상세 보기
Deep Learning-Based Semantic Image Communication for Satellite IoT Networks: A Survey
- Wondmagegn, Ayalneh Bitew;
- Dinh, Ton That Tam;
- Win, Thwe Thwe;
- Won, Dongwook;
- Cho, Sungrae
SCOPUS
0초록
Satellite IoT networks are expected to support image-intensive applications such as Earth observation, remote sensing, environmental monitoring, and disaster response under stringent bandwidth, latency, energy, and channel constraints. Conventional image transmission schemes are mainly bit-centric and target accurate signal reconstruction, which is often inefficient when task-relevant semantic information is more important than exact pixel fidelity. Recently, deep learning-based semantic communication has emerged as a promising alternative by integrating joint source-channel coding, adaptive feature extraction, and task-aware reconstruction into a unified framework. This paper presents a concise survey of deep learning-based adaptive semantic image communication for satellite IoT networks. We review the limitations of conventional satellite image transmission, summarize the fundamentals of semantic communication and deep joint source-channel coding, and critically compare recent adaptive transmission methods, including SNR-adaptive, bandwidth-adaptive, content-adaptive, entropy-aware, and channel-structured designs. We further discuss quantitative evaluation trends and practical deployment issues, including realistic satellite channel modeling, remote-sensing domain shift, semantic evaluation, onboard efficiency, security, and benchmarking standardization. Finally, we outline future research directions toward reliable and practical satellite semantic communication.
키워드
- 제목
- Deep Learning-Based Semantic Image Communication for Satellite IoT Networks: A Survey
- 저자
- Wondmagegn, Ayalneh Bitew; Dinh, Ton That Tam; Win, Thwe Thwe; Won, Dongwook; Cho, Sungrae
- 발행일
- 2026-07
- 유형
- Conference Paper
- 페이지
- 104 ~ 108
- 언어
- ENG
- 출판사
- IEEE Computer Society
- 발행국가
- 미국
- 분량
- 5 페이지
- ISSN
- E 2165-8536
P 2165-8528