Channel Modeling and Estimation For Wireless Communication Systems Using a Time-Frequency Approach

Broadband wireless communication is a very fast growing communication area. Multicarrier modulation techniques like Orthogonal Frequency Division Multiplexing (OFDM Biorthogonal Frequency Division Multiplexing (BFDM), Pulse Shaping (PS) and Multi-Carrier Spread Spectrum (MCSS) have recently been introduced as robust techniques against intersymbol interference (ISI) and noise, compared to single carrier communication systems over fast fading multipath communication channels. Therefore, multicarrier modulation techniques have been considered as a candidate for new generation, high data rate broadband wireless communication systems and have been adopted as the related standards. Several examples are the European digital audio broadcasting (DAB) and digital video broadcasting (DVB), the IEEE standands for wireless local area networks (WLAN), 802.11a, and wireless metropolitan area networks (WMAN), 802.16a. However, Doppler frequency shifts, phase offset, local oscillator frequency shifts, and multi-path fading severely degrade the performance of multicarrier communication systems. For fast-varying channels, especially in mobile systems, large fluctuations of the channel parameters are expected between consecutive transmit symbols, because channel state of wireless communication systems may change rapidly. Therefore, the estimation of channel parameters and channel equalization at the receiver becomes essential. Ultimately, introducing new and successful approaches for the representation and modelling of signals and systems in multicarrier communication systems will trigger an important progress in communication engineering. In fading channels with very high mobility, the time variation of channel over an OFDM symbol period results in a loss of subchannel orthogonality which leads to inter-carrier interference (ICI). Receivers based on conventional estimation techniques that assume time-invariant channel for one OFDM symbol has error floor for high mobility cases. In this thesis, two-dimensional (2D) pilot-symbol assisted channel estimation for wireless OFDM systems is presented. This linear interpolation algorithm has the advantage of minimizing the system complexity and processing delay while providing a good approximation to real mobile environment. The performance of the 2D frequency domain estimation algorithm is compared to coherent modulation with perfect channel estimation as well as other conventional methods, and investigated under different detection techniques. Simulation results show that the performance of proposed channel estimation algorithm outperforms the conventional channel estimation algorithms and gives very close results to the known channel state case for high mobility environments.