The Edge computing (also called Fog computing) paradigm is becoming a new solution that can overcome the limitations of traditional Cloud solutions, especially for distributed application with strong constraints on response time. Our research aims at proposing innovative techniques for the management of an Edge computing platform. We consider a distributed scenario where a set of sources produce data to be processed by an intermediate layer of edge computing nodes before being forwarded to one or more Cloud computing data centers. The research aims at tackling the problems of managing such infrastructure considering multiple aspects ranging from computing load to network characteristics and up to power efficiency. All these research topics should be analyzed combining multiple techniques ranging from mathematical models (e.g., performance models, energy models, optimization problems) to simulation-based performance evaluation and up to prototype development and testing.