Tactile internet over fiber-wireless enhanced hetnets using edge intelligence.

Ebrahimzadeh, Amin (2019). Tactile internet over fiber-wireless enhanced hetnets using edge intelligence. Thèse. Québec, Doctorat en télécommunications, Université du Québec, Institut national de la recherche scientifique, 203 p.

Prévisualisation

PDF Télécharger (23MB) | Prévisualisation

Résumé

Today’s telecommunication networks enable people and devices to exchange a tremendous amount of audiovisual and data content.With the advent of commercially available haptic/tactile sensory and display devices and conventional triple-play (i.e., audio, video, and data) content communication now extends to encompass the real-time exchange of haptic information (i.e., touch and actuation) for the remote control of physical and/or virtual objects through the Internet. This paves the way towards realizing the so-called Tactile Internet. Through human-machine interaction, the Tactile In ternet is expected to convert today’s content delivery networks into skillset/labor delivery networks. The Tactile Internet holds great promise to have a profound socio-economic impact on a broad ar ray of applications in our everyday life, ranging from industry automation and transport systems to healthcare, telesurgery, and education. In most of these industry verticals, very low latency and ultra-high reliability are key for realizing immersive applications such as robotic teleoperation. While necessary, though, the design of ultra-reliable and low-latency converged communication net work infrastructures is not sufficient to unleash the full potential of the Tactile Internet. In this thesis, we put forward the idea that the Tactile Internet may be the harbinger of human augmen tation and human-machine symbiosis envisioned by contemporary and early-day Internet pioneers. In search for synergies between humans and machines/robots, we explore the idea of treating the human as a “member” of a team of intelligent machines rather than keep viewing him as a con ventional “user” while putting a particular focus on developing systems that are human-aware and help advance the human condition, e.g., economic inequality. After describing the Tactile Internet’s human-in-the-loop-centric design principles and haptic communications and traffic models, we elab orate on the development of decentralized cooperative dynamic bandwidth allocation algorithms for end-to-end resource coordination in fiber-wireless (FiWi) access networks. We then use machine learning to decouple haptic feedback from the impact of extensive propagation delays. Next, we propose a context- and self-aware allocation scheme for both physical and digital tasks to coor dinate the automation and augmentation of mutually beneficial human-machine coactivities while spreading ownership of robots across users. In addition to realizing collective context-awareness via task coordination, we aim to exploit local self-awareness in order to improve the energy-delay performance of mobile robots. Further, we study the problem of joint prioritized scheduling and assignment of delay-constrained teleoperation tasks to human operators. Finally, this doctoral the sis investigates the performance gains of cooperative computation offloading for multi-access edge computing (MEC) enabled FiWi enhanced heterogenous networks (HetNets) with capacity-limited backhaul links. After presenting the envisioned two-tier MEC architecture for a FiWi based net working infrastructure, a simple but efficient offloading strategy is proposed, which relies on the flexible trilateral cooperation between end-devices, edge servers, and the remote cloud.

Type de document:	Thèse Thèse
Directeur de mémoire/thèse:	Maier, Martin
Mots-clés libres:	AI; computation offloading; context-awareness; FiWi enhanced HetNets; motion planning; multi-access edge computing; OPEX; self-awareness; teleoperation; task allocation
Centre:	Centre Énergie Matériaux Télécommunications
Date de dépôt:	19 nov. 2021 14:22
Dernière modification:	19 nov. 2021 14:22
URI:	https://espace.inrs.ca/id/eprint/12068

Gestion Actions (Identification requise)

Modifier la notice