Merged Source Word Codes for Efficient, High-Speed Entropy Coding
Conference
·
OSTI ID:15005833
We present our work on fast entropy coders for binary messages utilizing only bit shifts and table lookups. To minimize code table size we limit our code lengths with a novel type of variable-to-variable (VV) length code created from source word merging. We refer to these codes as merged codes. With merged codes it is possible to achieve a desired level of efficiency by adjusting the number of bits read from the source at each step. The most efficient merged codes yield a coder with an inefficiency of 0.4%, relative to the Shannon entropy, in the worst case. On one of our test systems a current implementation of coder using merged codes has a throughput of 35 Mbytes/sec.
- Research Organization:
- Lawrence Livermore National Lab., CA (US)
- Sponsoring Organization:
- US Department of Energy (US)
- DOE Contract Number:
- W-7405-ENG-48
- OSTI ID:
- 15005833
- Report Number(s):
- UCRL-JC-150989
- Country of Publication:
- United States
- Language:
- English
Similar Records
Length-Limited Variable-to-Variable Length Codes for High-Performance Entropy Coding
Length-Limited Variable-to-Variable Length Codes for High-Performance Entropy Coding
High-speed maximum-entropy encoder for images
Conference
·
Sun Nov 16 23:00:00 EST 2003
·
OSTI ID:15013735
Length-Limited Variable-to-Variable Length Codes for High-Performance Entropy Coding
Conference
·
Sun Jan 04 23:00:00 EST 2004
·
OSTI ID:15013628
High-speed maximum-entropy encoder for images
Conference
·
Fri Dec 31 23:00:00 EST 1982
·
OSTI ID:6228206