Search

Spatial correlation is a crucial impairment for practical Multiple Input Multiple Output (MIMO) wireless communication systems. It is important to consider spatial correlation between antennas in actual communication scenario at both sides, i.e. transmit and receive side. Furthermore, we have considered nonorthogonal full rate STBC in MIMO systems equipped with four transmit antennas and four receive antennas in a quasi-static Rayleigh fading channel. In this paper, Bit Error Rate (BER) performance of full rate non-orthogonal Space Time Block code (STBC) is obtained using simulations in MIMO wireless communication systems. Spatial correlation is assumed between antennas at both the transmitter and the receiver sides. For this analysis, we have considered various transmit antenna spacing values of dt = 0.1π, 0.2π, and 0.4π keeping fixed dr = 0.1π and various receive antenna spacing values of dr = 0.1π, 0.2π, and 0.4π keeping fixed dt = 0.1π. The obtained results show that the transmitting diversity is more serious than the receiving diversity in the spatially correlated antenna environment for the specified signal-to-noise ratio. This study will be useful for actual wireless communication implementation where spatial correlation antennas are present in the practical environment of the MIMO wireless communication system.

In this paper, we consider Mt×M system (Mt > M), where Mt and M are the numbers of antennas at the transmitter and receiver, respectively. We select M out of Mt transmit antennas using two different antenna selection schemes. In scheme 1, we select the subset of M antennas out of total (Mt_M) subsets. In the selected subset, the minimum SNR is maximum compared to minimum SNR of all the remaining subsets. In scheme 2, Mt available transmit antennas are divided into Mtg disjoint groups of successive antennas, where Mtg = Mt/N. It means that there are N antennas in each group, where N ≤ M. Further both Mt and M are divisible by N and the total possible combinations of available groups are given as (Mtg_C), where C є {1, …, Mt/N}.Then we select the subset of M antennas out of total (Mtg_C) subsets. In the selected subset, the minimum SNR is maximum compared to minimum SNR of all the remaining subsets. In this scheme, N and C are chosen to meet the requirements of M selected antennas for transmission. After antenna selection, the resulting system will be M×M. We use Minimum Mean Square Error (MMSE) Vertical Bell Laboratories Layered Space Time (VBLAST) detection at the receiver for both the antenna selection schemes. We present MIMO Symbol Error Rate (SER) versus MIMO Symbol SNR using simulations for M-QAM constellations with Rayleigh fading channels. We have compared the performance of the considered systems with prevailing high complexity schemes ML and MMSE Improved VBLAST. The considered scheme 1 with MMSE VBLAST outperforms the prevailing schemes while scheme 2 with MMSE VBLAST provides similar performance in a wide range of SNR compared to prevailing schemes. However, the number of feedback bits used in scheme 2 is less as compared to scheme 1. There is a tradeoff between the SER performance and the number of required feedback bits. As N decreases, scheme 2 performance starts moving towards the performance of scheme 1. Both the systems provide the diversity gain in the fading channel.

A novel point-to-point symbol detection algorithm in multiple input multiple output (MIMO) system is proposed. This algorithm is an augmentation of two popular algorithms, namely, vertical Bell laboratories layered space-time (VBLAST) and maximum a posteriori probability (MAP). Here, layers are distinguished or ordered based on the a posteriori probabilities of output symbols and not on signal-to-noise ratio (SNR). For each layer, a set of a posteriori probabilities is computed for all output symbols using all possible signal constellations. The layer corresponding to the output symbol having minimum a posteriori probability is selected first from the set of a posteriori probabilities for detection by doing the comprehensive search over all the possible signal constellations. Then, the remaining layers are detected by the conventional VBLAST MAP technique. The relationship of MIMO symbol error rate (MIMO SER) versus MIMO symbol SNR is presented using simulations for 16x16 MIMO systems and 16-QAM constellation. The results show that the proposed algorithm outperforms conventional VBLAST MAP and improved VBLAST algorithms.