NeTS: Small: Optimized Mobile Data Off-loading Architectures and Mechanisms

Yale University, New Haven CT

Investigators

Abstract

The unprecedented growth of mobile data traffic, that is expected to surpass 25 exabytes/month in 2020, poses major technical and economic challenges to the cellular operators, and necessitates rapid capacity increase. Several network expansion methods based on technology upgrades and spectrum acquisitions, are costly and more long-term approaches. There is a need for alternative solutions, and mobile data offloading to Wi-Fi networks appears as one of the most promising. Despite several recent initiatives, the research community still lacks a systematic approach for achieving a seamless integration of Wi-Fi networks and for ensuring the maximum possible offloading benefits. This project will fill this gap by proposing, analyzing, and validating a techno-economic framework that enables optimized and efficient mobile data offloading solutions. First, it will analyze how an operator can deploy a Wi-Fi overlay network (large-scale, long-term optimization), and how it can be used to offload cellular traffic in an efficient fashion (dynamic, real-time optimization). Then, it will study how operators can outsource offloading by leasing dormant user-owned network resources, such as residential Wi-Fi access points, and by cooperating with Wi-Fi sharing communities or Wi-Fi operators. Such approaches not only enable on-the-spot and on-demand offloading, but also promote sustainable networking solutions and create opportunities for novel business models in this emerging mobile ecosystem. The proposed plan lies at the nexus of network optimization and network economics, and employs convex, stochastic and discrete optimization in conjunction with game theory and auction theory tools. In particular, the project comprises the following two Thrusts: (I) Wi-Fi Access Point Deployment, Pricing Strategies, and Traffic Offloading Policies, and (II) Wi-Fi Spare Capacity Markets and Mechanisms. Thrust I will study network policies for overlaying Wi-Fi auxiliary networks in cellular networks. These policies will be jointly designed with pricing schemes for charging the users who offload mobile data. This approach departs significantly from the current practice of gratis offloading services, and is expected to bring significant economic benefits for the mobile network operators (MNOs). Moreover, the project will study dynamic (online) policies that adapt on traffic variations and optimize, for a given network infrastructure, the offloading benefits while taking into account the users? requirements. The theoretical analysis in this Thrust involves challenging optimization problems, which are often NP-hard, and require sophisticated and/or lightweight solution algorithms. The second Thrust will study how the operators can effectively lease dormant user-owned Wi-Fi networks for offloading purposes. This study refers to decentralized architectures as the network infrastructure is owned by different entities, and often to user-initiated offloading schemes where the mobile users decide when to offload their traffic. Various scenarios will be analyzed based on whether the leased networks are owned by independent individual users, by Wi-Fi operators, or Wi-Fi sharing communities. Each case raises unique issues as both the required offloading mechanisms, and the incentive design problems are different. Broader Impacts: This project is motivated by current challenges of the telecommunications industry, builds upon real data sets, and is supported by extensive testbed experimentation. The proposed research, incorporating communication system design, network optimization, and network economics, will have broad impact on science, education, and communication markets. First, it will increase substantially the cellular capacity by optimized Wi-Fi networks and offloading mechanisms. This will be a decisive step towards accommodating the surging mobile data traffic. It can also inspire new business models, and collaboration schemes among cellular operators, Wi-Fi operators, and Wi-Fi sharing communities. Besides, this project adopts a radical approach for employing sharing economy models to wireless networks, which is expected to lead in sustainable networking solutions. Moreover, the project will serve as an excellent conduit between academia and industry, involving actively top research labs such as the Alcatel-Lucent Bell Labs. Top-notch network testbed facilities will be employed in order to build realistic models and evaluate the proposed architectures. This will create datasets that can be reused by the research community for studying various aspects of the next generation heterogeneous wireless networks; this will also serve as a method for further disseminating the research outcomes. Special emphasis will be given on recruiting women and under-represented minority (URM) students to the project, as well as undergraduate students via Research Experience for Undergraduates (REUs). A systematic effort will be made in outreach activities to identify, engage, and nurture exceptional talents for careers in STEM.