无线传感器网络下移动扩散源追踪算法

Mobile diffusion source tracking in wireless sensor networks

中文关键词  离散化浓度模型  无线传感器网络  最小二乘  Sage-Husa卡尔曼滤波  移动扩散源追踪

英文关键词  discrete concentration model  wireless sensor networks  least squares  Sage-Husa Kalman filter  mobile diffusion source tracking

中文摘要

      与移动瞬时源追踪相比, 移动扩散源追踪相对困难. 本文分析了移动扩散源扩散过程, 给出了离散化浓度 场模型. 将连续线源目标追踪问题转化为离散点源目标追踪的次优问题, 提出了一种离散化移动扩散源追踪算法. 在该算法中, 先采用约束最小二乘方法估计目标实时位置、到达时间等相关参数, 并进一步采用仅针对位置序列的 Sage-Husa卡尔曼滤波方法优化位置估计. 该算法克服了一般基于动态序列的追踪方法无法直接应用于离散移动扩 散源追踪问题的不足. 在仿真实验中, 分别在匀速率平滑曲线运动与变速非平滑曲线运动的情形下进行追踪实验, 分析了追踪精度与采样间隔以及观测节点密度的关系. 仿真结果说明了提出的移动扩散源追踪算法的有效性.

英文摘要

      Generally it is more difficult to track a mobile diffusion source than to track a mobile instantaneous source. The diffusion process of mobile sources is analyzed, and a discrete concentration model is proposed. The problem of tracking a continuous line diffusion source is transformed into a suboptimal problem of tracking a discrete mobile diffusion source, to which the discrete tracking algorithm is proposed. In the algorithm, a constrained least squares method is adopted to estimate the related parameters including positions in real-time and arrival time. Next, the Sage-Husa Kalman filter is used to obtain the optimal estimation of the target positions. The algorithm overcomes the shortcoming that the general tracking methods based on dynamic sequence cannot be directly applied to the discrete tracking problem of the mobile diffusion source. The simulation experiments are carried out in two scenarios, in one of which the target moves along a smooth curve with a constant rate, and in another the target moves along a non-smooth curve with a varying velocity. In the simulations, the relation between the tracking accuracy and the sampling density as well as the node density is investigated. The results illustrate the effectiveness of the proposed mobile diffusion tracking algorithm.