Cluster-oriented Fault Diagnosis And Recovery Mechanism In Mobile Ad Hoc Networks

Mobile ad hoc network is a new wireless network structure of self-organization,non-infrastructure and easy deployment.The future mobile communication network will require higher performance,lower failure rate and higher reliability to carry more complex services.Since once a node fails,there will have an overall impact on the system performance of the network,it has important academic value and significant social and economic benefits to research fault diagnosis and recovery mechanism in order to ensure the reliability of mobile ad hoc networks.In mobile ad hoc networks,clustering network architecture can optimize network bandwidth utilization,reduce the cost of routing maintenance and improve application scalability.The mobility of node is one of the important characteristics of the mobile ad hoc network,as well as the key factors influencing network performance.In this paper,by introducing the group mobility model,we forecast the motion of nodes and calculate the mobility metric on the base of improved Gaussian Markov model.We select best node as head node according to proposed comprehensive assessment model of cluster node.We can achieve a more stable and more reliable cluster network structure by measuring node mobility accurately.On the basis of the stable cluster network structure created by clustering algorithm,there is a further research on cluster-oriented fault diagnosis and recovery mechanism.In this paper,we propose cluster-oriented fault diagnosis strategy based on the invalidation rule of comparative model.The cluster head is responsible for the cluster node fault diagnosis with a diagnosis field in the normal communication data packet.Using above information,we propose an efficient local failure recovery mechanism.When a failure of any of the nodes is detected,we choose an optimum node from its neighbor nodes to replace it.The simulation results show that the proposed mechanism is able to fix network with lower communication cost and shorter time delay to ensure the continued operation of network when communication is disrupted.