A data-driven scheduler performance model for QoE assessment in a LTE radio network planning tool

Abstract The use of static system-level simulators is common practice for estimating the impact of re-planning actions in cellular networks. In this paper, a modification of a classical static Long Term Evolution (LTE) simulator is proposed to estimate the Quality of Experience (QoE) provided in each location on a per-service basis. The core of the simulator is the estimation of radio connection throughput on a location and service basis. For this purpose, a new analytical performance model for the packet scheduling process in a multi-service scenario is developed. Model parameters can easily be adjusted with information from radio connection traces available in the network management system. The simulation tool is validated with a large trace dataset taken from a live LTE network.