Lightweight learning algorithms for massive IoT and analysis of their performance - Archive ouverte HAL Access content directly
Reports (Technical Report) Year : 2022

Lightweight learning algorithms for massive IoT and analysis of their performance

Abstract

Executive summary We address in this deliverable parameters optimization as well as the automation of devices configuration in massive IoT LoRaWAN scenarios. The utilization of automation techniques for devices configuration is a crucial evolution in IoT LoRa radio access in the way for network virtualization and automation. The challenges in LoRa radio access networks virtualization consists on partitioning the resources between different services and devices that are connecting in an ALOHA-like access. We will investigate how to perform an automatic orchestration of radio resources between different devices. In particular, we will focus on a reducing the overhead required to ensure a good functioning of the automated devices configuration. We intend to (i) develop strategies enabling IoT devices automated configuration (ii) explore possible strategies enabling to follow a certain goal, such as maximize the energy efficiency, or the reliability, represented here by the Packet delivery ratio, and (iii) prepare a platform for service differentiation of different IoT slices.
Fichier principal
Vignette du fichier
Intelligentsia___D3_1 (1).pdf (1.78 Mo) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-03844765 , version 1 (09-11-2022)

Identifiers

  • HAL Id : hal-03844765 , version 1

Cite

Ghina Dandachi, Yassine Hadjadj-Aoul, Patrick Maillé, Renzo Efrain Navas. Lightweight learning algorithms for massive IoT and analysis of their performance. INRIA Rennes - Bretagne Atlantique and University of Rennes 1, France. 2022. ⟨hal-03844765⟩
129 View
107 Download

Share

Gmail Mastodon Facebook X LinkedIn More