Adaptive Unscented Kalman Filter for Time Varying Channel Estimation in OFDM Systems

This paper presents a comprehensive analysis of channel estimation and symbol detection in orthogonal frequency division multiplexing systems using both the unscented Kalman filter and adaptive unscented Kalman filter techniques. The study investigates the performance of these filters under various modulation schemes, including binary phase shift keying, quadrature phase shift keying, and 16 quadrature amplitude modulation. Using the Rapp model, a simulation study is evaluated for the channel model incorporating non-linear effects, and adaptive filtering parameters are meticulously tuned. Performance metrics such as bit error rate and root mean square error are analyzed as a function of signal-to-noise ratios to compare UKF and AUKF techniques comprehensively. Results indicate that AUKF significantly enhances channel estimation accuracy and reduces BER compared to the traditional UKF, especially under high SNR conditions. The study underscores the potential of adaptive filtering in improving communication reliability in OFDM systems, paving the way for more resilient and efficient wireless communication networks.