UKF-Based Iterative Channel Estimation Using Two-Dimensional Block Spread Coding for Uplink Transmission in Multicarrier CDMA Networks

Future broadband wireless applications must effectively work under highly dynamic and strong multipath channel conditions, where severe time- and frequency-selective fading effects will coexist. Although some advanced techniques, such as 2-D block spread coding and antenna diversity techniques, can be applied to the code-division multiple-access (CDMA) networks to improve the performance, an accurate estimation of the channel is normally required. Conventional channel estimators (CE) such as the minimum-mean-square-error (MMSE)-based CE cannot effectively track the channel when the time-selective fading of the channel becomes severe. The performance will even be deteriorated in a multiuser environment, where multiple-access interference (MAI) exists. Therefore, in this paper, we demonstrate the challenges to CE in 2-D block spread coding networks through analytical derivation and then propose an unscented Kalman-filter (UKF)-based iterative CE for uplink transmission in multicarrier CDMA networks. By jointly working with the MMSE-CE, our proposed UKF-based approach can effectively track the dynamic changes in the time-selective fading channel and compensate for the estimation lag caused by the Doppler effects. Our scheme can also effectively mitigate the impact of the residual MAI, which resulted from the detection process. Simulation analysis demonstrates the good estimation performance and the quick convergence of the proposed MMSE-UKF CE, particularly under fast-varying channel conditions.