Salil Sawhney

Performance of CS Fallback from LTE to UMTS

Circuit Switched Fallback (CSFB) is the most commonly used method to support voice services over Long Term Evolution (LTE) networks today, as the deployment of IP Multimedia Subsystem (IMS) is still in its infancy. In this article, we discuss the performance of CSFB voice traffic redirection from LTE to UMTS using data from live commercial networks. The key factors impacting CSFB call setup delay are highlighted. Data from multiple cell geometries including stationary and mobility conditions are presented for Mobile Originated (MO) calls towards landlines, Mobile Terminated (MT) calls from landlines and Mobile-to-Mobile (M2M) CSFB calls. Analysis results in well optimized deployments show that on an average, MO/MT call-setup time for CSFB from LTE to UMTS is around 1 sec greater than legacy UMTS. However, the results can vary depending on the network configuration and the conditions of the measurement. To illustrate the optimization of CSFB performance in real networks, we also highlight the principal call set up optimization and implementation factors impacting CSFB call setup delay and success rates. This article demonstrates that in well optimized networks, CSFB to UMTS call setup success rates can be very close to those achieved in the legacy UMTS systems.

Layer management through idle-mode parameter optimization in multi-carrier LTE networks

In this paper, we investigate the load balancing problem in multi-carrier LTE networks. We propose a probabilistic framework for idle-mode parameter optimization in Self-Optimizing Networks (SON) in order to achieve the target traffic distribution per carrier. In particular, we consider different layer management strategies such as hierarchical, equal priorities. For each strategy, we mathematically optimize the underlying idle-mode parameters defined by 3GPP specification, based on the statistical information of radio frequency (RF) propagation in order to achieve the target traffic distribution. Moreover, we characterize the underlying tradeoffs between the traffic controllability and user experience for each strategy. Finally, we show the performance of our proposed framework in real network deployments.