Exponential Strong Converse for Content Identification with Lossy Recovery

Zhou, Lin; Tan, Vincent Y. F.; Yu, Lei; Motani, Mehul

doi:10.48550/arxiv.1702.06649

Cited by 2 publications

(1 citation statement)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In fact, [10] proposed a new and widely applicable converse technique based on functional-entropic duality and reverse hypercontractivity, and it appears that the entire blowing-up lemma business for strong converses [2][3] was merely suboptimal approximations of this approach. After the publication of [10], Zhou-Tan-Yu-Motani [11] and Oohama [12] improved a previous technique of Oohama and claimed that an Op ? nq converse for WAK also follows from that technique.…”

Section: Figure 1 Source Coding With Compressed Side Informationmentioning

confidence: 99%

Dispersion Bound for the Wyner-Ahlswede-Körner Network via Reverse Hypercontractivity on Types

Liu

2018

2018 IEEE International Symposium on Information Theory (ISIT)

View full text Add to dashboard Cite

This paper introduces a new converse machinery for a challenging class of distributed source-type problems (e.g. distributed source coding, common randomness generation, or hypothesis testing with communication constraints), through the example of the Wyner-Ahlswede-Körner network. Using the functional-entropic duality and the reverse hypercontractivity of the transposition semigroup, we lower bound the error probability for each joint type. Then by averaging the error probability over types, we lower bound the c-dispersion (which characterizes the secondorder behavior of the weighted sum of the rates of the two compressors when a nonvanishing error probability is small) as the variance of the gradient of infP U |X tcHpY |U q`IpU ; Xqu with respect to QXY , the per-letter side information and source distribution. In comparison, using standard achievability arguments based on the method of types, we upper-bound the c-dispersion as the variance of cı Y |U pY |U q`ıU;XpU ; Xq, which improves the existing upper bounds but has a gap to the aforementioned lower bound. September 14, 2018 DRAFT

show abstract