To reduce the time consumption in the reconstruction of electrocardiograph（ ECG） signals for the wireless sensor networks（ WSNs）

a dichotomy sparsity adaptive matching pursuit（ DSAMP） algorithm for compressed sensing without knowing sparsity is proposed.The algorithm calculates the stage value for approaching the real sparsity based on dichotomy

and determines the direction of stage switching by the absolute value of the difference in residual norm between two adjacent iterations.Results show that the runtime of DSAMP can be reduced compared with that of sparsity adaptive matching pursuit and is similar to those of subspace pursuit and orthogonal matching pursuit.The peak signal to noise ratio of the DSAMP is improved 6.29 % and 5.43 %

compared with the that of orthogonal matching pursuit and subspace pursuit on average

respectively.