Research On Collaborative Task Allocation Mechanism For Wireless Sensor Networks Based On MAS

Wireless sensor network(WSN) is related to a variety of subjects, such as computer technology, micro-electromechanical system technology, wireless communication technology, automatic control technology and artificial intelligence. It is a highly cross and more cutting-edge research focus, and it is used widely. The communication ability, energy and resource of a single node in the wireless sensor network is limited, so that the nodes must work together to complete the global task. Therefore, collaborative task allocation in wireless sensor networks becomes one of the key problems to be solved in this field.The collaborative task allocation problem of wireless sensor networks is studied, through the learning of multi-agent system theory and the analysis of wireless sensor networks. Firstly, several commonly used collaborative technologies of wireless sensor networks are introduced. And the collaborative technologies are based on multi-agent theory. Then, a collaborative task allocation mechanism for wireless sensor networks based on MAS is proposed through in-depth analysis of these methods. Finally, the detailed process of the task allocation mechanism is given.In cluster network, a task allocation algorithm for wireless sensor networks based on dynamic alliance is proposed, which combined dynamic alliance with genetic annealing algorithm. Firstly, the nodes in the cluster are classified into three groups, and then the advanced nodes are selected to form an initial alliance. Then, a multiple objectives cost function is constructed through considering the energy consumption, task execution time and remaining energy in the network. The genetic annealing algorithm is adopted to solve the multiple objectives cost function. Finally, the algorithm is simulated on the MATLAB, and the performance of the algorithm from time, energy consumption and loading balance in the network is analyzed. Experimental results show that the task allocation algorithm proposed in this paper has better optimization effect.An efficient task immigration algorithm based on contract net is proposed to deal with the failure problem of the single node in this paper. The algorithm is presented by using multi-attribute utility function to design a synthesis evaluation function adding a threshold, while considering the dependencies of task and the dynamic characteristics in the network when new nodes are joined in. The task of the allocation is handled by the union leader under the situation in which there are no nodes responses to the bidding invitation. The experimental results show that the algorithm can ensure the success of the allocation and reduce the time of the task allocation.