Reduced-complexity sequential decoding detector for continuous phase modulation based on green decomposition

The memory of Continuous Phase Modulation (CPM) leads to the high complexity of MLSD. Linear decomposition of CPM signals has been one of the important approaches to reduce the receiver complexity [1][2][3]. Most recently, Wylie Green generalized the Laurent representation, and derived a generalization of linear representation for any variant of CPM [4]. However, the Green decomposition could not reduce the trellis significantly due to the linear components are data-dependent [5]. Based on this, we consider using the sequential decoding algorithm to detect the CPM signals, the complexity of which is independent with the number of states in the trellis [6]. Then we simplify the computation of branch metric of the sequential decoding by utilizing the features that some different branches in the trellis share the same sent signal waveforms.