Multiple antenna system and channel estimation for multiband orthogonal frequency division multiplexing in ultra-wideband systems

Abstract

Multiband Orthogonal Frequency Division Multiplexing (OFDM) has been deployed for practical implementation of low cost and low power Ultra-Wideband (UWB) devices due to its ability to mitigate the narrowband interference and multipath fading effects. In order to achieve high data rates, the deployment of multiple antenna techniques into a UWB system has gained considerable research interest. In a UWB system, both the spatial and multipath diversities exist in UWB system can be exploited via the use of Multiple-Input Multiple-Output (MIMO) antenna system and Space-Time Codes (STC) by leveraging Alamouti scheme. This work shows that MIMO system outperforms Alamouti technique in providing a power combining gain in the receiver. Given that channel estimation for timefrequency multiplexed such as a multiband OFDM system is unexplored largely, this thesis also addresses this issue. In literature, most of the conventional Channel Frequency Response (CFR) estimations require either pre-storing a large matrix or performing real-time matrix inversion. In general, these requirements are prohibitive for practical implementation of UWB devices. In this thesis, the implementation issues of STC-based on Alamouti scheme are investigated for the multiband OFDM system. The research quantifies and analyses existing channel estimation in frequency domain such as Least-Square (LS) and Minimum Mean Square Error (MMSE) techniques. Consequently, low-complexity channel estimation based on Singular Value Decomposition (SVD) technique is developed for multiband OFDM system evaluates under modified Saleh-Valuenzela (S-V) channel modelling represents the realistic wireless indoor environment. This work implies that the SVD technique gives an improvement of 3-5 dB compared to LS technique. Even though SVD performs similarly to MMSE, it managed to reduce significantly the complexity by or to 57.8%.