基于μ+λ进化计算的传感器网络定位算法  

Location Estimation in Wireless Sensor Networks Based on μ+λEvolutionary Algorithm

作　　者：叶苗[1] 程小辉[1] 邓昀[1] 

机构地区：[1]桂林理工大学信息科学与工程学院,广西桂林541004

出　　处：《微电子学与计算机》2013年第2期9-12,17,共5页Microelectronics & Computer

基　　金：国家自然科学基金项目(61063001;61262075);广西自然科学基金项目(0832264);广西高等学校重大科研项目(201201ZD012)

摘　　要：基于接受信号强度(RSS)测距的定位方法是无线传感器网络中成本低而普遍使用的方法,但容易受到干扰降低定位精度.本文通过运用贝叶斯法则对RSS信号测距的概率模型进行详尽分析后,依据最大似然估计法则建立了更加合理的概率定位模型,然后针对模型不好求解的特点,结合传感器网络传输特点设计了基于μ+λ进化计算的求解算法.最后通过仿真实验证明了建立的概率定位模型和设计的基于μ+λ进化计算求解算法能降低环境干扰的影响,提高传感器节点的定位精度.One of the most commonly--used location methods in distance measurements is based on received signal strength （RSS） because this method is cheap for wireless sensor networks. However, its precision in location is easily affected by the interference of the circumstances. This paper employs the principle of the Bayesian chaining rule to thoroughly analyze the probability model of RSS--based distance measurements and set up a more reasonable probability location model based on the principle of maximum likelihood estimation. Considering the characteristics of the transmission of sensor networks, this paper designs an algorithm based on evolutionary algorithm since the model is difficult to work out. Through simulation experiment, the newly established probability location model and the algorithm based on evolutionary algorithm are finally proven to reduce the circumstantial interference and improve the location precision of the sensor nodes.

关 键 词：传感器网络 RSS定位 贝叶斯法则 最大似然估计 概率模型 μ+λ进化算法 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]