A Method of Power Aware Large Data Download on Smartphone

Jie Ren, Ling Gao, Hai Wang, Yan ChenIEICE 2015 · DOI

Abstract: The endurance of smartphones remains constrained by limited battery capacity. Applications that download large amounts of data over slow or unstable networks significantly exacerbate this issue. This paper proposes a smartphone download strategy designed for low energy consumption, termed CLSA (Concentrated Download and Low Power and Stable Link Selection Algorithm). CLSA reduces the overhead of large data downloads by intelligently delaying requests based on three factors: current network conditions, the length of the download request queue, and local device information. Evaluated using a music player implementation on an Android-based ZTE V880, CLSA demonstrates a superior trade-off between energy consumption and delay compared to standard "Minimum Delay" and "WiFi Only" strategies.

Introduction

As smartphones become central to daily life, the energy consumption of their Wireless Network Interfaces (WNI)—specifically WiFi and cellular radios—has become a critical bottleneck. Large data downloads, such as those required by media streaming or cloud syncing, are particularly taxing. Standard operating system behaviors often prioritize immediate data retrieval, which can lead to inefficient energy use when network conditions are poor or when requests are frequent and small.

The Energy Efficiency Challenge

Two primary phenomena contribute to battery drain during data transfers:

  1. The "Tail Effect": Radios do not return to an idle state immediately after a transmission ends. They remain in a high-power state for a "tail" period, waiting for potential subsequent data. If an application makes many small, scattered requests, the radio stays active far longer than necessary.
  2. Signal Quality and Throughput: In areas with low Received Signal Strength Indicators (RSSI), the radio must increase its transmission power while experiencing lower data rates. This results in more Joules spent per byte downloaded.

The CLSA Strategy: Concentrated Waking-up

The core innovation of the CLSA algorithm is the shift from a "first-come, first-served" model to a "concentrated waking-up" strategy. Instead of activating the radio for every individual request, the system batches requests and waits for optimal conditions.

Comparison of frequent vs. concentrated wake-up strategies.

By increasing sleep durations and reducing the frequency of wake-ups, the system minimizes the cumulative time spent in the high-power "tail" state. CLSA utilizes the Lyapunov optimization framework to manage this delay, ensuring that while downloads are batched to save power, the queue remains stable and does not exceed user-tolerable limits.

Software Architecture and Implementation

To implement this strategy, the researchers modified the standard Producer/Consumer software pattern. In a typical model, the consumer thread wakes up as soon as the buffer is non-empty. In the CLSA model, a dynamic "upper threshold" is introduced.

Flowchart of the modified consumer thread model incorporating an upper threshold.

As shown in the flowchart, the consumer thread remains in a sleep state until the data backlog reaches a specific threshold or the network conditions (RSSI) improve significantly. This threshold is not static; it is adjusted based on a control parameter ($W$) that allows the system to balance the priority of energy savings versus delivery speed.

Experimental Results

The algorithm was tested in real-world environments, including dormitories and libraries, using a custom music player. The researchers measured precise power consumption using Agilent digital multimeters and DC power supplies.

Key findings include:

  • Energy Efficiency: In poor network environments (e.g., a dormitory with weak WiFi), CLSA reduced energy consumption per byte by over 60% compared to the "Minimum Delay" strategy ($1.88 \times 10^{-6}$ J/byte vs. $5.41 \times 10^{-6}$ J/byte).
  • Balanced Performance: While a "WiFi Only" strategy might use less energy, it often results in unacceptable delays if a hotspot is not immediately available. CLSA achieved a manageable delay while maintaining high energy efficiency.
  • Optimization: The study found that a weight parameter ($W$) of approximately 22 provided the optimal balance for the test device, effectively minimizing the product of energy and delay.

Conclusion

CLSA provides a robust framework for managing large data downloads on mobile devices. By leveraging Lyapunov optimization and RSSI-based forecasting, it allows smartphones to "wait for the right moment" to download, significantly extending battery life without compromising the user experience. Future work may involve integrating mobility patterns and GPS data to further refine network selection and timing.

Summary updated

Abstract

The endurance time of smartphone still suffer from the limited battery capacity, and smartphone apps will increase the burden of the battery if they download large data over slow network. So how to manage the download tasks is an important work. To this end we propose a smartphone download strategy with low energy consumption which called CLSA (Concentrated Download and Low Power and Stable Link Selection Algorithm). The CLSA is intended to reduce the overhead of large data downloads by appropriate delay for the smartphone, and it based on three major factors: the current network situation, the length of download requests' queue and the local information of smartphone. We evaluate the CLSA using a music player implementation on ZTE V880 smartphone running the Android operation system, and compare it with the other two general download strategies, Minimum Delay and WiFi Only. Experiments show that our download algorithm can achieve a better trade-off between energy and delay than the other two.
BibTeX
@inproceedings{ren2015method,
  author = {Jie Ren and Ling Gao and Hai Wang and Yan Chen},
  title = {{A Method of Power Aware Large Data Download on Smartphone}},
  booktitle = {IEICE},
  year = {2015},
  doi = {10.1587/transinf.2014edp7178},
}
Figure