Deployment problems in Mobile Networks – DEMON

For decades, the study of cellular networks was limited to analytical and simulation studies. Indeed, running experiments on operational cellular networks was limited to very specific operator studies, and building a cellular network testbed seemed prohibitive, considering the equipment costs. However, open implementations of the cellular architecture, such as srsRAN, running on standard computers and using relatively cheap software-defined radio equipment, have surfaced recently. Therefore, the project has an important experimental component, an original methodological feature in the study of cellular networks, based on an experimental platform using srsLTE, formed of one network core and multiple operational base stations, running on USRP NI 2901 equipment. This platform can run on static computers, but it also integrates a terrestrial Diablo robot. Drone integration has given mitigated results, as the USRP equipment is relatively heavy and expensive, and transporting it with drones is not totally safe. These experimental results complete an extensive set of simulations, conducted on an event-driven simulator specifically developed for this task.

The project proved experimentally the feasibility of migrating the user context from a mobile base station to another, without any service interruption noticeable to the users. We studied the use of mobile base stations in multiple scenarios: i) interventions of first responders, with or without pre-existing fixed infrastructure; ii) tactical networks with strict security constraints; iii) commercial mobile networks in urban areas; iv) service providers willing to deploy their own infrastructure. In all these use cases, integrating mobile base stations resulted in significant gains over multiple metrics, such as achievable throughput, quality of service and operational costs.

The establishment of a reliable backhaul link between mobile base stations and the core network remains a barrier at present, but the development of beamforming techniques may represent a potential solution.

The adoption of mobile base stations could represent a real paradigm shift in the world of cellular networks, with significant cost reductions. This could democratize access to the mobile network market, allowing new players to emerge. Overall, this architectural evolution could improve connectivity in difficult situations, while reducing its financial and environmental costs.

Cellular networks have been continuously reshaping communication in our society, through the rapid evolution of standards, products, and use cases. However, the careful planning and deployment strategies, implemented by mobile operators in order to provide coverage and data services to users, were constants during all these decades. But recent evolutions in terms of communication equipment miniaturization, network virtualization and autonomous vehicles are challenging this vision. In this context, the DEMON project proposes a major evolution of the cellular network architectures, by using mobile base stations. The main objective of the DEMON project is to provide dedicated solutions for this new architecture, demonstrating the advantages of a self-deployable approach, detecting key challenges in this new paradigm and addressing these challenges