Efficient Data Transmitting Model in Wireless Sensor Networks

Considering the high similarity among sensor data sequences in wireless sensor networks(WSNs), we propose an efficient data transmitting model—EDTM, which optimizes the data collection scheme utilizing the correlation among sensor nodes. The aim of EDTM is to extend the lifetime of the sensor networks as much as possible, meanwhile guaranteeing the accuracy of data provided to the user which is measured by two parameters, error bounds and confidence. To realize this purpose, EDTM exploits the matching relationship, and if some nodes are matching with each other to a certain degree, it selects the optimal representative nodes Vs to undertake the transportation task. Thus, the measurements of some nodes can be substituted and the energy is conserved. A challenge of the matching algorithm is the time drifting and data lost problem. To obtain the real variation curves of sense attribute, we piecewise fit data sequences using least squares and the fitting error is eliminated in the matching estimation. EDTM computes a parameter IDST based on the sensing curves and we prove that according to the IDST we could analyze the matching degree of any two nodes. At the end, the sensor nodes are distributed into matching sets. We also design the optimal representative node selection scheme including three phases. Every node in the matching sets need to send a cmsg packet to sink in order to inform sink its energy cost information. Sink computes each node's parameter L, which reflects the energy consumption and the influence of energy balance of the node, selects Vs, and informs the transmitting node Vs by sending a tmsg packet. In order to test our approach, an experimental simulation has been done over real-world data and the performance of EDTM has been analyzed. The outcome demonstrates the low error ratio of EDTM, and its effectiveness in decreasing the energy cost.