2014/11/27 by Ahmad Beirami, Beirami, Ahmad, Liling Huang +5
Biochemistry, Genetics and Molecular Biology · Computer Science · Engineering · #Algorithms and Data Compression #Cellular Automata and Applications #DNA and Biological Computing #FOS: Computer and information sciences #Information Theory (cs.IT) #Wireless Communication Security Techniques
paper · pdf · doi:10.48550/arxiv.1411.7607
openalex publication_date 2014/11/27 · openalex created_date 2025/10/27 · openalex updated_date 2026/07/28
This paper investigates the benefits of the side information on the universal\ncompression of sequences from a mixture of K parametric sources. The output\nsequence of the mixture source is chosen from the source i \∈ 1,\…\n,K with a di-dimensional parameter vector at random according to\nprobability vector \w = (w1,\…,wK). The average minimax\nredundancy of the universal compression of a new random sequence of length n\nis derived when the encoder and the decoder have a common side information of\nT sequences generated independently by the mixture source. Necessary and\nsufficient conditions on the distribution \w and the mixture\nparameter dimensions \d = (d1,\…,dK) are determined such that\nthe side information provided by the previous sequences results in a reduction\nin the first-order term of the average codeword length compared with the\nuniversal compression without side information. Further, it is proved that the\noptimal compression with side information corresponds to the clustering of the\nside information sequences from the mixture source. Then, a clustering\ntechnique is presented to better utilize the side information by classifying\nthe data sequences from a mixture source. Finally, the performance of the\nclustering on the universal compression with side information is validated\nusing computer simulations on real network data traces.\n