Minimal length multirate zero ISI filters

A novel adaptive method for eliminating intersymbol interference (ISI) in a non-ideal channel, based on multirate signal processing, is proposed. The data is upsampled and filtered (by an appropriately designed channel-dependent transmitter filter) prior to transmission. The received signal is downsampled at the receiver. Zero ISI is achieved if the channel is known precisely. This method can be applied to multicarrier communication systems, such as discrete multitone transmission (DMT). This method is compared with the cyclic prefix method that is used in DMT systems, and it is found that comparable performance can be achieved, but with significantly reduced receiver complexity.