Improved multi-scale principal component analysis based OFDM channel estimation

High data rate wireless communication system introduces the unavoidable inter symbol interference (ISI), for which orthogonal frequency division multiplexing (OFDM) gives a best solution by utilizing the property of orthogonality among the sub carriers. In addition, estimating the channel characteristics under non-linear time varying random channel environment, is always a tough task for a successful communication of wireless data. By combining the merits of wavelet transform (WT) with principal component analysis (PCA), a signal processing method called improved multi-scale principal component analysis (I-MSPCA) was utilized here to solve the problem of channel estimation in OFDM communication system. Multi level decomposition nature of WT until the frequency range of interest gives a clear way to eliminate the correlated noise component, whereas PCA enables to identify the principal components of the received signal, which in turn make the process of channel estimation easier. General comparative measure, the bit error rate (BER) was calculated and compared to establish the efficacy of the proposed I-MSPCA method.