Authors: Apostolos Apostolaras, Kostas Chounos, Leandros Tassiulas and Thanasis Korakis

Conference:  in proceedings of the 18th International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks (WiOPT), Volos, Greece, 15-19 June 2020.

Abstract: We consider the problem for mobile network op-erators (MNOs) of leasing resources, servicing and pricingmobile users, in the context of 5G systems that facilitate theuse of software-defined radio access network (SD-RAN) andnetwork function virtualization (NFV) technologies. We studythe case where the service capability of a MNO cannot satisfythe total users’ demand who are characterized by inelasticbehavior against the servicing rate that they experience. TheMNO addresses this temporal depletion of its resources andacquires dynamically, through leasing, additional resources froman infrastructure provider (InP) to adequately comply with itsmobile users’ demand. We model and analyze the interactionsamong the MNO, and the users, as a Stackelberg game. Tomodel users’ inelastic behavior, we use a sigmoid utility function.Furthermore, we show the optimal pricing decisions when MNO’ssupplying capacity satisfies users’ demand. Given an excesson MNO’s supplying capacity, we employ the generalizedr-Lambert function to determine the optimal pricing. When MNO’ssupplying capacity is not ample, we determine, besides pricing,an approximation of the optimal amount of the additionalresources to purchase, given a leasing cost imposed by the InP. Aninteresting finding shows that the amount of additional resourcesto be purchased can be larger than the MNO’s minimumcapacity gap. Simulation and testbed experimentation validatethe feasibility of the proposed pricing and leasing scheme anddemonstrate its practical application.

Download paper: [Paper]