SNR estimation using extended kalman filter technique for orthogonal frequency division multiplexing (OFDM) system

• Main Author: Ong, Sylvia Ai Ling
• Format: Thesis
• Language: English; English
• Published: 2012
• Subjects: TK Electrical engineering. Electronics Nuclear engineering; TK5101-6720 Telecommunication. Including telegraphy, telephone, radio, radar, television
• Online Access: http://eprints.uthm.edu.my/2468/

Signal to Noise Ratio (SNR) estimation of a received signal is an important and essential information for Orthogonal Frequency Division Multiplexing (OFDM) system. This is because in OFDM system, robustness in frequency selective channels can be achieved using adaptable transmission parameters. Therefore, to reckon these parameters, knowledge of SNR estimates obtained by channel state information is required for optimal performance. The performance of SNR estimation algorithm is contingent on channel estimates obtained through channel estimation schemes. In this project, two estimators which are Least Square (LS) and Minimum Mean Square Error (MMSE) estimators are simulated and analyzed. From the result obtained, LS shows better performance than MMSE in terms of Symbol Error Rate (SER) and Mean Square Error (MSE) via computer simulation. With different number of sub carriers implemented for the system model, 16, 32, 64, the result apparently shows that the SER curve of the estimator with the highest number of sub carriers, 64 is significantly lower compare with the other estimators with sub carriers of 16 and 32. Therefore, a system model which contribute to 64 sub carriers are implemented. However, in case of wireless channels, they possess non linearity where the LS and MMSE, linear estimators yield inefficient results. Therefore, to improve the SNR estimation, an efficient non linear Extended Kalman Filter (EKF) estimation, is implemented into the OFDM system. The EKF estimator outperforms the LS and MMSE estimators in terms of SER and MSE for AWGN channel. The beauty of the estimation is that it can estimate the past, present and future