QoS provisioning in LTE-A networks with relay nodes

In the vision of heterogeneous networks, the Quality of Service provided by classic cellular systems is improved thanks to a jointly adoption of macro and small-range cells (i.e., micro, pico, and femtocells) working in the same area. This rationale has been also assimilated by 3GPP, which now proposes its possible exploitation in the upcoming Long Term Evolution-Advanced architecture. In this context, the wireless relay technology represents an interesting solution for extending the coverage of a macrocell by means of a relay node which offers wireless connectivity to mobile users and, at the same time, is connected to the eNode-B (and hence to the backbone) through a wireless backhaul link. But, the standard does not specify any optimal resource allocation strategy for the two hop path formed by the user and the backhaul links, which is indeed very relevant to satisfy the level of QoS expected by end users. To this aim, in the present contribution, we present a novel scheduling strategy, conceived as an extension of the recently proposed Two- Level Scheduler approach, able to effectively support real-time applications in the presence of relay nodes. Its performance has been tested in several network scenarios that involve a variable number of users hosting real-time multimedia applications and best effort flows. A comparison with respect to Proportional Fair and Logarithmic strategies has been also provided, showing that the proposed approach is able to greatly reduce the Packet Loss Ratio experienced by multimedia applications, at the expense of a slight impairment of the throughput of the best effort flows.