Osman Nuri Ucan (ONU) receivers in partial, response channels

Abstract

In this paper, for the first time in literature trellis coded M-PSK modulated signals are analysed in partial response channels and the Osman Nuri Ucan (ONU)-receiver is introduced. The ONU-receiver is composed of a k unit memory decoder, k/k+1 rated convolutional encoder, cascaded with a PRS channel model. The combined trellis structure of the ONU-receiver reduces the state number from 2/sup k/2/sup /spl nu//2/sup k+1/ resulting from k-unit memory of the precoders, /spl nu/ unit memory of the encoder and (k+1) unit memory of the PRS channel model to only 2/sup k+/spl nu//. Therefore considering the PRS channel together with the classical trellis coded M-PSK does not increase the state number and decoding complexity. Appropriate precoders structures are introduced for M-PSK (M=2, 4, 8). The ONU receiver improves both bandwidth efficiency and error performance with no resulting decoding complexity. As an example, the error event upper bound of the ONU-receiver is analysed for trellis coded 4-PSK having encoder memory /spl nu/=1 and /spl nu/=2. In AWGN and colored noise environment, the error event lower bounds of the considered schemes are derived. An interesting result is that ONU receivers improve the error performance in colored noise environment.