Professional Documents
Culture Documents
Distributed Arithmetic Coding For Sources With Hidden Markov Correlation
Distributed Arithmetic Coding For Sources With Hidden Markov Correlation
Abstract—Distributed arithmetic coding (DAC) has been shown R ≥ H(X). In the DAC [4], interval lengths are proportional
to be effective for Slepian-Wolf coding, especially for short data to the modified probabilities (1 − p)γ and pγ , where
blocks. In this letter, we propose to use the DAC to compress
momery-correlated sources. More specifically, the correlation H(X|Y )
arXiv:2101.02336v1 [cs.IT] 7 Jan 2021
iteratively mapped onto sub-intervals of [0, 1), whose lengths be a sequence of observations. A hidden Markov process is
are proportional to (1 − p) and p. The resulting rate is defined by λ = (A, B, π):
A = {aji }: state transition probability matrix, where aji =
This research was supported by Seoul Future Contents Convergence (SFCC) P (st = i|st−1 = j);
Cluster established by Seoul R&BD Program. B = {bi (k)}: observation probability distribution, where
The authors are with the Lab. of ICSP, Dep. of Electronics and Commu-
nications Engineering, Hanyang University, Haengdang-dong, Seongdong-gu, bi (k) = P (zt = k|st = i);
Seoul 133-791, Korea (e-mail: yfang79@gmail.com). π = {πi }: initial state distribution, where πi = P (s1 = i).
2