Slicing enterprise WiFi networks for differentiated IoT service provisioning

Abstract

The increasing number of differentiated IoT services introduced to wireless networks are coming with diverse and sometimes conflicting requirements that brought new challenges for controlling and managing such networks. Network slicing is among the most wanted features of the 5G to address these diverse requirements. However, adopting it for Enterprise WiFi networks is challenging due to the lack of wireless virtualization supports in hardware. This thesis proposes a new slicing solution for enterprise WiFi networks that requires no virtualization support in WiFi access points. Our solution relies on a dynamic user association mechanism that takes into account the amount of network-wide utilization for each slice. To define the association decisions, we formulate an optimization problem to maximize the total throughput of the network with respect to different constraints of each slice. These constraints include the physical limitations for the wireless connections, system integrity limitations related to the WiFi networks, IoT requirements, and the network-wide utilization for each slice. Due to highly dynamics of user and traffic in Enterprise WiFi networks, this high-complexity optimization should be solved in a timely manner. We propose a heuristic based on a many-to one matching game. We adopt the Deferred Acceptance Algorithm to find the stable matching solution while considering the IoT slice requirements and the time required to find proper association for the existing consumers in the network. Simulation results show substantial potential for extending the current work, which can hugely impact the WiFi networks’ social and economical growth. Results show that the proposed solution approximates the optimal results and outperforms the traditional RSSI approaches while guaranteeing the requirements of different IoT slices.

Date	25 Aug 2022
Original language	American English
Awarding Institution	École de technologie supérieure
Supervisor	Kim Khoa Nguyen (Supervisor)