Trajectory Planning Algorithms For 3D Mobile Wireless Sensor Networks

Mobile Wireless Sensor Network(MWSN)solves the problems of communication interruption,monitoring blind spot,and energy hole in stationary sensor networks to a certain extent by introducing mobile nodes,and is widely used in space monitoring.Node trajectory is the key to ensure the quality and efficiency of space monitoring tasks.However,due to the different requirements and applications of different tasks,it is difficult to find a general path planning algorithm.Area search is one of the typical tasks of space monitoring,which refers to searching a specified area by planning the trajectory of the node.It drops the area uncertainty and determined the location of targets.After target is sensed,trajectory needs to be designed for sensor nodes to track it.In addition,how to design the trajectory of sink to efficiently collect the information of multiple monitoring nodes in the network and support the decision center to formulate strategies is also a fundamental task of space monitoring.Aiming at the above three task scenarios,this thesis mainly studies the following three aspects of the trajectory planning of MWSN:(1)In order to ameliorate the target detection efficiency of area search and reduce the uncertainty of the area,a algorithm based multivariate combination objective function(MCOFA)method is proposed to realize the sensor node's movement path planning and use GA to solve it.This method constructs three objective functions with three variables of entropy decrease(ED)?time increase(TI)?and threat degree(TD),and transfers the continuous time path planning problem into a discrete-time direction selection problem.GA is designed to obtain the node trajectory of the node under the condition that the sensor movement directionality and regional boundary constraints are satisfied,while the traditional algorithm has a single objective function and high three-dimensional space optimization complexity question.Experimental results show that MCOF drops the repeated access frequency of the area effectively,raises the target search efficiency,and avoids the node from visiting the high threat area during the movement process,and reduces the possibility of the node being attacked and damaged.(2)For improving the tracking efficiency and formation efficiency,an improved A^*(Improved A^*,IAStart)trajectory planning algorithm(Fuzzy System based Improved AStart Algorithm,FUZIAStart)based on fuzzy logic system(FLS)is proposed for nodes-to-multiobjects tracking path design.This method uses extend kalman filter(EKF)to obtain the target state,and uses FLS to allocate the tracking target of the node according to the heterogeneity of the node,and selects the IAStart algorithm or the flocking algorithm on the basis of the node-target distance.The flocking control algorithm(FCA)completes the trajectory planning.Experimental results show that compared with the flocking algorithm and the shortest path tracking(SPT)along the line between the departure position and the destination position,the algorithm can avoid the low-speed area and track the target quickly,form a stable tracking formation lastly.(3)In order to ease the time consumption and energy consumption of collecting node information by MWSN,a two-stage sink stay-point determination algorithm(Virtual Force Algorithm and Improved LEACH,V-ILEACH)combining improved LEACH(ILEACH)and virtual force algorithm(VFA)is proposed.This method uses ILEACH to select the first-generation cluster head,then uses VFA to adjust the sink's stay position,and finally uses the GA to get the path traversing all sink's stay points.Compared with traveling salesman problem(TSP)and common clustering methods,V-ILEACH shortens sink trajectory and reduces network energy consumption on condition that complet information collection.In addition,a trade-off between data transmission delay and network lifetime can be achieved by dynamically adjusting the number of sink stops.In this thesis,three algorithms:MCOFA,FUZIAStart and V-ILEACH are proposed for three application scenarios of space monitoring to solve the path planning of mobile nodes in 3D space.These three algorithms improve the target detection probability in complex environments,reduce the tracking and formation time of nodes,shorten the sink information collection path and prolong the network life,laying a solid foundation for MWSN to complete the space monitoring task.