Research On Wireless Sensor Network Repair Algorithm

Wireless sensor network has attracted considerable attention under the background of rapid development of wireless network. The application of wireless sensor network covers various aspects of national defense and military, environmental monitoring, city management, medical care, electronic information as well as industrial and agricultural development. Wireless sensor network mainly consists of sensors which are featured as small size and low power, with functions as sensing, computing and transmission. Physical information is able to be transferred into digital one with the help of sensors, which provide precious data for the research and control of various industries.Wireless sensor network has both structures as single-tiered and double-tiered, which means sensor nodes could send data directly to the sink node, or via cluster head node, on the basis of which there exist different kinds of routing protocol accordingly. Steiner tree is used as one of the research tools in the wireless sensor network and has a broad scope of research, including bottle-neck Steiner tree problem, Steiner tree problem for minimal Steiner points, Steiner tree scheduling problem, minimum spanning tree, which serves as important theoretical basis for the research of wireless sensor network.This paper focuses on the restoration algorithm of wireless sensor network. Node failure will result in partition of network, and the restoration work could start from two aspects:1) self-restoration through node movement:in those areas without human intervention, connectivity is restored via self-relocation of sensor nodes.2) relay node placement for restoration:connectivity is restored via relay node. Multiple types of optimization purposes as coverage, connectivity, robustness and energy efficiency should be considered in algorithm design.For the restoration type as self-relocation of sensor nodes, the problem could be concluded as special bottle-neck Steiner tree problem. We proposed an improved approximate algorithm in order to decrease the number of relocated nodes and total travel distance. The algorithm introduces node cooperation, in which different nodes undertakes different relocation strategy according to the node degree and the connectivity status. Those nodes with no connectivity task undertakes more relocation task, thus the total travel distance is lowed.Connectivity restoration could also be achieved through relay node placement, which is concluded as Steiner tree problem with minimal Steiner points, namely SMT-MSP problem. We propose an improved approximate algorithm for this problem with optimization purposes as energy saving and decreasing number of relay nodes. We also proposes new model for representation nodes choosing for each partition with the consideration of both nodes locations and energy left. Classical methods as Minimum spanning tree, Steiner tree problem is used in the decision of relay nodes location, together with the strategy of task sharing with relay nodes for low-energy sensor nodes. In addition to that, the location of relay nodes takes the energy of neighboring sensor node into consideration. All strategies above contribute to decreasing the number of relay nodes and energy saving of entire network.In the end, we take a look into the future work related to our research area.