NOMA Computation Over Multi-Access Channels for Multimodal Sensing

TL;DR

This paper introduces an improved eigenvector-based method for NOMA computation over multi-access channels, significantly reducing MSE in multimodal sensor networks and enhancing performance with more subcarriers and sensors.

Contribution

It develops a novel eigenvector decomposition approach for NOMA-CoMAC, achieving lower MSE and better scalability for multimodal sensor networks.

Findings

MSE reduced by approximately 0.7 at E_b/N_o=1 dB compared to average sum-channel method.

Performance gain increases with more subcarriers and sensors due to diversity gain.

Proposed scheme is highly suitable for next-generation multimodal sensor networks.

Abstract

An improved mean squared error (MSE) minimization solution based on eigenvector decomposition approach is conceived for wideband non-orthogonal multiple-access based computation over multi-access channel (NOMA-CoMAC) framework. This work aims at further developing NOMA-CoMAC for next-generation multimodal sensor networks, where a multimodal sensor monitors several environmental parameters such as temperature, pollution, humidity, or pressure. We demonstrate that our proposed scheme achieves an MSE value approximately 0.7 lower at E_b/N_o = 1 dB in comparison to that for the average sum-channel based method. Moreover, the MSE performance gain of our proposed solution increases even more for larger values of subcarriers and sensor nodes due to the benefit of the diversity gain. This, in return, suggests that our proposed scheme is eminently suitable for multimodal sensor networks.