User-Centric resource allocation through a power-efficient jamming-proof RAN on top of a multi-tenant backhaul

The exponential growth in the number of communicating devices and the increasing demand for better and high-performance communications have made allocation of network resources an issue of extreme importance. The problem is further exacerbated if we consider that, while the number of deployed devices is massive, the amount of network resources is still limited. In this thesis, we aim at providing a holistic approach for resource allocation in modern telecommunication networks. Specifically, we consider a network consisting of a backhaul interconnecting a Radio Access Network (RAN) that provides mobile users with wireless access to a Core Network (CN). The CN allows access to the Internet and enables end-to-end communications by routing users data and calls. We take both energy and security aspects into account, by proposing a power-efficient and jamming-proof resource allocation scheme for the RAN. Furthermore, we consider the relevant and emerging case where the backhaul is shared among different tenants. Accordingly, we exploit Software Defined Networks (SDNs) and Network Function Virtualization (NFV) paradigms to provide dynamic and flexible network and service management in the multi-tenant backhaul and CN, respectively. We formulate the resource allocation problem through both centralized and distributed approaches, we discuss the existence and uniqueness of efficient resource allocation solutions and we provide distributed privacy-preserving algorithms that provably converge allow to the optimal resource allocation policy by exploiting only local or shared information.