Compressed Sensing Assisted Joint Channel Estimation and Detection for DS-CDMA Uplink

Compressed sensing assisted Joint Channel Estimation and Detection (JCED) for asynchronous Direct-Sequence Code Division Multiple Access (DS-CDMA) system is considered. We present a new dictionary matrix to characterize the sparsity of multiuser signals transmitted over dispersive channels. Using the new dictionary, the amount of non-zero entries in the sparse representation is reduced to $1/Q$ of that in the literature, where $Q$ is the number of symbols in the observation window. The JCED algorithm relying on the new dictionary is at least $\frac{1}{2}\min\{Q,L\}$ times computationally more efficient than a competing solution generalized from existing works, where $L$ denotes the length of signature code. It also outperforms the latter in terms of mean squared error (MSE) and bit-error-rate (BER).