Research On Energy-efficient Data Routing Protocol And Transmission Technology Of Wireless Sensor Networks

Advances in low-power wireless communication, micro-sensor technology, micro-electronic-mechanics system technology, embedded-system technology and computer network technology have enabled the development of small, relatively inexpensive and energy-efficient wireless sensor nodes, which can be connected via wireless network. As a new method of collecting data and an important creative project in computer network area, wireless sensor networks have a variety of applications including military surveillance, environmental monitoring, medical treatments, and commercial business.Design energy-efficient data routing protocols which satisfied the characteristics of networks is a key issue in wireless sensor network researches. Cluster-based routing protocols with better performance in scalability, efficient communication and energy efficiency overthrow other kinds of protocols obviously, and become the major development direction of data routing protocols. However, the problems of unbalanced energy consumption, high-energy communication between cluster heads, and optimal cluster formation in cluster-based routing protocols affect the lifetime performance of network.Based on systematically summarizing the relative works on cluster-based routing protocols, this dissertation focuses on energy consumption distribution in network, low-energy data transmission, and cluster formation in cluster-based routing protocols of wireless sensor networks, and gains several achievements. The major contributions of this dissertation are listed as below:(1) Proposed an energy-balanced cluster range control algorithm (ECRC). In cluster-based multi-hop WSN, because the cluster heads closer to the sink are burdened with heavy relay traffic and consume more energy for relaying data, they will die much faster than the other cluster heads. The energy consumptions among all clusters are not uniformly distributed in network. Based on amount of relaying data, the ECRC algorithm formats nodes into clusters of unequal range, and clusters closer to the sink have smaller ranges than those farther away from the sink. Thus cluster heads closer to the sink can preserve energy for inter-cluster data relaying. Simulation results show that ECRC algorithm leads to more uniform energy dissipation in clusters and achieves an obvious improvement on the network lifetime than EEUC(Energy Efficient Unequal Clustering).(2) Proposed an analysis model for energy consumption of MIMO scheme communication combined cooperative MIMO techniques in cluster-based WSNs. In order to analyze the performance of energy consumption in inter-cluster communication using cooperative MIMO techniques, this proposed model onverts the topology of sensor network with cooperative MIMO scheme into equivalent SISO scheme network in which the optimal system parameters can be easily solved by using optimization programming. Simulation results show that data transmission with cooperative MIMO scheme has outperformed the protocols without MIMO techniques significantly in terms of energy efficient and lifetime of WSNs.(3) Proposed strategies of constructing optimal cluster formations in WSNs. The total energy consumption of a cluster is decided by the intra-cluster communication pattern and topology in a cluster. These strategies solve three problems: the optimal communication radius of node in intra-cluster communication of single-hop and multi-hop; establishment of a minimum energy relay link between cluster head and member nodes based on optimal umber of hops; energy-balanced cluster formation in network with clusterheads and nodes unevenly distributed. Simulation results show that these strategies reduce the energy drainage of cluster significantly and meet the requirements of WSNs for low-energy consumption.(4) Proposed a distributed ID generation algorithm for node in wireless sensor networks based on probability theory and birthday paradox. This distributed ID generation algorithm reveals the relationship between the name space size and the amount of sensor nodes with arbitrarily small probability of two nodes with the same ID. It is distributed executed without information exchange among sensor nodes, and is suitable to context-aware applications which require every node has an unique node ID to identify itself.