Information Flow Decomposition in Feedback Systems: Linear Time-Invariant Systems with Gaussian Channels
Bertrand Wechsler, Dan Eilat, Nicolas Limal
TL;DR
This paper explores how information flow in feedback systems with LTI components and Gaussian noise can be decomposed and characterized, confirming a conservation law of information flow.
Contribution
It extends a previous information identity to LTI feedback systems with Gaussian channels, linking the decomposition to sensitivity functions and verifying the conservation law.
Findings
Decomposition characterized by sensitivity functions
Conservation law of information flow verified in LTI systems
Quantitative analysis of information flow in noisy feedback systems
Abstract
In our companion paper [1], an information identity decomposition has been derived, which can be interpreted as a law of conservation of information flows in feedback systems. In this paper, we further investigate this decomposition result when specified to linear time-invariant(LTI) systems connected with additive white Gaussian noise(AWGN) channels. It is shown that the quantities in the decomposition are characterized in sensitivity function and the law of conservation is verified.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsNeural Networks and Applications · Quantum Computing Algorithms and Architecture · Distributed Sensor Networks and Detection Algorithms