The Switching And Management Strategy Of Sensor Clusters Based On Hybrid Estimation Algorithm

Wireless sensor network (WSN) with integrated sensor technology, wirelesscommunication and distributed information processing technology, was a new informationacquisition and processing technology, and demonstrates the potential of huge value in themilitary, disaster relief, environmental monitoring and other aspects. Target tracking was animportant application of the WSN, and the requirements for the cooperative schedulingefficiency of distributed sensing nodes and positioning algorithm increased with the numberof tracking targets as well as the diversity of target motion characteristics. For multi-targettracking problem, the study built a distributed multi-target tracking WSN with high noise andinvested multi-target tracking method on the platform to realize rapid multi-target tracking aswell as accurate feedback, which would offer help to high-quality system analysis and design.To facilitate the research of node scheduling, the ultrasonic node sensing domain andinterference domain were prescribed, which led to the fact that nodes in the domain of mutualinterference must be in asynchronous measurements. For the clock synchronization problem,periodic sequence synchronization strategy was proposed in accordance with the cycles of theexperimental platform. To solve the wireless channel contention, a time-shared transferprotocol based on TDMA is provided. Sensor node measuring interference and channelcontention were attributed to network resource management and scheduling problems and atask-based dynamic resource management method and target location-based dynamicclustering scheduling policy were put forward. Experimental results showed that the abovestrategies can effectively avoid the problems of the ultrasonic measurement interference andradio channel contention, and achieved efficient scheduling and management of sensor nodesin the multi-target tracking system.In addition, an adaptive hybrid estimation algorithm was introduced against the inherentdefects of the traditional method of least squares and the conventional Kalman filter algorithm.The algorithm was based on extended Kalman filter, and combined with the correction offaster convergence method of least squares, in order to achieve faster and more accuratetracking. Finally, in order to achieve the movement mutations in the grid switching, aestimation-based hybrid strategy in wireless sensor cluster switching and management wasproposed, with the addition of target recovered mechanism for grid switching monitoring andtargets losing.Experimental results showed that the hybrid estimation algorithm was not only be able toaccurately track the moving target without speed mutation, but also can achieve the effect of fast tracking for moving target with speed mutation. Mobile robots in grid switching wouldbring switching of observational node clusters, and sensor node cluster switching andmanagement algorithm based on hybrid estimation strategies for effective collaborativemanagement of observational node could achieve accurate and smooth switching.