Date of Award
2026
Document Type
Thesis
Degree Name
Master of Science in Engineering (MSE)
Department
Electrical and Computer Engineering
Committee Chair
Laurie Joiner
Committee Member
David Pan
Committee Member
Adam Panagos
Research Advisor
Laurie Joiner
Subject(s)
Wireless communication systems, Image transmission, Deep learning (Machine learning), Combined source channel coding
Abstract
The growing use of AI-enabled applications has increased the demand for adaptive, efficient, and low-latency image transmission of large data volumes. This has created challenges for conventional separate source-channel coding that focuses on reliable bit sequence transmission without considering the contents or communication objectives. Deep joint source-channel coding (DeepJSCC) potentially solves this by combining source compression and error correction into a unified process. This thesis focuses on developing adaptive and computationally lightweight DeepJSCC-based transceivers for image transmission across heterogeneous channels. The framework further investigates DeepJSCC’s performance in OFDM with robust channel estimation and importance-based subcarrier allocation strategies. Comprehensive evaluation using Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM) across different channels and transmission conditions prove DeepJSCC’s ability in adaptation, efficiency, and robustness, establishing it as a promising alternative to traditional separation-based algorithms and a potential solution for next-generation wireless communication.
Recommended Citation
Suvro, Sudipto Das, "Deep joint source-channel coding for semantic-aware adaptive wireless image transmission" (2026). Theses. 848.
https://louis.uah.edu/uah-theses/848
Presentation