Title :
Asymptotic properties on codeword lengths of an optimal FV code for general sources
Author :
Koga, Hiroki ; Yamamoto, Hirosuke
Author_Institution :
Graduate Sch. of Syst. & Inf. Eng., Univ. of Tsukuba, Ibaraki, Japan
fDate :
4/1/2005 12:00:00 AM
Abstract :
This correspondence is concerned with asymptotic properties on the codeword length of a fixed-to-variable length code (FV code) for a general source {Xn}n=1∞ with a finite or countably infinite alphabet. Suppose that for each n ≥ 1 Xn is encoded to a binary codeword φn(Xn) of length l(φn(Xn)). Letting εn denote the decoding error probability, we consider the following two criteria on FV codes: i) εn = 0 for all n ≥ 1 and ii) lim supn→∞εn ≤ ε for an arbitrarily given ε ∈ [0,1). Under criterion i), we show that, if Xn is encoded by an arbitrary prefix-free FV code asymptotically achieving the entropy, 1/nl(φn(Xn)) - 1/nlog2 1/PXn(Xn) → 0 in probability as n → ∞ under a certain condition, where PXn denotes the probability distribution of Xn. Under criterion ii), we first determine the minimum rate achieved by FV codes. Next, we show that 1/nl(φn(Xn)) of an arbitrary FV code achieving the minimum rate in a certain sense has a property similar to the lossless case.
Keywords :
convergence; entropy codes; error statistics; optimisation; variable length codes; asymptotic optimality; binary codeword length; countable infinite alphabet; distribution convergence; entropy; error probability; fixed-to-variable length code; information spectrum; optimal FV code; Binary sequences; Convergence; Decoding; Entropy; Error probability; Probability distribution; Random variables; Systems engineering and theory; Asymptotic optimality; convergence in distribution; fixed-to-variable length coding; general source; information spectrum;
Journal_Title :
Information Theory, IEEE Transactions on
DOI :
10.1109/TIT.2005.844098