Universal Quadratic Lower Bounds on Source Coding Error Exponents

Cheng Chang, Anant Sahai
2007 2007 41st Annual Conference on Information Sciences and Systems  
We consider the problem of block-size selection to achieve a desired probability of error for universal source coding. While Baron, et al in [1], [9] studied this question for rates in the vicinity of entropy for known distributions using central-limit-theorem techniques, we are interested in all rates for unknown distributions and use error-exponent techniques. By adapting a technique of Gallager from the exercises of [7], we derive a universal lower bound to the source-coding error exponent
more » ... ng error exponent that depends only on the alphabet size and is quadratic in the gap to entropy.
doi:10.1109/ciss.2007.4298398 dblp:conf/ciss/ChangS07 fatcat:o3kxbec4nbetrmwiibhicg7xle