Analyses of ping-pong handovers in real 4G telecommunication networks

Abstract

The last decade has been characterized by a rapid increase in the usage of mobile communications. One of the main aspects of mobile communications is mobility. This means that mobile phones have to switch between base station cells in order to support the uninterrupted usage of all available services within the area of the mobile network coverage. The process of switching user devices between base station cells is called a handover. Accordingly, base stations are optimized to serve mobile phones with certain moving velocities based on an area and user characteristics to better handle handovers. However, issues appear when a mobile phone starts switching rapidly between two different cells in a short period of time, referred to as ping-pong handover. These occurrences cause negative effects in terms of increased telecom operators signaling overhead and user dissatisfaction. The 3rd Generation Partnership Project (3GPP) defines a metric called ping-pong rate in order to analyze what is the most adequate base station configuration concerning different user velocities and how well base stations handle handovers. The ping-pong rate is defined as a share of ping-pong handovers in total handovers, without taking into consideration the conditions under which they occur. In this paper, the conditions under which ping-pong handovers occur are analyzed by using live, non-simulated data. Findings show that the handovers under different conditions have different probabilities to become ping-pong handovers. Furthermore, these findings are utilized for defining a new empiric metric used for evaluating the handover handling of pre-configured base stations. Consequently, the novel empiric metric provides better insights into the base station handling of handovers than the ping-pong rate defined by 3GPP.