error: u'In this master thesis, I discuss the problem of channel estimation for Long Term Evolution (LTE). LTE uses coherent detection, that requires channel state information. LTE provides training data for channel estimation. To assess the performance of di\x1berent channel estimators, I utilize the LTE link level simulator developed at the Institute of Communications and Radio- Frequency Engineering (INTHFT), Vienna University of Technology. The channel estimators are compared in terms of throughput of the complete system for slowly and rapidly changing channels. The Least Squares (LS) channel estimator with linear interpolation is loosing 2 dB, and the Linear Minimum Mean Square Error (LMMSE) channel estimator 0.5 dB with re- spect to the system with perfect channel knowledge. In order to reduce the complexity, while preserving the performance of the LMMSE channel esti- mator, Approximate Linear Minimum Mean Square Error (ALMMSE) chan- nel estimators are also investigated. I present implementations of such an approximate channel estimator for slowly changing channels and for rapidly changing channels. The ALMMSE estimator for block fading uses the correla- tion between the L closest subcarriers. In the fast fading case, the ALMMSE estimator utilizes the structure of the channel autocorrelation matrix.
Fur- thermore, this thesis shows some simulations, from which the block fading assumption can be proofed valid for velocities up to approximately 20 km/h.
At higher velocities, the subcarriers are not perfectly orthogonal to each other. Thus, Inter Carrier Interference (ICI) occurs. At higher velocities, the Signal to Interference and Noise Ratio (SINR) should be considered in- stead of Signal to Noise Ratio (SNR).'