Blind channel identification for OFDM based on subspace decomposition

A blind identification algorithm based on subspace decomposition for OFDM channels is proposed. Taking the received OFDM signal as an equivalent Single Input-Multiple Output (SIMO) oversampled signal, the channel's parameters are estimated by subspace method. Without any training sequences and periodic pilot signals, blind channel identification for OFDM system is realized. The broadband OFDM mobile communication system usually has a number of subchannels and its channel's response period is less than the OFDM symbol's period, so it is better to divide the whole system into several subsystems, then each of them can be estimated separately with less complication.