Information Compression And Routing Protocols In Wireless Sensor Networks

Nowadays, with the development of wireless communication and electronics, a new kind of network has become a hot research area, that is wireless sensor networks. Because of the needs for small size sensor nodes, the power supply of this kind of network is always limited. Thus, the most important factor of designing the protocols for wireless sensor networks is energy efficiency. In this paper, the energy efficiency of wireless sensor networks is discussed in two aspects, information compression and energy efficient routing protocols.Due to the limitation of power supply, avoiding the unnecessary communication among sensors is preferred. One way that can achieve this goal is to adopt distributed algorithms. According to Shannon's theory, joint entropy is enough for transferring correlated sources when they can communicate with each other. But, if distributed compression is needed, the joint entropy can also be sufficient?Distributed source coding theory demonstrates that the joint entropy limit can also be achieved by distributed compression. In this paper, distributed source coding theory is firstly introduced, especially Slepian-Wolf and Wnyer-Ziv's milestone theories. After that a distributed source coding method by using syndromes is presented. This paper explores the decoding process and gives the modification methods. After analyzing the classified decoding errors, the results demonstrate that the proposed methods can effectively improve the decoding performance. This method is also implemented under binary channels. The simulation results show that in order to obtain a better decoding performance, the powerful channel codes such as Turbo code and LDPC code must be adopted.Next, the LDPC code which attracts much attention during these years in channel code field is introduced. And its applications in distributed source coding areas are also presented. A syndrome based method is presented and analyzed in a new view. The simulation does not focus on getting better performance, but on comparing the error probabilities under different code lengths. The results show that the decoding performance drops quickly with the decreasing of code length. It means that LDPC method should better be used in a larger code length circumstance. At the end, a non-binary LDPC method is proposed. As difference from others work, here, a more general source generating method is used. The simulation results show that this method achieves much better decoding performance under lower compression rate.Routing is one of the most important techniques in wireless sensor networks. Because of its robustness, flexibility and energy efficiency, clustered routing protocols attract much attention. This paper does not focus on modifying the protocols, but on adopting other technologies to cooperate with routing protocols to realize energy efficiency. First, by analyzing the energy model of clustered routing protocols, a local SNR aid method is proposed. By choosing suitable inner cluster sensors for data transferring, it avoids consuming the energy on all sensors and save a large amount of energy. The simulation results show that this method can balance the different kinds of energy consumptions, save the power of sensor nodes, prolong network life time and do not affect the target state estimation results. Second, a target tracking method by using the proportion of distance is proposed. In some cases, it is hard to decide the distance between sensor nodes and the target, but the value of most target state estimation result is proportion to the distance between target and sensor nodes. So this proportion can be used to estimate the target position. The simulation results show that the locus of target drawn by this method is close to the original locus, so it can be used estimate the target location.At last, a simple architecture to combine distributed source coding and clustered routing protocols is proposed. The information on cluster head can be used as side information and help decoding the compressed information transferred from inner cluster sensor nodes. The simulation results show that with the cooperation of these two technologies, it effectively realizes the energy efficiency.