水下无线传感器的分布式唤醒/睡眠IMM滤波跟踪

Distributed wake-up/sleep IMM filter tracking for underwater wireless sensors

研究了在水下无线传感器网络(Underwater Wireless Sensor Networks,UWSN)环境中,具有多种运动轨迹的单个目标的精确跟踪问题。首先,针对组成UWSN的每个水下无线传感器节点都是电池供电的,节能问题是一个关键问题,利用分布式体系结构估计目标状态,提高能源效率的问题,并给出了利用所提出的唤醒/睡眠(Wake-Up/Sleep,WUS)方案提高传感器节点能量效率的算法。然后,将交互式多模型(Interacting multiple model,IMM)滤波器应用于所提出的分布式体系结构,根据机动目标的运动特性组合不同的模型,以适应不同的轨迹变化,计算各模型的概率,以应对水下无线传感器网络目标运行的机动性。仿真结果表明,该算法具有较高的估计精度和能量效率。

This paper studies the problem of precise tracking of single target with multiple trajectories in underwater wireless sensor networks(UWSN).Firstly,for each underwater wireless sensor node that constitutes UWSN is battery-powered,energy saving is a key issue.The distributed architecture is used to estimate the target state and improve energy efficiency.An algorithm for improving energy efficiency of sensor nodes by using the proposed Wake-Up/Sleep(WUS)scheme is presented.Then,the Interacting Multiple Model(IMM)filter is applied to the proposed distributed architecture.Different models are combined according to the motion characteristics of maneuvering targets to adapt to different trajectory changes and calculate the probability of each model to cope with the maneuverability of underwater wireless sensor networks.Thesimulationresultsshow thatthealgorithm has high estimation accuracy and energy efficiency.