In the recent years, we have witnessed an increased interest in space-time coding, because of its effectiveness in overcoming the impairments due to the signal fading phenomenon. A large number of such codes require channel knowledge for decoding. Therefore, research in the area of channel estimation for space-time coding is very important.; This thesis deals with the application of the Kalman Filter (KF) to channel estimation in space-time coded systems. For the purpose of deriving such an estimator, we model the multiple antenna channel as a first order Autoregressive process. As test cases, we employ the Alamouti's code as well as a rate ¾ orthogonal code for 3 transmit antennas. We consider a single antenna receiver, with different decoders: the joint optimal, the QR decision feedback and the symbol by symbol decoders.; Through computer simulations, the effect of channel estimation on the system's performance is investigated. With known channel (perfect estimation), the results show that the joint optimal decoder achieves the best performance, followed by the QR decision feedback and the symbol by symbol decoders, respectively. Also, simulation results show that the tracking ability of the KF in its predictive form deteriorates as the Doppler rate increases. When the channels are estimated by the KF, the simulations show that the three decoders yield essentially the same performance, because of estimation errors. The same behavior for the three decoders, with channel estimation, is observed for different modulation formats (BPSK and QPSK) and orthogonal codes. Finally, the numerical evaluation of the error probability for the joint optimal and symbol by symbol decoders, by the use of the residue theorem, shows similar results, backing up the computer simulations. Therefore, to enjoy the benefits of improved decoders, an efficient channel estimator should be also used.
展开▼
