Achieving Fair LTE/Wi-Fi Coexistence with Real-Time Scheduling

Carrier-Sensing Adaptive Transmission (CSAT) is a promising approach to address coexistence between LTE and Wi-Fi in unlicensed bands. Under CSAT, a key problem is the design of a scheduling algorithm to allocate radio resources (across multiple channels and a large number of sub-channels) for LTE and Wi-Fi users. This paper investigates this scheduling problem through an optimization formulation with the objective of minimizing LTE's adverse impact on Wi-Fi users. Special considerations of each LTE user's uplink/downlink rate requirements and channel conditions are given in this optimization formulation. We show that this scheduling problem is NP-hard and propose to develop a near-optimal solution. A major challenge here is to ensure the scheduler can obtain a solution on similar to 1 ms time scale - a stringent timing requirement to meet LTE standard. Our main contribution is the development of CURT, a scheduling algorithm that can obtain near-optimal solution in similar to 1 ms under standard LTE small cell scenarios. CURT exploits the unique structure of the underlying optimization problem and decomposes it into a large number of independent sub-problems. By taking advantage of GPU's parallel processors, we allow the large number of sub-problems to be run in parallel and independently from each other. By implementing CURT on Nvidia GPU/CUDA platform, we demonstrate that CURT can deliver near-optimal scheduling solution in similar to 1 ms for LTE small cells with no more than 20 users following 3GPP's evaluation methodology.