Poster: Context-Triggered Mobile Network Measurement

Abstract

While the availability and accessibility of cellular network connectivity have improved in recent years, our ability to diagnose and debug network problems in this environment has not. One key challenge is that many of the network problems occur near the edge of the network where only mobile devices can perceive them, but network and battery resources to conduct measurements from these mobile devices are scarce. Traditional network measurement approaches that use continuous, periodic, or random measurements are either infeasible or ineffective in this environment. In this work, we propose an alternative approach: triggering measurements based on relevant device context such as signal strength and historical performance data, which can inform decisions for when to measure current network performance. This context can be collected locally on the device as well as aggregated at a global scale to schedule measurement based on data collected from multiple devices. By carefully selecting when to conduct a measurement, and using prediction to improve the likelihood that triggered measurements will succeed, we can more reliably measure important network phenomena with less overhead. Using Mobilyzer [3] as a platform for evaluation, we propose an architecture that is sufficiently general to support a wide range of triggered measurement experiments. We demonstrate the use of this framework for measurements on mobile platforms that are traditionally difficult to capture, e.g., handoff measurement. Further, we can use the global scheduler to predict which devices will likely satisfy the preconditions for the triggered measurement to improve the measurement success rate. Compared to previous work [2, 1, 4], ours is the first to propose a general framework to enable context-triggered mobile measurement, leveraging both local and global visibility into context while ensuring low overhead.