Average-Case Optimized Buffered Decoders

Abstract

In many error-correction coding systems, the required decoding time is a random variable which depends on the noise severity. When there are few errors, the decoder has a relatively easy time, but when there are many errors, the decoder must work much harder. If the "many errors" situation is relatively rare, the average decoding time may be much less than the maximum decoding time. If this is so, it will be possible to increase the decoder's effective speed dramatically, through the use of a buffered decoder architecture.

Additional details