Optimal Pilot Structure Design and ANN based CSI Prediction Technique for Channel Estimation in MIMO-OFDM Systems

  • K. Sripal Reddy, D. Praveen Kumar

Abstract

Inter-user interference of MIMO-OFDM is prevalent during the channel estimation. Also, the overhead of CSI feedback increases the CSI estimation cost and affects the medium access channel (MAC) efficiency. This paper presents an Optimal Pilot structure design and ANN based CSI prediction technique of MIMO-OFDM system for accurate channel estimation. This technique utilizes the information of pilot channel properties at non – pilot sub carriers to estimate the channel response. The information given by the pilots of each sub-channel is used as input to the Neural Network (NN) estimator. In Back-Propagation neural networks (BPNN), the corresponding errors are estimated at the output stage and based on these errors weights are updated at the input stage to minimize the errors. ICI cancelation is then performed by sending feedbacks of the corresponding channel estimates to model the current ICI. By MATLB simulation result, it was observe that the presented technique minimizes overhead and enhances the throughput.

Published
2020-03-30
How to Cite
K. Sripal Reddy, D. Praveen Kumar. (2020). Optimal Pilot Structure Design and ANN based CSI Prediction Technique for Channel Estimation in MIMO-OFDM Systems. International Journal of Advanced Science and Technology, 29(3), 13699 - 13706. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/31709
Section
Articles