A Unified Framework for Unbiased Non-Coherent Over-the-Air Computation
IEEE Transactions on Communications · 2026 · 74 · 11124-11139
DOI: 10.1109/tcomm.2026.3698903 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Communications, 2026. Verfasst von Martin Dahl und 2 weiteren Autor:innen. Die Arbeit nennt 45 Literaturangaben. Gefördert von: Excellence Center at Linköping – Lund in Information Technology, Swedish Research Council (VR), and the Knut and Alice Wallenberg (KAW) Foundation.
Abstract (Originalsprache)
Over-the-Air Computation (OAC) enables efficient data aggregation in large-scale distributed systems by exploiting the superposition property of wireless multiple-access channels. In contrast to most existing studies on OAC assuming exact channel state information, we consider non-coherent OAC (NC-OAC) where the channel phase is unknown at the transmitters and receiver. A three-step framework for NC-OAC with a mapping between source data and codewords is proposed: 1) Devices encode their data to non-negative codewords; 2) Devices transmit a sequence of symbols with amplitude proportional to their codewords, such that the receiver can estimate the codeword sum. Estimation of the codeword sum is studied under two scenarios of global channel amplitude knowledge: statistical or instantaneous; 3) The estimated codeword sum is decoded to the desired source data sum at the receiver. With the proposed framework, we first study prior work on NC-OAC and map these to the framework. Next, we define and compare the two most commonly (often implicitly) used mappings for NC-OAC: the Affine and the Augmented Affine mappings. Under the constraint of unbiased estimation, we show that with uniformly distributed data and standard channel assumptions, the Augmented Affine mapping exhibits an order of magnitude lower estimation variance than the Affine mapping with both statistical and instantaneous channel knowledge. This result is validated by extensive simulations. Finally, we propose and analyze a new mapping, which demonstrates superior performance over the previous two affine mappings.
Förderung
- Excellence Center at Linköping – Lund in Information Technology
- Swedish Research Council (VR), and the Knut and Alice Wallenberg (KAW) Foundation
Zitieren
APA
Dahl, M., Chen, Z., & Larsson, E. G. (2026). A Unified Framework for Unbiased Non-Coherent Over-the-Air Computation. IEEE Transactions on Communications, 74, 11124-11139. https://doi.org/10.1109/tcomm.2026.3698903
Vancouver
Dahl M, Chen Z, Larsson EG. A Unified Framework for Unbiased Non-Coherent Over-the-Air Computation. IEEE Transactions on Communications. 2026;74:11124-11139. doi:10.1109/tcomm.2026.3698903
BibTeX
@article{Dahl2026,
author = {Dahl, Martin and Chen, Zheng and Larsson, Erik G.},
title = {A Unified Framework for Unbiased Non-Coherent Over-the-Air Computation},
journal = {IEEE Transactions on Communications},
year = {2026},
volume = {74},
pages = {11124--11139},
doi = {10.1109/tcomm.2026.3698903}
}
EUVIMED