A Buffer Cache Algorithm Using the Characteristic of Mobile Applications based on Hybrid Memory Architecture

In general, current mobile devices employ a buffer cache mechanism which has been designed to mitigate the performance gap between CPU and storage. However, it cannot fit in mobile devices because it does not consider the characteristics of the mobile applications (e.g., foreground or background state). Therefore, existing buffer cache mechanisms can cause performance degradation by I/O interference between foreground applications and background applications. In addition, DRAM accelerates energy consumption by performing periodic refresh operations. Therefore, DRAM-based memory system can cause critical power problem due to its limited battery capacity in the mobile devices. In this paper, we propose a novel buffer cache algorithm for mobile devices based on hybrid memory architecture composed of DRAM and non-volatile PCM. Proposed algorithm is based on key observation that background applications rarely issue I/O requests and they may interfere with foreground applications. For evaluation, we implemented the proposed algorithm and compared the performance with two widely known algorithms. Our experimental results show that our algorithm reduces the elapsed time of the foreground applications by 70% on average and the power consumption by 67% on average while operating same workloads.

[1]  Slo-Li Chu,et al.  Design a Low-Power Scheduling Mechanism for a Multicore Android System , 2012, 2012 Fifth International Symposium on Parallel Architectures, Algorithms and Programming.

[2]  Hyokyung Bahn,et al.  CLOCK-DWF: A Write-History-Aware Page Replacement Algorithm for Hybrid PCM and DRAM Memory Architectures , 2014, IEEE Transactions on Computers.

[3]  Dennis Shasha,et al.  2Q: A Low Overhead High Performance Buffer Management Replacement Algorithm , 1994, VLDB.

[4]  Tajana Simunic,et al.  PDRAM: A hybrid PRAM and DRAM main memory system , 2009, 2009 46th ACM/IEEE Design Automation Conference.

[5]  Young Ik Eom,et al.  User-aware power management for mobile devices , 2013, 2013 IEEE 2nd Global Conference on Consumer Electronics (GCCE).

[6]  Sean Eilert,et al.  Phase Change Memory: A New Memory Enables New Memory Usage Models , 2009, 2009 IEEE International Memory Workshop.

[7]  吉田 則裕,et al.  Android Open Source Projectを対象としたパッチレビュー活動の調査 , 2012 .

[8]  Vijayalakshmi Srinivasan,et al.  Scalable high performance main memory system using phase-change memory technology , 2009, ISCA '09.

[9]  Young Ik Eom,et al.  Low Power Killer: Extending the battery lifespan by reducing I/O on mobile devices , 2015, 2015 IEEE International Conference on Consumer Electronics (ICCE).

[10]  C. Bonnet,et al.  Android power management: Current and future trends , 2012, 2012 The First IEEE Workshop on Enabling Technologies for Smartphone and Internet of Things (ETSIoT).

[11]  Gernot Heiser,et al.  An Analysis of Power Consumption in a Smartphone , 2010, USENIX Annual Technical Conference.