Deep Joint Source-Channel Coding for Wireless Image Transmission
- Submitting institution
-
The University of Essex
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 1451
- Type
- D - Journal article
- DOI
-
10.1109/tccn.2019.2919300
- Title of journal
- IEEE Transactions on Cognitive Communications and Networking
- Article number
- -
- First page
- 567
- Volume
- 5
- Issue
- 3
- ISSN
- 2332-7731
- Open access status
- Deposit exception
- Month of publication
- May
- Year of publication
- 2019
- URL
-
-
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
2
- Research group(s)
-
C - Communications and Networking (Comms)
- Citation count
- 19
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This seminal work departs from the conventional separation based design of source and channel coding methods and introduces a novel joint source channel coding algorithm for wireless image transmission. Importantly this was the first use of deep models and the autoencoder framework for encoding/decoding function optimization. Extensive evaluation on high-resolution images shows that this method is significantly more robust to channel variations than state-of-the-art approaches and eliminates the "cliff-effect", a long-standing problem in digital communications. This highly-cited work is referenced by renowned international researchers (Y.Polyanskiy-MIT, A.Goldsmith-Stanford, O.Simeone-King's College, F.Fekri-GaTech) and has triggered a wave of similar papers.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -