scispace - formally typeset
Search or ask a question

Showing papers by "Bruno de Sousa published in 2023"



Proceedings ArticleDOI
08 Jan 2023
TL;DR: In this paper , a multicommodity flow heuristic, based on a Mixed Integer Linear Programming (MILP) formulation, is implemented in the ONOS controller to enable a fair utilization of links in a topology replicating a real-world scenario.
Abstract: The utilisation of resources from a fairness perspective has been an extensively researched topic, either to enhance the quality of services or to allow spectrum sharing, among others. One method to define forwarding routes is through flow algorithms, which aim to find optimal paths between nodes where capacity constraints are of interest. SALEM is an approach to enable fairness in services competing for resources in wireless mesh environments. SALEM implements a multicommodity flow heuristic, based on a Mixed Integer Linear Programming (MILP) formulation, to consider fairness in reliability, delay and energy objectives and is constrained by bandwidth and link usage criteria. SALEM is implemented in the ONOS controller to enable a fair utilisation of links in a topology replicating a real-world scenario. Within a time complexity that is directly related to the topology, the number of users and available services, the results of SALEM demonstrate its feasibility to run in production-ready networks with minimal overhead.