🤖 AI Summary
This paper investigates the asymptotic exponential optimization of the Type-II error probability in distributed binary hypothesis testing under a two-terminal architecture with decision-making solely at the receiver. Focusing on the “quantization and random binning” scheme proposed by Shimokawa–Han–Amari, we redesign the receiver’s decoding rule: retaining identical rate, communication load, and computational complexity, we replace the original typicality-based guessing mechanism coupled with random binning with a tighter information-spectrum-based decision rule. Theoretical analysis demonstrates that the proposed approach strictly improves the achievable Type-II error exponent, surpassing the fundamental limit of the original scheme—without incurring additional communication or coding overhead.
📝 Abstract
Shimokawa, Han, and Amari proposed a"quantization and binning"scheme for distributed binary hypothesis testing. We propose a simple improvement on the receiver's guessing rule in this scheme. This attains a better exponent of the error probability of the second type.
