Price-aware traffic splitting in D2D HetNets with cost-energy-QoE tradeoffs

With the advances in wireless technologies, there has been tremendous increase in user demand for resource-intensive mobile Internet services. This has been coupled with the limited capabilities of mobile devices and the high level of user expectations in terms of both quality of experience (QoE) and cost. An attractive enhancement technique is to utilize the co-existence of multiple interfaces in wireless devices to connect simultaneously to different access networks including cooperation over device-to-device (D2D) links. To this end, we present in this work optimized user-centric traffic splitting strategies in heterogeneous networks to achieve a high QoE level for video on demand applications with low cost and energy consumption for end users. We formulate a multi-objective optimization problem considering different pricing models with prediction whereby a user can estimate the links' performance for future time slots and make suitable decisions accordingly. We evaluate the proposed strategies and demonstrate their effectiveness using parameters determined via experimental measurements to provide an evaluation under realistic operational conditions. Results provide useful insights on the tradeoffs between energy consumption, cost and quality of experience.