Home // International Journal On Advances in Telecommunications, volume 3, numbers 1 and 2, 2010 // View article

Network Prediction for Energy-Aware Transmission in Mobile Applications

Authors:
Ramya Sri Kalyanaraman
Yu Xiao
Antti Ylä-Jääski

Keywords: prediction; adaptation; SNR; power; context-awareness; policy-based.

Abstract:
Network parameters such as signal-to-noise-ratio (SNR), throughput, and packet loss rate can be used for measuring the wireless network performance which highly depends on the wireless network conditions. Previous works on energy consumption have shown that the performance of wireless networks have impact on the energy efficiency of data transmission. Hence, it is potential to gain energy savings by adapting the data transmission to the changing network conditions. This adaptation requires accurate and energy-efficient prediction of the network performance parameters. In this paper, we focus on the prediction of SNR and the prediction-based network adaptations for energy savings. Based on the SNR data sets collected from diverse real-life networks, we first evaluate three prediction algorithms, namely, Autoregressive Integrated Moving Average, Newton Forward Interpolation, and Markov Chain. We compare these three algorithms in terms of prediction accuracy and energy overhead. Later we propose a threshold based adaptive policy which controls the data transmission based on the predicted SNR values. To evaluate the effectiveness of using network prediction in adaptation, we use a FTP as a case study and compare the network goodput and energy consumption under different network conditions. The experimental results show that the usage of adaptations improves the network goodput. Furthermore, the adaptations using prediction can save up to 40% energy under specific network conditions when compared to the adaptation without prediction.

Pages: 72 to 82

Copyright: Copyright (c) to authors, 2010. Used with permission.

Publication date: September 5, 2010

Published in: journal

ISSN: 1942-2601