Source coding, a central concept in information theory, is the study of encoding and decoding data. Depending on the topological structure of the sources, i.e. how the sources are connected with encoders, different rate distortion functions are used. In this thesis two different encoding schemes---distributed and decentralized---are discussed and compared with a benchmark (centralized) coding structure. Specifically, all structures for two and three sources are discussed and a special case for the multi-source (more than three sources) is calculated. This work gives a pathway to characterize the generalized multiterminal source coding systems by finding the difference in the rate distortion limits from the optimal centralized coding system. It is shown that in specific cases, some decentralized systems can achieve the Shannon lower bound in a high resolution regime. / Thesis / Master of Science (MSc)
| Identifer | oai:union.ndltd.org:mcmaster.ca/oai:macsphere.mcmaster.ca:11375/23323 |
| Date | January 2018 |
| Creators | Tu, Xiaolan |
| Contributors | Balakrishnan, Narayanaswamy, Chen, Jun, Mathematics and Statistics |
| Source Sets | McMaster University |
| Language | English |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.0023 seconds