Multiple Input Multiple Output-Orthogonal Frequency Division Multiplexing (MIMO-OFDM) is currently one of the key technologies to continue increasing the data rates in wireless communications systems. One of the limitations of MIMO-OFDM is that in order to improve the performance, the number of antennas in the transmitter or receiver is increased. But this increase implies higher power consumption and more electronic complexity. The combination of MIMO-OFDM with Electronically Steerable Passive Array Radiator (ESPAR) antenna has been recently proposed as a radio frequency signal processing solution to overcome this limitation. This combination can improve the bit error rate performance and obtain additional diversity gain without increasing the hardware complexity. However, due to the large size of the channel matrix, the computational cost required for the detection process using Vertical-Bell Laboratories Layered Space-Time (V-BLAST) detection is very high to be implemented in practice.
Two approaches to reduce the computational cost of the detection process of MIMO-OFDM with ESPAR-antenna-assisted receiver are presented. First a Minimum Mean Square Error sparse-Sorted QR Decomposition (MMSE sparse-SQRD) detection algorithm, which is based on the MMSE-SQRD algorithm, is introduced. This algorithm exploits the sparse structure of the channel matrix to reduce the computational cost of the detection process. The computational cost analysis and simulation results are presented, and show that this algorithm achieves an important reduction of the average computational cost required for the detection process. But when compared with a conventional MIMO-OFDM detection scheme, the computational cost is still higher. Second, in order to further reduce the computational cost another solution that combines the MMSE Sparse-SQRD algorithm with a sub-matrix division of the channel matrix is also proposed for the detection process. The computational cost analysis and simulation results of several schemes with different sub-matrix sizes are presented. Using this proposed scheme the computational cost required for the detection process of MIMO-OFDM with ESPAR-antenna-assisted receiver is reduced to about 0.002 % of the original when V-BLAST detection is used. Also this proposed scheme achieves a complexity that is similar to the conventional MIMO-OFDM detection schemes.