信息驱动的海洋传感网目标定位最优部署策略

Information-driven optimal placement strategy for target localization in ocean sensor networks

针对海洋传感网(OSNs)最小化目标定位误差的最优部署策略进行研究,推导得到接收信号强度指示值(RSSI)的费雪信息矩阵,并基于此得到目标函数.由于受到非高斯噪声的影响,原信息矩阵无法得到其闭环表达式.为此,利用一种蒙特卡罗(MC)策略,基于粒子化的思想得到信息矩阵的闭环表达式.随后推导出该函数的一种可行解,进而得到一种OSNs的最优部署策略.当目标节点初始位置未知时,该部署策略无法实施,进一步提出一种目标节点定位方法(TLM),将原定位问题转化为广义信赖域子问题用二分法进行求解.该方法避免了在非高斯噪声情况下利用最大似然算法因高度非凸性及计算复杂性而导致定位困难的问题,为最优部署策略提供了有效的目标初始位置.仿真实验结果证明所提出最优部署策略及定位方法的有效性.

An optimal placement strategy for minimizing the error of target localization in ocean sensor networks(OSNs) was investigated,and the object function was obtained via the Fish information matrix of received signal strength indication(RSSI).As it was infeasible to acquire the closed-form expression for the Fish information matrix when the noise was non-Gaussian,thus a Monte Carlo(MC) strategy was utilized,with which the closed-form of information matrix was obtained based on a particle theory.An optimal placement strategy for target localization in OSNs was obtained after figuring out the feasible solution of the function.Nevertheless,it was hard to exploit the strategy when the initial position of the target was unknown.Therefore,a target localization method(TLM) that converted the localization problem into a generalized trust region sub-problem was proposed,in which a bisection method was used to acquire the solution.The method could be an alternative to the maximum likelihood algorithm,which was challenging to figure out the target location when existing non-Gaussian noise due to the high non-convexity and computational complexity,and could provide effective initial position of the target for the optimal placement strategy.Simulation experiment results show the effectiveness of the proposed optimal placement strategy and the localization method.