User-in-the-loop for hethetnets with backhaul capacity constraints

A popular method to model heterogeneous networks is the use of two independent homogeneous Poisson point processes to locate UEs and BSs with unlimited backhaul capacity. Despite the analytical tractability, this approach is far from accurate. First, the distribution of UEs in real scenarios is neither homogeneous nor independent of BSs. Besides, the assumption of unlimited capacity for backhaul connections is optimistic, especially in the future 5G HetNets with small cells. In this article, we propose a novel modeling approach for heterogeneous networks with heterogeneous spatial traffic distribution (HetHetNets). Specifically, in the proposed model, a particular ratio of UEs are collocated with the BSs while the rest of UEs are independently and homogeneously distributed in the network. Moreover, the proposed model presumes backhaul connections with constrained capacity. We study the impact of this more realistic network modeling on the effectiveness of the spatial user-in-the-loop (UIL) schemes in HetHetNets. Spatial UIL assumes that (some) UEs can be influenced by the operator to move in the network. Finally, we propose a new objective for the UIL mechanism that takes into account the impact of the BS loads and the backhaul capacities on the network performance.