The Research On Data Fusion Algorithm And Model For Wireless Sensor Networks

Wireless sensor networks which has attracted the academia and industry high value is efficient, flexible, and many other advantages of high research value are very broad potential applications in the military, industrial and agricultural production, urban management, health care, environmental monitoring and disaster relief, and many other fields have important applications. But energy, computing power, storage capacity and communication capacity of wireless sensor networks are severely limited. The research indicates that the 70% of the total energy is consumed in transmission, so WSNs life time should be lengthen and the energy should be reasonably. Data fusion is one of the most important technologies which can reduce energy consumption of WSNs.Aiming at the severe energy and computing resource constraints of wireless sensor networks, based on the analysis and study of data aggregation technology, a data fusion model and a data fusion algorithm for a large scale wireless sensor network are proposed. Under the deep analysis on the relationship between network density, energy cost, advantages and disadvantages of sensor nodes with even and uneven deployed, a scheme with uneven distribution in the network to reduce the whole network energy cost is proposed.A wireless sensor networks data fusion simulator is developed with.NET platform and multi-thread technology, which provides data fusion and related performance indicators simulation analysis. Based on the analysis of the relationship between sensor nodes density and energy cost in network, a low energy cost and fast data fusion tree model (LEFDAT) is proposed in the data communication stage and a distributed distinct data fusion based on fuzzy c-means (DFCMDA) is also proprosed in the data fusion stage. The LEFDAT with sensor nodes uneven distribution provides more than one parent sensor nodes with more residual energy, distance adjustment mechanism and strong transmission capacity for evergy node. To decrease the communication traffic, data dimension is reduced and data redundancy is eliminated after the raw-data is processed by DFCMDA with the limited computing capacity of WSNs. The two stages construct the integret data fusion structure for WSNs.In this thesis, a wireless sensor networks data fusion simulator for data fusion model and algorithm simulation is developed. The simulation result indicates that, under the guarantee of the data precision with a minimum dalay oppositely, the data gathering model and algorithm can decrease the engery cost and lengthen the WSNs life time.