| Wireless sensor networks are energy-limited and self-organizing networks, which often operates unattended in harsh environment such as battlefield reconnaissance, fire and so on. With the need of application in real life, the actor node applied to wireless sensor networks has the high probability to fail because of the restriction in energy, storage and computer ability. The whole network could be disabled on account of simultaneous failure of multiple nodes. The fault discussed above may partition the network into disjoined segment and put a negative effect on the performance that mainly refers to environmental data collection. Therefore, there is a need to research on the fault recovery mechanism for wireless sensor networks.In order to resolve the partition problem hampering the data transmission in the network, the paper provides two different recovery strategies with respect to the type of fault:1)A method applying connected dominating set is carried to decide whether the faulty node has destroyed the connectivity of network or not. Then, a neighbor node is picked to replace the faulty node according to the priority mechanism. This scheme proposed for single node fault will eventually achieve the goal of connectivity recovery of network.2) Due to the fact that the neighbor node of fault node might become invalid in the multiple simultaneous failure, the recovery method for this type of fault will be different from above method. A policy based on the density of nodes divides the whole network into many connected sets and sectors. It is required that there should be a path responsible for connecting the connected neighbor sets in a sector. Then, these connected sets can ease the burden of simplification communication between them.The validation of the recovery scheme is carried out through the simulation experiment. The simulation results prove that the proposed scheme have the excellent performance in energy consumption and total time spent in the recovery process. |
PDF Full Text Request