迭代粒子群优化的水下无线传感器网络节点自定位算法
A Localization Method for Underwater Wireless Sensor Networks Based on Modified Particle Swarm Optimization Algorithms

传统MDS-MAP（multi-dimensional scaling MAP）算法使用节点间的最短路径作为真实距离计算节点位置，但当水下无线传感器网络（underwater wireless sensor networks，UWSN）构型非均匀时，最短路径将严重偏离节点间真实距离，位置计算将产生较大误差。针对此不足，文中设计了一种基于迭代粒子群优化的RQ-PSO定位算法。该方法利用MDS-MAP算法对传感器节点完成粗定位，引入几何约束来限制粒子群初始种群范围，并采用鲁棒四边形规则对未知节点位置进行优化求解。通过理论分析和仿真，结果表明，该算法收敛速度明显高于传统粒子群算法（PSO），定位精度高于传统MDS-MAP与PSO算法，且RQ-PSO算法具有较强的鲁棒性。
For the problem that the localization error of the traditional Multi-dimensional Scaling MAP (MDS-MAP) algorithm is oversensitive to the distance matrix between nodes in Underwater Wireless Sensor Network (UWSN), a robust quadrilateral based modified Particle Swarm Optimization (RQ-PSO) is proposed. Inspired by the robust quadrilateral, geometrical constraint is introduced to narrow down the range of the initial particle swarm after the rough localization by applying MDS-MAP algorithm. The experiment results show that the proposed algorithm can decrease the location error, improve accuracy and the rate of convergence with strong robustness, compared with MDS-MAP and classical PSO