Channel Estimation in Space Time Block Coded MIMO-OFDM System using Genetically Evolved Artificial Neural Network

Abstract

Robust wireless communication systems are expected to have high transmission data rate, system capacity and quality of service. Utilizing multi-carrier modulation technique like orthogonal frequency division multiplexing (OFDM) with multiple-input and multiple-output (MIMO) signal processing provides a promising solution for acquiring high data rates and system capacity respectively. Additionally, the performance of the communication system can be improved by applying space-time block coding (STBC). In such aggregated system channel estimation is essential for accurate reception of the transmitted signals. The main problem associated with state of art channel estimation is the use of slow gradient descent based learning algorithms that suffers from local minima entrapment. Therefore, this paper suggests a genetic algorithm optimized artificial neural network (GA-ANN) based channel estimation technique for STBC-MIMO-OFDM system because of its global optimization property. The performance of the proposed estimator is evaluated using bit error rate (BER) to signal to noise ratio (SNR) graphs and compared with least square (LS) and minimum mean square error (MMSE) algorithms. According to simulation results the GA-ANN performed better than LS and MMSE estimators at higher SNR values but was close to the MMSE algorithm at lower SNR values. As the selected algorithm uses a supervised learning process, a tradeoff is made between bandwidth efficiency and accurate estimation.