Error correction techniques for the general class of correlative encoded continuous phase modulation signals (CECPM)

Abstract

Nonredundant Error Correction (NEC) is a convolutional-code-equivalent error correction technique. Like any other correction technique, it depends on extra parity-like data transmitted along with the source data. Unlike any other technique, however, neither any modification is required at the transmitter nor any adjustment of the channel bandwidth is required to accommodate an increased bit rate. NEC operates through the employment of multiple differential detectors at the receiver in order to extract the parity information, which resides within the source data. Based on this parity information, it applies some syndrome rules to check if any errors exist among the received data. It has been proved to operate well not only under additive white gaussian noise but also under Intersymbol Interference and Cochannel Interference contaminated environments. Correlative Encoded Continuous Phase Modulation (CECPM) is a general class of signals characterised by their constant envelope, small bandwidth occupancy and the substantially small out-of-band radiation. Among the main representatives of this class of signals, there are Tamed Frequency Modulation, Duobinary Coded Minimum Shift Keying and Duoquaternary Frequency Shift Keying. With correlation length extending over two or three symbol periods and symbols being either binary or quaternary, these modulations are theoretically analysed together with NEC under additive white gaussian noise conditions. Additionally, Quadrature Amplitude Modulation (star-QAM) although not a member of the CECPM class of signals, is also proposed with NEC and is simulated under AWGN and/or Rayleigh fading. Monte Carlo simulation results prove the correctness of the theory, that single and double errors can be detected and corrected. They also prove that the proposed error correction technique is useful even to signals whose data symbols are not independent with each other. The evaluation results indicate significant performance improvements over "conventional" differential detection.