eTrain: Making Wasted Energy Useful by Utilizing Heartbeats for Mobile Data Transmissions

Abstract

With the rapid proliferation of smartphones, hundreds of millions of mobile users are attracted to Instant Messaging (IM) apps. While such apps have brought convenience to our life, it comes with the price of great energy consumption, as these apps keep sending heartbeat messages to the server periodically in order to maintain an always-online connection. These frequent and fragmented transmissions result in a considerable amount of energy waste.In this paper, we investigate the "cost and potential of heartbeats". We quantify power consumption of heartbeats of real-world IM apps through extensive measurements. The measurement results confirm that huge power consumption is induced by heartbeats. The goal of this paper is to save energy by turning the energy wastage of heartbeats into transmitting useful data. Thus, we develop eTrain, a transmission management system running on Android phones, which takes advantage of IM heartbeats (as trains) to piggyback aggregated delay-tolerant apps' data such as e-mail and Weibo (as cargoes) via an online transmission strategy, so as to minimize the cumulative tail energy without sacrificing user-specified deadlines. Compared to other existing works, eTrain can reduce more energy consumption under the same settings. Experiments conducted on smartphones show that eTrain can achieve 12%-33% energy saving in various application scenarios.

Energy Minimization for Federated Asynchronous Learning on Battery-Powered Mobile Devices via Application Co-running

“…The efforts of energy optimization on mobile devices has been revolving around software and hardware components to elongate battery lifetime, e.g., dynamic voltage and frequency scaling, resolving "energy bugs" from unexpected energy consumption [16] and coalescing packets to reduce tail energy on the wireless network interface [17], [18] using the Lyapunov framework [27]. For on-device training [7], delegating the long-running, training workloads as a background service is a viable way to avoid interrupting normal usage.…”

“…Since a running foreground application would have already activated shared resources on the big cores and the background processes are typically dispatched to the little cores, co-running training with applications could take advantages of such energy disproportionality [19]. Similar to packet coalescing [17], [18], this idea of task bundling dates back to piggyback sensing activities with applications such as web browsing and phone calls [20]. However, these early works cannot be readily applied to federated learning to achieve the energy-staleness trade-off as well as bounded staleness with statistical stability.…”

“…However, these early works cannot be readily applied to federated learning to achieve the energy-staleness trade-off as well as bounded staleness with statistical stability. The closest works to ours are [17], [18] that adopt the Lyapunov framework to achieve energy-delay trade-offs. This paper takes a step forward to consider gradient staleness induced by delayed execution and attempts to fill the gap between the machine learning algorithms and mobile systems for optimal energy efficiency.…”

Towards Mobile Edge Computing: Taxonomy, Challenges, Applications and Future Realms

“…Zhang et al [136] proposed a technique to make the wasted energy useful. The power consumption of heartbeats is quantified by extensive measurement.…”

Profiling Energy Efficiency and Data Communications for Mobile Internet of Things

“…Different tools have been developed to measure energy consumption on smart devices and smartphones. For example, power monitor meter has been used to provide the current with constant voltage 3.7 V to the smartphone instead of using the battery [14]. This hardware setup can provide accurate power consumption measurements, but it is a bulky solution not suitable for ordinary users in their daily usage.…”