Distributed Arithmetic Coding for Memory Sources

Slepian-Wolf problem regards the compression of multiple correlated information sources that do not communicate with each other. In real applications, the information sources such as video sequences are usually with memory where they are interdependent between symbols. Existing researches mainly consider the memoryless sources and employ channel coding to solve the Slepian-Wolf problem. In this paper, we use distributed arithmetic coding instead of channel coding to solve the Slepian-Wolf problem, benefiting the advantages of source coding in memory sources by eliminating redundancy between symbols. The proposed scheme is very competitive comparing to the existing schemes when applied to memory sources. Simulation results show that the proposed scheme has performance gains when applied to the first-order memory sources with the different overlapping factor. We also analyzed the performance applied to the second-order memory sources with different block lengths.