Clustering Algorithm Based On Slepian-Wolf Theorem For Wireless Sensor Network And Improvement

Combining the micro-sensors and computer network, wireless sensor network has become an emerging technology in the 21st century. It breaks the traditional way of gathering information about the physical world. Wireless sensor networks have a great influence in scientific researches and wide applications in every area of society.In general, the sensor nodes have limited resources, especially energy resources. In large-scale wireless sensor networks, massive data are generated every day. The geographically proximate sensors usually have temporal and spatial correlation. It results in data redundancy. Transmitting and processing these redundant data not only waste energy but also reduce the perception of valid data. Under the circumstances, how to improve the energy efficiency of the network has become a research hot spot.To solve the above problems, this paper mainly focuses on the design of clustering routing protocol (also known as clustering algorithm). The core work is described as follows:Firstly, this paper gives the overview of routing protocol and introduces the advantages of clustering algorithm in wireless sensor network. Through the analysis of different clustering algorithms proposed in last few years, we point out their existed shortcomings. In order to solve the data correlation problem, a local data correlation aware clustering algorithm based on Slepian-Wolf theorem (LDCA) is proposed in this paper. The algorithm comprehensively considers several key factors, such as the correlation, communication distance, residual energy etc. We define the average entropy and connection degree as determined conditions for cluster head selection. Then, a distributed algorithm is designed to improve the network performance and energy efficiency. Simulation results indicate that LDCA clustering algorithm can not only generate a better topology, but also reduce the communication data volume. Besides, the performance on energy balance is better than the other algorithms.Secondly, in order to avoid the extra energy consumption and time overhead during the update process, this paper improves the LDCA algorithm and proposes an energy-efficient clustering algorithm using random update (EECRU). In the clustering update phase, the algorithm combines the random update policy and cluster head rotation scheme to solve the disadvantages mentioned above. In data transmission phase, a sampling rate control method is adopted to improve the data process efficiency and make the sensors smart. Meanwhile, we proposed a clustering update theorem to prove that the random update algorithm can achieve a higher efficiency. After that, the mathematical proof is given. Our simulation results show that EECRU algorithm can achieve a better energy efficiency and prolong the network lifetime more effectively compared to other clustering algorithms.Finally, we give the conclusion of this paper and the emphasis of our future work.