Research On Some Energy Efficiency Related Key Problems In Wireless Sensor Networks

With the advancement in wireless communications and electronics, the wireless sensor network which is the new-type network has been the focus of research and application. The wireless sensor network has limited resources, large network density and restricted communication range, the sensor node is easy to fail and has no global ID. As the wireless sensor network has these unique characters, thus the protocols of traditional wireless network aren't fit to wireless sensor networks. Furthermore, the new applications propose new requirements and challenges to wireless sensor networks which include protocols of network layers, energy efficiency problem, safety problem, location problem, middleware protocols and design tools, etc. In these problems, energy efficiency is the most important and kernel basis one.This paper focuses on energy efficiency problem on data link layer and network layer of wireless sensor networks, the research contents are:First of all, the sleep schedule protocol in wireless sensor network with periodic data transmission pattern is proposed to save energy. The energy consumption of a sensor node in its work state is more than node in sleep state; while in periodic data transmission pattern, the node's work state is disciplinarian, so it is necessary to design a sleep schedule protocol to achieve energy efficiency and guarantee transmission delay based on the rule. MAC protocols given in the past improve on S-MAC protocol from different points of view, while the periodic transmission is not taken into consideration. This paper proposes a full MAC algorithm TB-MAC(A Time Based MAC algorithm) which uses the disciplinarian and the record of historical traffic to predict future traffic and to compute the interval of the sleep period. Simulation results show that TB-MAC saves much energy than S-MAC and achieves less transmission delay.Secondly, this paper discusses the problem of how to maximum the network lifetime to gather more data and enhance node's usage ratio in wireless sensor networks. Current research always assume the wireless sensor network has only one sampling frequency, while in real applications, to achieve enough redundancy to let network more robust and to reduce repetition of sampling data, the network usually has several different sampling frequencies. Focusing on this situation, the paper proposes a tree-based and energy balancing data gathering algorithm EBDGA(An Energy Balancing Data Gathering Algorithm). EBDGA centrally constructs a spanning tree connecting all the sensor nodes with the base station as its root, and computes the cost of each added node based on node's sampling frequency. Simulation results show that, even if the base station changes its location or the node density is changing, EBDGA prolongs the lifetime of the first 80% dead node than PEDAP-PA and prolongs the lifetime of the first 75% dead node than PEDAP, EBDGA also gets notable energy balancing ability than both algorithms and thus enhances a majority of sensor nodes' usage ratio. The paper also proposes a popular algorithm EBDGA-A which adapts to the changing of aggregation function. Based on the basis principle of EBDGA, EBDGA-A uses the theory of the sets to compute the real transmission amount, thus to compute the adding cost after each node is added into the spanning tree. Simulation results show that, when the aggregation function changes, EBDGA-A prolongs the lifetime of first 80% dead node than PEDAP-PA when the base is in the center of sensor nodes. When the base is distant from the monitoring zone, EBDGA-A has bad performance.At last, to achieve robust ability and fault tolerance of routing protocols, this paper discusses the problem of designing multi-path routing protocols in wireless sensor networks. The research on location algorithm and the existence of GPS location system make high precision on node's location coming into reality. The paper proposes LBMPR protocol(A Load(Energy) Banlancing Multi-Path Routing protocol) which creates multiple paths from source node to the sink node based on sensor node's geographical location information. LBMPR constructs a rectangle based on source and sink node's location, the resulting multiple paths are contained in this rectangle. The data transmission is equally distributed on different paths, thus to achieve energy balancing ability among sensor nodes and to prolong network's lifetime. Simulation results show that LBMPR can achieve demanded application delay based on the proper selection of parameters a and b, at the same time, it can also achieve better energy balancing and energy efficiency ability than Mesh-2 and Mesh-3.