Randomly distributed sensor nodes in the sensor network may receive weak signals.Signal combining technology can be used to enhance the performance of signal perception for the sensor network system.This paper focuses on the combining weight estimation for weak signal combining for the wireless sensor network.Taking the combined signal autocorrelation coefficient as the object function

this paper proposes an eigen value decomposition-based combining weight estimation algorithm.The algorithm does not require to estimate the noise correlation matrix

and can be applied to the non-uniform noise variance environment.The numerical simulation results show that the proposed algorithm outperforms classical eigen-based algorithm using the combined signal SNR or power as the object function

under low SNR environment with non-uniform noise variance for weak sensor nodes signal combining.