Pricing Based Distributed Traffic Allocation for 5G Heterogeneous Networks

Abstract

Next generation (5G and beyond) mobile networks will highly rely on distributed heterogeneous infrastructure, which is foreseen to create a versatile resource ecosystem. In such network deployments, efficient traffic allocation and management is of prominent priority. The increasing congestion conditions of cellular infrastructures have steered extensive research focus towards Heterogeneous Network (HetNet) access solutions. The deployment of dense multiple radio access technology (RAT) networks combined with the enhanced capabilities of mobile devices to maintain concurrent connections through different RATs, add up to the complexity of the traffic allocation problem. Traffic allocation becomes more challenging when using diverse technologies concurrently to serve multiple traffic classes of different Quality of Service (QoS) demands. Towards a viable solution, we have proposed a pricing-based network selection process for heterogeneous networks, running in a distributed manner, which has proved to lead to equilibrium under the developed policies. In this paper, we extend our pricing algorithm by providing a new approach that introduces personalized pricing based on the load pressure that an allocation decision will impose to a RAT. The central system decisions are confined to dynamically select the UE polling rank policy that is presenting the closest potential performance to that of the optimal centralized solution, used as the benchmark ranking. The optimal policy is based on proportional fairness and is used as a reference centralized policy to drive our distributed solutions. We assess the enhanced pricing scheme proposed in this work, and we employ extensive simulations and testbed experimentation to evaluate its performance in terms of efficient bandwidth allocation, price variation, fairness and induced cost to clients served by the HetNet.