Distributed Compression, the Information Bottleneck and Cloud Radio Access Networks: A Unified Information Theoretic View: Plenary talk

This talk focuses on connections between relatively recent notions and variants of the Information Bottleneck and classical information theoretic frameworks such as: Remote Source-Coding; Information Combining; Common Reconstruction; The Wyner-Ahlswede-Korner Problem; The Efficiency of Investment Information; CEO Source Coding under Log-Loss and others. We overview the upink Cloud Radio Access Networks (CRAN) with oblivious processing,which is an attractive model for future wireless systems and highlight the basic connections to distributed Gaussian information bottleneck framework. For this setting, the optimal trade-offs between rates (i.e. complexity) and information (i.e. accuracy) in the discrete and vector Gaussian chemes is determined, taking an information-estimation viewpoint. Further, the performance cost of the simple 'oblivious' universal processing in CRAN systems is exemplified via novel bounding techniques. The concluding overview and outlook addresses in a unified way the dual problem of the privacy funnel and recent observations on the additive noise channels with a helper. Further, connections to the finite block length bottleneck features (related to the Courtade-Kumar conjecture) and entropy complexity measures (rather than mutual-information) are shortly discussed. Some challenging problems are mentioned such as the characterization of the optimal power limited inputs ('features') maximizing the 'accuracy' for the Gaussian information bottleneck, under 'complexity' constraints