An alternative approach is collecting network performance
data on end-user devices in a crowd-sourcing based man- ner [4]–[6]. This allows estimating and predicting from those the cellular network quality at certain times and locations, as experienced by the end-user. Still, active measurements are expensive in terms of energy and may even cause monetary cost for the users. Knowing the network and cell utilization, a number of optimizations can be derived for the handset. For example background traffic may be delayed as long as the network performance is low. This allows App developers to optimize their application to adjust the quality of the consumed content a priori, and to improve the user experience by reducing delays, reducing the power consumption, and such increasing the quality of experience (QoE) of the end-user.
This paper focuses on analyzing the parameters affecting the cellular network performance. Special emphasis was put on the RTT, as an increasing number of services are delay sensitive (e.g. mobile cloud gaming, or health care). Hence, this paper aims to address the following questions:
1) How does network configuration and management influ- ence the cellular network performance?
2) Can the performance of the cellular network be predicted based on the available measurements?
3) What is the optimization potential in the given network configuration, and how can this be exploited?
4) How is the user-perceived mobile service quality affected by routing decisions of the mobile operator?