Date of Award
5-1-2026
Document Type
Dissertation
Degree Name
Ph.D. in Engineering Science
First Advisor
Lei Cao
Second Advisor
Hamid Bahrami
Third Advisor
Ramanarayanan "Vish" Viswanathan
School
University of Mississippi
Relational Format
dissertation/thesis
Abstract
Continuous Phase Modulation (CPM) is an excellent digital modulation scheme recognized for its constant envelope property, continuous phase transitions, and inherent memory. These characteristics enable excellent spectral and power efficiency, allow compatibility with nonlinear amplifiers, and make CPM particularly attractive for power- and bandwidth-constrained applications such as satellite and deep space communications. Serially Concatenated Continuous Phase Modulation (SCCPM) with iterative decoding further improves error performance, and its flexibility enables it to evolve alongside state-of-the-art channel coding technologies and show significant potential for performance enhancement.
This dissertation investigates SCCPM systems with binary Markov and Hidden Markov Model (HMM) data sources and develops iterative decoding algorithms that exploit source statistics. First, for convolutional code (CC)-coded SCCPM systems, we design unidirectional and bi-directional modifications of the Soft-Output Viterbi Algorithm (SOVA) to incorporate source memory into the outer code decoding process. The CC-coded SCCPM architecture consists of a recursive systematic convolutional (RSC) code as the outer code concatenated with an ??-ary CPM as the inner code, forming an iterative decoding structure in which SOVA operates as the soft-input soft-output (SISO) decoder for both components. The proposed modified RSC SOVA decoders combine soft information derived from source statistics with the extrinsic information from the CPM SOVA. This integrated soft information is used as a refined a priori input to the RSC decoder, and the updated extrinsic information is iteratively exchanged between the CPM and RSC decoders. In addition, the impact of signal-to-noise ratio mismatch (SNRM) on the SOVA decoders in CC-coded SCCPM is also investigated.
Second, for LDPC-coded SCCPM systems, we extend the bi-directional modification framework to develop Markov and HMM-based source decoders within the LDPC sum-product algorithm (SPA). The modified SPA decoder jointly processes soft information from the systematic variable nodes and source statistics to generate additional soft information, which is fed back to the variable nodes to enhance decoding performance.
Extensive simulation results demonstrate that the proposed modified outer code decoders, including both SOVA and LDPC, effectively exploit source memory, yielding substantial improvements in error performance and accelerating decoding convergence across a wide range of SCCPM configurations.
Recommended Citation
Guo, Zhijie, "Iterative Decoding for SCCPM with Markov and Hidden Markov Sources" (2026). Electronic Theses and Dissertations. 8823.
https://egrove.olemiss.edu/etd/8823