Dynamic Task Scheduling In Wireless Sensor Networks

Wireless sensor network (Wireless Sensor Network) is the current concern in the international community which involving a high degree of cross-multidisciplinary, highly integrated knowledge of cutting-edge research areas. In this paper, we mainly research in tasks dynamic scheduling in wireless sensor networks. Compared with other algorithms ant colony algorithm has positive feedback, distributed, and many other advantages, so in this paper, we select ant colony algorithm for the dynamic scheduling algorithm. In Wireless sensor networks, node's energy is constrained, so the task scheduling algorithm not only to ensure the success rate of tasks dynamic scheduling, but also to extend network life. Based on the above two requirements and the ant colony algorithm's disadvantage of local optimization, we improve ant colony algorithm from pheromone dynamic diffuse and heuristic function improving based on energy cost. Simulation test proved that the improving way not only to enhance the success rate of task scheduling, but also to extend the network lifetime. In order to improving success rate further when nodes begin to dead, we put forward a task migration strategy. In fact this strategy achieved the initial purpose.Really, data that tasks need always from others. In view of this task which contact with each other, this paper put forward a CGDC policy which is based on Partition-Reconfiguration arithmetic. CGDC policy consists of three part which is Critical Path Task Gather, Task Dynamic Copy and redundancy task delete. Task clusters come into being by CGDC policy and schedule by ant colony algorithm. Finally compared with other task dynamatic scheduling algorithm that based on DAG, CGDC policy achieved better results.