Network-Coded Macrocell Offloading in Femtocaching-Assisted Cellular Networks
The femtocaching idea was proposed as a solution to compensate for the weak backhaul capacity, by deploying coverage-limited nodes with high storage capacity called femtocaches (FCs). In this paper, the macrocell offloading problem in femtocaching-assisted cellular networks is investigated. The objective is to minimize the number of transmissions by the macrocell base station (MBS) given that all requests should be served simultaneously to satisfy quality-of-experience (QoE) of the clients. We first formulate this MBS offloading problem as an optimization problem over a network coding graph, and show that it is NP-hard. Therefore, we propose an ONC-broadcast offloading scheme that exploits both broadcasting and opportunistic network coding (ONC) to minimize the number of required MBS transmissions. We utilize a random graph model to approximate the performance of the proposed ONC-broadcast scheme in terms of the resultant average number of transmissions by the MBS. Moreover, despite the complexity of finding the optimal solution for each and every case, we prove that this ONC-broadcast scheme is asymptotically optimal, i.e., for large number of requests, the ONC-broadcast scheme achieves a similar macrocell offloading performance to that of the optimal solution. To implement the ONC-broadcast scheme, we devise a heuristic that employs a dual conflict graph or broadcasting at the FCs such that the remaining requests can be served using the minimum number of transmissions at the MBS. Simulations show that the dual graph scheme improves MBS offloading as compared to the traditional separate graph scheme. Furthermore, the simple heuristic proposed to implement the ONC-broadcast scheme achieves a very close performance to the optimal ONC-broadcast scheme.