摘要
为使目标跟踪系统的精度和能量消耗达到最佳平衡,节点选择策略在基于无线传感网络的目标跟踪系统中起着关键的作用。文章将节点选择归类为数学中的背包问题,同时通过预测信息熵及能量消耗,提出一个创新性的目标方程去判定该节点是否需要参与定位。通过自适应调整背包问题中的价值阈值的节点选择策略决定背包问题中的总容量,也就是参与运算的节点数。在计算效率上我们通过贪婪竞争策略减少不合适的候选节点,同时采用遗传算法去解决随后出现的约束性优化问题。仿真结果证明该算法可以提供更高的跟踪精度以及更少的系统能量消耗。
Sensor selection is a key issue for target tracking in wireless sensor networks（WSNs）,in order to achieve an optimal tradeoff between tracking accuracy and energy consumption.This paper establishes a novel objective function to evaluate whether a sensor is appropriate for task according to its predicted information entropy and energy cost,and formulizes the problem of sensor selection as a knapsack problem.The knapsack capacity,the number of active sensors,is determined by the adaptive sensor strategy relying on innovation hard threshold.To alleviate invalid candidates,a greedy strategy is proposed for computational efficiency.Further,a genetic algorithm is introduced to solve the constrained optimization.Simulation demonstrates its superior performance both in tracking accuracy and energy efficiency.
出处
《电子技术（上海）》
2017年第2期35-38,共4页
Electronic Technology
基金
上海市自然科学基金基金编号15ZR1439800所资助
关键词
无线传感网
遗传算法
自适应选择算法
背包问题
Wireless Sensor Network（WSN）
Genetic Algorithm（GA）
Greedy Adaptive Sensor Selection（GASS）
Knapsack Problem
