We derive finite-blocklength bounds on the minimum achievable energy per bit over a Gaussian unsourced multiple access (UMA) channel in the presence of heterogeneous path-loss conditions. We consider a setting in which the path loss is known to the users, which enables the use of location-based codebook partitioning [\c{C}akmak et al., 2025]. Through numerical simulations and a large-system analysis based on the replica method, we quantify the performance gain of this strategy relative to the conventional UMA approach in which all users employ a common codebook.
Cite
@article{arxiv.2604.20643,
title = {Minimum Energy per Bit of Unsourced Multiple Access with Location-Based Codebook Partitioning},
author = {Deekshith Pathayappilly Krishnan and Kaan Okumus and Khac-Hoang Ngo and Giuseppe Durisi},
journal= {arXiv preprint arXiv:2604.20643},
year = {2026}
}
Comments
6 pages, 1 figure; accepted for presentation at ISIT 2026