A Power-Aware Distributed Wi-Fi Access Point Scheduling Algorithm Asma Enayet, Nusrat Mehajabin and Md. Abdur Razzaque ∗ Green Networking Research Group University of Dhaka, Bangladesh asmaenayet, nusrat.47.cse@gmail.com, razzaque@cse.univdhaka.edu Choong Seon Hong † Networking Laboratory, School of Electronics and Information, Kyung Hee University 1732 Deokyoungdaero, Giheung-gu, Yongin-si, Gyeonggi-do, 446-701, Korea, South Korea cshong@khu.ac.kr ABSTRACT Smartphone has implemented several power saving strate- gies for its Wi-Fi radio as the Wi-fi radio consumes signif- icant amount of energy. Previously, Wi-Fi energy saving mechanisms have been designed in such a way that add un- fairness to the Wi-Fi network. Prime reasons of the un- fairness are retransmissions due to collisions in the channel, unnecessary waiting in high power mode and hidden termi- nals in the vicinity. We propose PowerNap an improved distributed energy effi- cient access point (AP) scheduling algorithm that addresses the above issues. PowerNap schedules the APs to scale down the overlapping among the transmission time of APs in the same vicinity. This trims out the energy consumption of Smartphone and also avoids unfairness. PowerNap sched- ules the APs in a weighted manner and also supports dy- namic rescheduling. The proposed algorithm could be im- plemented via software installation in APs. Implementation of the PowerNap prototype improves Smartphone battery life upto 49- 60%. Categories and Subject Descriptors H.4 [Mobile networking]: Miscellaneous; D.2.8 [Mobile networks]: Energy savings—Algorithms, performance mea- sures General Terms Algorithms, Wi-Fi, Smartphone Keywords Wi-Fi Network, Energy Optimization, Distributed Schedul- ing Algorithm, Traffic Measurement ∗ Dept. of Computer Science & Engg., University of Dhaka † Dept. of Computer Engineering, Kyung Hee University Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. ICUIMC(IMCOM)’13 , January 17-19, 2013, Kota Kinabalu, Malaysia Copyright 2013 ACM 978-1-4503-1958-4 ...$15.00. 1. INTRODUCTION Today, Smartphones support numerous applications, added processing power, and ease of the Internet access. They re- quire extended battery life to satisfy the user needs. More explicitly, users demand longer battery life in wireless hand- held devices such as Smartphones to receive uninterrupted services on the go without compensating the number of run- ning applications. Thus, it has always been a critical goal for Smartphones to optimize energy. Improved user experience and productivity will result from this optimization. This cer- tainly necessitates optimizing the energy hungry processes so that the power consumption is minimized. The major power consuming components are display, sensors, commu- nication and computation circuitries [7, 8]. Widespread mo- bile networking technologies, for instance 3G, GSM and Wi- Fi, necessitate power draining at high frequency. For Wi-Fi, a large portion of consumed power is wasted at the Access Point (AP) level due to network contention [8, 5]. In this work, our goal is to diminish this loss of energy by scheduling the AP’s transmission periods in such a way that their time slot overlapping and the traffic contentions are reduced. The optimization of network activity regarding Smartphone with a view to prolong battery life has several state-of-the- art implementations and solutions. The 802.11 PSM (power saving mode) [4] is implemented as the default power conser- vation strategy for Wi-Fi with the limitation of being able to only economize energy if the network activity is idle for a certain predefined duration. Also, PSM does not account for situations where multiple clients may wake up at the same time to use the channel, which results in heavy contention in the network. This problem has been addressed in NAPman [8] that prioritizes the packets and provides a virtual copy of an AP to every connected user. Thus, NAPman supports multiple clients under an AP without collision through vir- tualization. However, the virtualization of one AP does not care for the existence of multiple physical APs in the same vicinity. Therefore, traffic contention will be increased in the environment, wherein, there are many overlapping APs. Another power saving approach is Catnap [2] that introduces operating system (OS) buffer to store the packets from wired part of the network and postpone the delivery to the wireless section until the deadline. But it imposes a constraint that the wireless bandwidth has to be larger than the wired to be able to save energy. A different approach towards power saving in Smartphones via packet aggregation is found in