On the impact of backhaul network on distributed cloud computing

Nowadays, heterogeneous networks are seen as a key enabler to meet the challenging performance required for the evolution of LTE networks. At the same time, cloud computing is an emerging trend that outsources computing and storage resources, and extends mobile handsets capability of processing large data amounts. Cooperation in heterogeneous networks allows better (improved) network management by reducing power consumption, enhancing network coverage, and reducing deployment and communication costs. Within the cloud paradigm, federating femtocells in clusters not only helps with network management, but also opens the possibility to use these clusters for both communication and computation. Clusters of femtocells with virtual machine hosting capabilities is a novel disruptive paradigm which adds a dimension to cooperative management of heterogeneous networks. However, effective cooperation among geographically neighboring femtocells is prone to backhauling energy costs and delay limitations. In this paper, we propose a framework for evaluating backhauling performance in a femto-cloud architecture, considering several backhaul technologies and topologies. Our simulations explicit the limitations of clustering in femto-cloud imposed by backhauling.