Blind channel MMSE equalization with transmultiplexers based on filter bank precoders

Transmultiplexers based on filter bank structure can be used in estimation and equalization of finite impulse response(FIR) channels.In this paper we introduce a new equalization scheme under the FIR filter bank framework.Firstly,we use the correlation matrix to identify the unknown channel,and then equalize the received data with a FIR post-equalizer under the minimum mean square error(MMSE) criterion.It is shown that the new MMSE-based method outperforms zero forcing(ZF) equalizer.It can mitigate the intersymbol interference(ISI) and depress the noise amplification effect simultaneously,and subsequently maximizes the signal to noise ratio of the system at the receiver.We obtain the SNR performance in a closed form.Simulation examples demonstrate its good performance.