Journal article
Onboard Spectral Analysis for Low-complexity IoT Devices

Authors/Editors
No matching items found.


Research Areas
No matching items found.

Publication Details
Author list: Grimaldi, Simone
Publication year: 2020
Start page: 43027
End page: 43045
Number of pages: 19
View additional information: View in Web of Science™

Abstract

The lack of coordinated spectrum access for IoT wireless technologies in unlicensed bands creates inefficient spectrum usage and poses growing concerns in several IoT applications. Spectrum awareness becomes then crucial, especially in the presence of strict quality-of-service (QoS) requirements and mission-critical communication. In this work, we propose a lightweight spectral analysis framework designed for strongly resource-constrained devices, which are the norm in IoT deployments. The proposed solution enables model-based reconstruction of the spectrum of single radio-bursts entirely onboard without DFT processing. The spectrum sampling exploits pattern-based frequency sweeping, which enables the spectral analysis of short radio-bursts while minimizing the sampling error induced by non-ideal sensing hardware. We carry out an analysis of the properties of such sweeping patterns, derive useful theoretical error bounds, and explain how to design optimal patterns for radio front-ends with different characteristics. The experimental campaign shows that the proposed solution enables the estimation of central frequency, bandwidth, and spectral shape of signals at runtime by using a strongly hardware-limited radio platform. Finally, we test the potential of the proposed solution in combination with a proactive blacklisting scheme, allowing a substantial improvement in real-time QoS of a radio link under interference.


Projects
No matching items found.

Keywords
No matching items found.

Documents
No matching items found.