Construction of Efficient and Error Resistant Reversible Variable Length Codes

Project: National Science and Technology CouncilNational Science and Technology Council Academic Grants

Project Details


Reversible variable length code (RVLC) is one kind of variable length code (VLC). RVLC can be decoded instantaneously both in the forward and backward directions. Upon detecting an errors and jumping to the next Resync-Marker, the decoder can start decoding backwards thereby giving an opportunity to recover the packet as much as possible. In order to recover the transmission error, RVLC can be included in MPEG-4 and H.263 video coding standards. When a bit error occurs in a RVLC bit stream, it is possible that the bit error can be detected or synchronized. The codeword length of error detection and error synchronization affects the decoding time and the data quality of receiver. Most RVLC construction algorithms proposed in the literature aim to minimize the average codeword length. In this project, we try to combine the RVLC construction rules and error resistant indices in the graph model for designing more efficient and error resistant RVLC. In this two year project, “Construction of Efficient and Error Resistant Reversible Variable-Length Codes”, this contains the following two parts: The first year, we will construct the RVLC graph model, and then combine the RVLC construction rules and error resistant indices in the model. The second year, we will design several RVLC construction algorithms, including more efficiency RVLC, high error resistant capability RVLC and the hybrid one based on the graph model. We will also compare the coding efficiency and error resistant capability of our construction algorithms with others in the literature

Project IDs

Project ID:PB9808-2402
External Project ID:NSC98-2221-E182-031
Effective start/end date01/08/0931/07/10


  • Reversible Variable Length Code
  • Error Resistant
  • Error Detection
  • ErrorSynchronization
  • Graph Model


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.