Abstract:
Orthogonal Frequency Division Multiplexing (OFDM) is an effective way to increase data rate and simplify equalization problem in wireless communications. OFDM systems require an efficient channel estimation procedure to demodulate the received data coherently.
Pilot based techniques have normally been employed for channel estimation in OFDM systems. The obvious drawback associated with pilot based techniques for channel estimation is bandwidth overhead. Blind channel estimation techniques act as alternative to pilot based techniques to increase the spectral efficiency. Second order statistics (SOS) based blind channel estimation methods overcome the slow and local convergence which persist in HOS methods. Among SOS based blind channel estimation methods, noise subspace method has proved to be effective in terms of convergence speed and estimation accuracy. In this dissertation work, we compare SOS based TXK (Tong ,Xu and Kailath) algorithm and noise subspace algorithm for blind channel identification of FIR channels using Single Input and Multiple Output (SIMO) model.
Conventional OFDM systems are based on sufficient cyclic prefix (CP) to facilitate simple receiver implementation, however, at the cost of significant channel utilization loss that may be the over riding constraint for future high speed services. Thus, there exists increasing interest in OFDM systems with less CP or no CP, for which conventional algorithms assuming sufficient CP may not be appropriate. This dissertation work is focused on study of blind channel estimation techniques for OFDM systems with insufficient CP or no CP. The redundancy introduced by oversampling or by virtual carriers (VCs) is exploited in above blind channel estimation techniques that give the OFDM systems the potential to achieve higher channel utilization.
MIMO-OFDM systems can achieve higher data rates over broadband wireless channels. For the purpose of subspace based blind channel estimation for MIMO OFDM, either cyclic prefix (CP) or zero padding (ZP) has been exploited. In this dissertation work, evaluation of both CP and ZP based techniques for channel estimation in MIMO OFDM is performed.
