributed estimation in wireless sensor networks has been widely studied, and specially emphasizes to prolong network lifetime under bandwidth constraints, battery capacity constraints and estimated performance constraints. But energy harvesting technologies change the monotone decreasing characteristic of battery power. The distributed estimation strategy in energy-harvesting wireless sensor networks, i.e., the distributed estimation based on game theory, is studied. It can make effective use of harvesting energy and create conditions to prolong network lifetime. The project focuses on the distributed estimation problem in energy-harvesting wireless sensor networks. Firstly, according to the concept of non-cooperative game theory, the project is divided into two parts: complete information game and incomplete information game. Then, considering energy-harvesting modes and distributed estimation under application scenarios, proper models of complete information game and incomplete information game on distributed estimation are constructed. Finally, finite strategy games will be depicted mainly by the game tree and its existence of equilibrium solution will be validated. The project aims to explore on distributed estimation in energy-harvesting wireless sensor networks based on game theory, and promote the further development of energy harvesting wireless sensor networks.

无线传感器网络分布式估计问题已经得到了广泛的研究，特别注重在节点带宽受限、电池能量受限及估计性能约束的情况下延长网络生命周期。而能量获取技术改变了电池能量的单调递减特征。研究一种基于能量获取技术的分布式估计策略，即基于博弈论的分布式估计策略，能够有效利用节点获取的能量，为进一步延长网络生命周期创造条件。本项目围绕能量获取无线传感器网络中的分布式估计问题开展研究，首先以非合作博弈理论为主线索，将本项目划分为分布式估计的完全信息博弈和不完全信息博弈；接着进一步综合考虑微能量获取规律和分布式估计应用场景，构建恰当的分布式估计完全信息和不完全信息博弈模型；最后，采纳博弈树等方式描述多参与者有限策略博弈，并验证解的存在性；旨在探索能量获取无线传感器网络中基于博弈论的分布式估计算法，推动能量获取无线传感器网络的进一步发展。

无线传感器网络分布式估计问题已经得到了广泛的研究，特别注重在节点带宽受限、电池能量受限及估计性能约束的情况下延长网络生命周期。而能量获取技术改变了电池能量的单调递减特征。本项目围绕能量获取无线传感器网络中的分布式估计问题开展了研究，首先以非合作博弈理论为主线索，综合考虑微能量获取规律和分布式估计应用场景，构建恰当的分布式估计完全信息和不完全信息博弈模型；同时，采纳博弈树等方式描述多参与者的对称和重复策略博弈，并验证解的存在性；在分布式估计算法中的量化策略方面，也提出了基于地理位置的自适应量化策略以及能容忍信道噪声的与位置相关的自适应量化策略；还有传感器与执行器网络方面，针对分布式估计问题，提出了一种基于Kriging的能量有效调度策略；最后验证系统方面，设计了太阳能采集系统和无线充电系统等。本项目在探索能量获取无线传感器网络中基于博弈论的分布式估计算法方面，做出了不错的示范作用。