Single carrier frequency division multiple access (SC-FDMA) technique is one of most powerful frequency-division multiple access scheme, which can reduce a peak to average power ratio (PAPR) comparing with the orthogonal frequency division multiplexing (OFDM), and enable to flexible frequency allocation for multiplexing users access with multiple-input multiple-output wireless communication system to achieve reliable and high data rate transmission over broadband wireless channels.
Electromagnetic wave as wireless channel conveys data information from transmitter to receiver side in wireless communication system. The wireless channel is modeled as stochastic, and its causes attenuation in amplitude and phase rotation of the signal. This is also called as multi-path fading. To guarantee the demands and quality of services, pilots-assisted channel estimation (CE) is able to extract the channel state information (CSI) to compensate the multi-path fading. There are many excited channel estimation schemes, which are based on linear regression method, obtain the CSI with utilizing amount of pilot signals to maintain the high accuracy. The balance between the data information and pilots is great challenge for CE. The compressed sensing (CS) could reconstruct CSI applying few random projections with partial pilot signals since the inherent sparsity of wireless channel in time domain, which means the quantity of path in channel is significantly smaller than the pilot signals. Therefore, the CE based on CS can reduce the computational cost and also utilize small number of pilots than conventional linear regression method.
This proposal achieves good performance with reducing complexity in CS based CE using orthogonal matching pursuit algorithm for SCFDMA Systems. We reduced the computational complexity by optimized the matrix vector of measurement matrix. In conventional compressed sensing, CS algorithms are processed one by one on each channel. Consequently, we proposed CS method by using the same measurement matrix for CS algorithm process, subsequently we selected a small number of pilots randomly in the receiver side to obtain low computational complexity. A small number of pilots obtain small size measurement matrix, and it has low complexity.