论文部分内容阅读
针对水下无线传感器网络节点选择“组合爆炸”的问题,研究了低计算复杂度节点选择问题。首先,在量化量测的条件下推导了后验克拉美罗下界(PCRLB)与节点位置的关系,为节点选择提供了准则;然后,将GBFOS算法、贪心算法和随机算法与推导的PCRLB相结合,设计了低计算复杂度的节点选择策略。实验结果表明,GBFOS算法和贪心算法可以在保持跟踪性能不退化的情况下,大幅度降低计算复杂度,非常适合解决密集水下网络节点选择问题。此外,还将GBFOS算法应用到非理想信道条件下节点选择问题,实验结果显示考虑非理想信道的影响可以大幅提高跟踪性能。
Aiming at the problem of “combinatorial explosion ” for underwater wireless sensor network node, the node selection problem with low computational complexity is studied. Firstly, the relationship between PCRLB and node location is deduced under the condition of quantitative measurement, which provides a criterion for node selection. Secondly, combining GBFOS algorithm, greedy algorithm and stochastic algorithm with PCRLB , A node selection strategy with low computational complexity is designed. Experimental results show that GBFOS algorithm and greedy algorithm can greatly reduce computational complexity while keeping the tracking performance unchanged, which is very suitable for solving the problem of node selection in dense underwater networks. In addition, the GBFOS algorithm is also applied to the node selection under non-ideal channel conditions. The experimental results show that the tracking performance can be greatly improved by considering the influence of the non-ideal channel.