Research On Reprogramming In Wireless Sensor Networks

Network Reprogramming is essential to update or reconfigure the function of nodes in wireless sensor networks. To solve the problems of current reprogramming methods, we proposed multi-source reprogramming scheme and target-nodes reprogramming scheme based on the Deluge, which can balance the network energy consumption and reprogram the code image of target nodes.Through deploying two or more source nodes in a network, the multi-source reprogramming scheme gets a better performance with the knowledge of hops and link quality. Moreover, five placement methods of multiple source nodes are carefully examined, including random, maximum, randomly in MDS (minimal dominating set), maximum in MDS, segmentation. In the real applications with TelosB, the results confirmed that multi-source reprogramming scheme achieved substantial improvement in both completion time and energy consumption. Compared with single-source reprogramming scheme, the completion time of multi-source reprogramming scheme is averagely reduced20.4%, the packets transmitted during reprogramming is averagely reduced23.5%. In the simulation experiments, we evaluated the performance of five placement methods. The results showed the segmentation was the most efficient in grid network. Compared with other four placement methods, the completion time of segmentation is averagely reduced38.3%, and the energy consumption is averagely reduced18.1%. Maximum in MDS worked best in random networks. Compared with other four placement methods, the completion time of maximum in MDS is averagely reduced7.4%, and the energy consumption is averagely reduced16.2%. Then we proposed a target-nodes reprogramming to update the function of nodes in target area, and an effective route algorithm based on GPSR to transmit code image. The route algorithm takes residual energy, link quality and completion time into consideration for a better comprehensive performance. Simulation experiments showed target-nodes reprogramming scheme can optimize the completion time and the lifetime of network effectively.The reprogramming schemes proposed in this paper can be applied to large-scale wireless sensor networks. The results of experiments showed multi-source reprogramming scheme can update the nodes in large-scale networks effectively, and target-nodes reprogramming scheme can update the nodes in special area if those nodes should perform different task.