Cross-layer Optimization Algorithms In Wireless Sensor Network

Wireless sensor network (WSN) is a kind of distributed self-organizing network, which combines data acquisition, procession and communication. Advances in microelectronics, wireless communications and embedded systems have enabled the creation and development of wireless sensor networks. Due to its special challenges and wide range of applications, WSN is attracting more researchers' interest and becoming one of hottest research areas. A cross-layer strategy that considers MAC layer and network: layer jointly should be adopted to improve the performance of the network due to the fact that both the layers have an great effect on the performance of WSN.First in this dissertation, the conception, architecture and key technologies of WSN are introduced briefly, and energy consumption characteristics of WSN are covered. Because sensor nodes are constrained by energy supply and they are difficult to resupply, it becomes the primary challenge to increase the efficiency of the batteries to maximize the network lifetime. This paper analyzes the energy consumption scale of different modules, for it is the basis for low-power design.Then introduces the current research in low-power protocol and analyzes several algorithms. In LEACH, the clustering structure, data fusion and transmission mechanism can optimize the routing, reduce data redundancy and reduce inter-cluster conflicts. But the selection mode of cluster head is random and the long inter-cluster transmission distance makes consumption relatively large. In PEGASIS, chained construction formed by greedy algorithm shorten the transmitting distance, However, the long chains bring greater delay and larger dependence on network topology. In TEEN , passive sending mechanisms cut down the volume of data to transfer, whereas they weaken the sensitivity of the network.The paper assimilates excellent ideas of the existing protocols and proposes one kind of highly effective cross-layer protocol (OCBC). Its basic ideals: the protocol can optimize the distribution of the cluster-head by partitioning the network with longitudes and latitudes, and decide cluster-head according to the position and remainder energy of the nodes. In this way, it can effectively balance the node energy consumption. Due to the optimal multi-hop algorithm, the expansibility of the network can be improved greatly, and the network load can be relieved by setting the threshold which can adapt the data sending frequency with the network status. The simulation result indicated that the dynamic clustering strategy has more advantages in expanding network and prolonging lifecycle than LEACH protocol.