The Research On Trajectory Planning Based On Aggregation Points Load Balancing And Dynamic Partitioning Algorithm In WSAN

How to use the limited energy of sensor nodes to extend the network lifetime efficiently and balanced under delay constrain, has become the main task of wireless sensor actuator networks. Both the selection of aggregation points in network, and the move strategy of the actuator will affect the energy consumption of the sensor nodes. For the unbalance of the load of the sensor nodes which is selected as aggregation points in network and the unbalance of the network energy consumption caused by the fixed actuator zones. In this paper, the trajectory planning of Wireless sensor actuator networks and dynamic partitioning of actuators are studied.Firstly, a centralized selection strategy of aggregation points based on load balancing is proposed, so that the problem of the unbalanced load of the sensor nodes as aggregation points in the WSAN can be resolved. By constructing the shortest path tree with the actuator as the root node, the communication distance between the sensor nodes and the root nodes will be shortest and the energy consumption of the network central portion the actuator will be shared by actuator, so that the energy consumption of network will be reduced. And then under the conditions of delay constraints, with the load of the sensor nodes as the primary objective to select aggregation points, our paper designed a actuator movement trajectory strategy based on load balancing to reduce the load of aggregation point which has the heaviest load, then the actuator visit the aggregation points by near neighbor algorithm, to collect the data of the network.Secondly, for the problem that the residual energy of the nodes is unbalanced caused by the traditional fixed actuator zones, our paper proposed dynamic partitioning algorithm based on the energy distribution of sensor nodes, which can control the movement of multi-actuator. Designing a energy distribution function to judge the energy distribution of the actuator zones, when the energy distribution function small than a threshold, our paper take advantage of the potential energy between actuators, and potential energy between actuator and the boundary, the actuator will self-deployment with a good coverage. And then using the remaining energy of the sensor nodes as a parameter of the potential energy which can control the actuators move to the regions of abundant remaining energy. Then partition the network with Voronoi diagram.Finally, simulations of energy consumption of sensor nodes, network lifetime and other aspects are taken in Matlab under different conditions, to verify the proposed aggregation point selection strategy and the effectiveness of the multi-actuator dynamic partitioning strategy.