Autonomous Interference Mapping for Industrial IoT Networks over Unlicensed Bands

TL;DR

This paper proposes real-time interference detection and classification methods for industrial IoT devices to autonomously create multidimensional interference maps, improving spectrum utilization and network performance in unlicensed bands.

Contribution

It introduces low-complexity interference detection techniques enabling autonomous interference mapping for IIoT networks without specialized hardware.

Findings

Effective real-time interference classification on low-complexity devices

Autonomous multidimensional interference map construction

Potential for enhanced spectrum management in IIoT networks

Abstract

The limited coexistence capabilities of current Internet-of-things (IoT) wireless standards produce inefficient spectrum utilization and mutual performance impairment. The entity of the issue escalates in industrial IoT (IIoT) applications, which instead have stringent quality-of-service requirements and exhibit very-low error tolerance. The constant growth of wireless applications over unlicensed bands mandates then the adoption of dynamic spectrum access techniques, which can greatly benefit from interference mapping over multiple dimensions of the radio space. In this article, the authors analyze the critical role of real-time interference detection and classification mechanisms that rely on IIoT devices only, without the added complexity of specialized hardware. The trade-offs between classification performance and feasibility are analyzed in connection with the implementation on low-complexity IIoT devices. Moreover, the authors explain how to use such mechanisms for enabling IIoT networks to construct and maintain multidimensional interference maps at run-time in an autonomous fashion. Lastly, the authors give an overview of the opportunities and challenges of using interference maps to enhance the performance of IIoT networks under interference.