Error rate performance of NEC receivers of duobinary coded MSK

Duobinary coded minimum shift keying (MSK) makes use of partial response signalling to achieve an efficient power spectrum. Therefore, it is a suitable candidate for applications of digital data transmission where spectrum efficiency is an important consideration. On the other hand this type of signalling, apart from reducing the out-of-band radiation, helps increase the symbol error rate. Nonredundant error correction (NEC) as an error correction technique based upon the principle of phase continuity of a broad class of signals, has been extensively used with MPSK modulation schemes in order to improve the signal-to-noise ratio. A little more complex receiver structure consisting of a bank of differential detectors not only detects errors but it also affords to correct them without any bandwidth or power increase. The results are encouraging in environments which are contaminated with Gaussian noise, intersymbol interference as well as cochannel interference. In this paper, such a receiver of duobinary MSK is proposed. It, utilises the bit sequence detected from the phase difference between alternate signalling intervals, along with the output bit sequence of the conventional differential detector to improve the error rate performance in accordance with a convolutional error-correcting code concept. Hence, the application of NEC to duobinary FSK is proposed, analysed and theoretically evaluated in detail assuming an AWGN environment. Computer simulated results of BER performance of this scheme are given. Moreover, an extension of duobinary MSK for use of multilevel input data is proposed