Dynamic Data Fusion Algorithm Based On Wireless Sensor Network (wsn) Research

As an intelligent information system, Wireless Sensor Network proposes greatpotential application ability in a lot of areas. But because of sensor’s constraint resource,such as, limited power, low bandwidth, low efficient process ability, its design forapplications meets a huge challenge. The key point is low the energy consume while atthe same time improve network lifetime.To large scale wireless sensor network, too much cheap sensors located at a smallarea will lead sensor information redundancy, while we know, directly relay theseinformation will no doubt increase the network transmission. To WSN, its transmissioncost more power compared to computation, so the key point for saving energy is low thenumber of packets. Data fusion is aim to delete the redundant and the ineffective and theunbelievable information, and combines all the information from different source nodes,reduce the data transmission while at the same time decrease the network power costand extend the network lifetime.In this thesis, based on the specific environment detection application, networktopology, energy model, firstly we made a deep research on three different algorithms,Leach、Pegasis, MST, and implemented all of them. By theory analysis and simulation,give a performance evaluation for the three algorithms in prolong the network lifetime.While its result tells, to the data compression and data completely correlated application,the algorithm MST is better than Leach and Pegasis. Later, based on the implementation,considering future wireless sensor network will support more complex applications, itwill bring complex fusion function. We proposed a dynamic data fusion algorithm basedon specific application. The algorithm take full consideration of the character ofapplication and data correlation, optimal the fusion function’s location in the initialplacement stage, and it will transfer the fusion function to other node based on the costfunction defined by user, balance the power cost in the network while prolong itslifetime.