Parameter estimation for mW composite sequence with block sparse compressed sensing

Due to the low power spectral density and controllable spectral nulls, m-Walsh (mW) composite sequence spread spectrum communication can be a strong candidate technique for secondary users in cognitive radio (CR). Through detecting the special spectral feature of mW composite sequence spread spectrum signal, we can estimate the carrier frequency and sequence number of Walsh sequence used to compound in mW composite sequence. With these parameters, the complexity of signal receiver can be greatly reduced. So it is very valuable for us to estimate these parameters. But the spectrum access range of the secondary user is so wide, so it requires high sampling rate to realize the spectrum detection. By exploiting the block sparsity property of mW composite sequence spread spectrum signal in the frequency domain, the transmit signal can be reconstructed from sub-Nyquist samples. Then it is available to realize the spectrum detection and parameter estimation for the wideband signal of secondary user. The simulation compares the results of parameter estimation based on BMP, BOMP and BCoSaMP block recovery algorithm in different conditions. The results verify the feasibility of parameter estimation for mW composite sequence by exploiting the block sparse property of signal.