Research On Topology Control Algorithm In Delay Tolerant Wireless Sensor Networks

As the purpose of data collection and its simplicity of deployment, wirelss sensornetwork (WSN) is used widely in every aspect of modern life and work. Yet the sensor node’svolume is so small that make energy supplement become an important aspect that constrictsits quality of service, such as lifetime in network, coverage, connectivity and so on. As aresult, the wsn is kind of delay tolerant network of limited energy. So we need varities ofprotocol that specify topology controlment, sleep scheduling of node to make better use oflimited energy and improve the QOS of wsn.In some wireless sensor network of particular purpose, there is no rigor requirement ofconnectivity, but need for long lifetime in network, so it is necessary for us to schedule sleepoccasion of nodes in network. In this aspect, we can draw lessons from idea of automaton.Through comparative research, we found that there are many similar aspects between wsn andautomaton of "the game of life", such as the cellular update of the automaton can describe thenode state transition of the wsn vividly, so the idea of life game can be used in sleepscheduling of wsn. Then through analysis of connectivity, coverage and lifetime in network,we can discover that there exists compromise among them, which means that we can prolongsome lifetime at the cost of connectivity and coverage.People have put forward many algorithms of topology control recently. All of them havetheir advantages and defects. The most classic clustering topology control algorithm isLEACH protocol which divides the network into clusters, whose data is forwarded to sinknode by its head. The implementation is simple and the idea of cluster is a big innovation inthis field, moreover, the head of each cluster can fuse information collected by its node,reduce the redundant data, thus save energy effectively. However, the LEACH protocol onlyconsider whether a node have been the head in the election, ignore its remaining energy; afterclustering, all of nodes in a cluster begin to work, there is no sleep scheduling mechanism tocontrol their sleeping; and they forward the received data to the head directly, as theconsumption of data forwarding is proportional to the4power of distance, so this kind of ahop will waste lots of energy.To solve defects of LEACH protocol, first, modifying the election strategy of clusterhead, consider the influence of remaining energy to avoid the node whose remaining energy istoo small to be the head of a cluster; second, applying the idea of “life game” to the LEACHprotocol, nodes in a cluster use current state of itself and its neighbors’states to determine their following work, sleep or collect data; then changing the communication mode in thecluster, replace the “only hop” with “multi hop”, nodes in a cluster forward their data betweeneach other till the data reaches the head of a cluster. Through the working mentioned above,the LEACH can make better use of energy, under the premise of ensuring the overall networkconnectivity, at the cost of part of network’s inconnectivity and more delaying of datatransmission, the lifetime will be prolonged, in some occasions, which need long lifetime, thiswill be of effective use.Indeed, the application of the idea of “life game” to the sleep scheduling of wsn is onlyin the exploration and experimental stage, this article is only a try in this aspect. Apart fromthe basic research, there are still many aspects that we do not consider, thus we need tocontinue more explorations and innovation. The simulation experiment mentioned in thisarticle, simulates only basic energy consumption and communication of sensor nodes simplyand randomly, and can not truely reflect working of network and communication mechanism.In the aspects of sensor topology control and sleep scheduling, there are still many problemswaiting for us to research and discuss, to explore and settle them with our great efforts.