Blind Channel Estimation in Multicellular OFDM Systems

The aim of this paper is to compare the performance of Blind channel Estimation in OFDM with pilot based channel estimation. A blind channel estimation is presented for uncoded OFDM. In this method, four steps is done. The four steps are decision algorithm, channel impulse response estimation, channel frequency response estimation and data recovery. The decision algorithm first makes the primary estimates based on minimum mean square criteria. Then these primary estimates are applied to MMSE channel estimation. At last hard decision block is used to recovered the transmitted signal by setting the optimum thresholds in the constellation regions. From the second OFDM symbol H[2,k] the algorithm switches to blind channel estimation and uses the channel estimates of first symbol is at hand. The first OFDM symbol is assumed by using any pilot based channel estimation. Bit Error Rate(BER) Performance (fade rate=0.015μs and TOFDM = 4μs) of Blind channel estimation(BCE) in OFDM is compared with the known technique such as block type pilot based channel estimation. Normalized Mean Square Error(NMSE) of the 1 iteration(fade rate=0.015μs and TOFDM = 4μs and SNR 20dB) is same for both technique such as block type and blind channel estimation. After that NMSE decreases for BCE when compared to other technique. BER performance for the BCE and other technique is calculated when Doppler frequency increases for the constant Signal to Noise Ratio (SNR) 20 dB.