Research On Resource Heterogeneity And Energy Efficiency In Wireless Sensor Network

A wireless sensor network (WSN), consisting of a large number of tiny sensor nodes, is a multi-hop self-organizing network system. As it can fuse the information world with the real world, WSN has a wide spectrum of promising applications in many fields, such as valley monitoring, military reconnaissance and industrial control. With the increasing expanding into other potential application prospects, the network size of WSN is growing, and the issue of resources especially energy in WSN receive increasing attention, many theoretical and technical problems remain open. Heterogeneous sensor network is composed of many wireless sensor nodes which are different in power energy, perception ability, communication ability, computing ability and storage ability. Energy heterogeneity is very common in practical applications of wireless sensor networks. This thesis focuses on the research of complex network characteristics in WSN, energy efficiency and network construction of heterogeneous sensor network, clustering algorithm for WSN and time-slot scheduling for MAC protocol of WSN.Firstly, a topology generator is designed based on the analysis of several basic topology models and various statistical characteristics of the complex network. As WSN has a large number of sensor nodes which exists close interaction with each other, there are some characteristics of complex network in WSN. This thesis analyzes the topology characteristics of WSN and its characteristics of complex network in order to provide a theory basis to improve the energy efficiency of WSN by constructing a heterogeneous sensor network with small-world characteristics.As the view of complex network, it has influence on WSN in terms of the small-world characteristics and energy efficiency deploying heterogeneous nodes and super links. For the sake of furthest mitigating the energy consumption of sensor nodes and prolonging the network life simultaneously, minor nodes with heterogeneous power capabilities would be deployed appropriately to undertake data relaying in WSN according to the preferential attachment. Accordingly, an optimal scheme of deploying heterogeneous nodes is presented by using genetic algorithm in order to minimize the average path length of WSN. The simulation results show that the performance of network has evidently improved and the energy efficiency has also increased through optimized deployment of heterogeneous nodes.Aiming at resolving the issue of complex terrain and wide region in actual valley monitoring, a strategy suitable for node-intensive environment is proposed to deploy proper heterogeneous nodes whose energy could be supplemented and construct a three-layer-structured multilevel energy heterogeneous sensor networks. The clustering algorithm of WSN based on correlation, geographical information and residual energy is described. Considering the network cost and wireless coverage, the thesis gives a solution to calculate the number of clusters and decide how many or where heterogeneous nodes should be deployed in WSN.As sensor nodes mainly consume their energy in data wireless communication, a clustering algorithm is proposed based on node energy evaluating (NEE) in order to resolve the problem of the existing distributed clustering algorithm based on residual energy, which usually influenced by man-made factors when run for cluster head, and other reasons such as energy information exchange frequently. NEE algorithm gives an energy evaluation function for sensor nodes to denote the ability parameter of competition for cluster head. At clustering stage, every node decides when it send information including energy and ability parameter to cluster head, and the cluster head manages the set of candidate heads, decides when the new cluster head would be selected for the reason of avoiding energy consumption in premature selection.There exists a problem with the traditional multi-channel MAC protocol which assigns equal-length timeslots to each channel. Therefore an energy-efficient adaptive time-slot scheduling strategy is proposed on the basis of analyzing AMAC and EATA. The weight-load adaptive MAC (WAMAC) can variably adjust a corresponding time-slot length for every node according to its weight-load that weighting the new load and the residual load. Simultaneously, WAMAC gives an approach to calculate time-slot length and weight. This thesis also elaborates the time-slot assignment, time-slot renewal and time-slot order. After that, the energy efficiency of WAMAC is also analyzed from the aspects of node overhearing, idle listening and algorithm complexity.The last part is the conclusion, which as well suggests the further research that should be done in the future.