Avoiding handover interruptions in pervasive communication applications through machine learning

There are an increasing diversity of devices and networks for making and receiving video and/or audio calls. The prevalence of smartphones, multiple smart speakers and various computing platforms in many homes has led to much greater flexibility in how and where users communicate. This trend is only set to continue as AR/VR becomes more commonplace and services such as healthcare, for which flexible, high-quality communications is a critical component, move online. Despite these trends, comparatively little has been done to offer a converged communications experience once a speech call session is in progress. For example, the simple act of switching an ongoing call from a smartphone to a smart speaker is often a highly manual process. Pervasive communication systems aim to address this by providing a seamless, flexible communications experience across multiple devices and, where required, multiple networks. A key part of achieving this experience is through seamless session handover. This paper considers how session initiation protocol systems can operate in a pervasive communication scenario, in particular when there is mobility causing vertical handover between delivery networks. The paper uses a machine learning-based approach to predict user’s transitions and thus overcome interruptions in speech due to signalling handover. As pervasive communication systems are not currently available for measuring the performance of the solution, the paper uses a commonly available dataset for vehicular mobility to assess likely handover performance. The results show that prediction can reduce the more concerning interruptions (>2s) due to vertical handover events by more than 99.9%.