| As a "next-generation broadband wireless mobile communicationsnetwork"organic supplement and peripheral networks,wireless sensor network is a newtype of distributed information retrieval system based on self-organizing wirelessnetwork technology.Setting a large number of tiny sensor nodes with communicationsand computing capabilities, by means of artificially laid, airdrop, artillery and othermethods to the scheduled monitoring of regional,We constitute a "smart"self-monitoring network system, in order to cooperate to perception, acquisition andprocessing of the information of objects which being coveraged by the network.However, due to the sensor node’s battery power, processing power, storage capacityand communication bandwidth and other resources are very limited, and one of thefeatures of the sensor network technology is to quickly process large amounts ofcollected data. So, facing with a large number of signal data and a lot of dataprocessing tasks, we need to fuse the data which being collected by the sensornetworks.This paper starts with the introduction of electromagnetic spectrum monitoringsystem using wireless sensor networks and the sensor nodes design.Then the datafusion technology, one of the key technologies in wireless sensor networks, isresearched.On this basis, we proposed different data fusion algorithm for differentapplications, including LZW lossless data compression algorithm, detect fusionalgorithm, location fusion algorithm and spectrum displaying fusion algorithm.Meanwhile, the characteristics of the hardware platform are taken into account duringthe design of these fusion algorithms. The results show that the designed data fusionalgorithms qutie suit for the system and achieve the desired requirements. Finally, thesimulation results show that the data fusion algorithms had a good performance inimproving the energy efficiency of wireless sensor networks and the data accuracy.Thus these fusion algorithms have some practical values in practical applications. |