In order to reduce redundant data in wireless sensor networks（ WSNs）

reduce communication energy consumption and extend network life cycle

a data fusion algorithm

heuristic firefly algorithm-back propagation neural network（ HFABPNN） which combines the heuristic firefly algorithm and BP neural network（ BPNN）algorithm is proposed.The position information of the fireflies is updated by introducing the " historical optimal" position as the heuristic information.At the same time

the weighting factor

together with step factor are counted to make sure its not trapped in the local optimum.Combining the heuristic firefly algorithm and the BP neural network

the optimal solution of the heuristic firefly algorithm is taken as the data fusion of the weight and threshold of the BP neural network.The simulation results show that the algorithm has less data redundancy and higher data fusion precision.