SelfTuning Wireless Network Power Management Manish Anand Edmund

Self-Tuning Wireless Network Power Management Manish Anand Edmund B. Nightingale Jason Flinn Department of Electrical Engineering and Computer Science University of Michigan 1 Manish Anand

Motivation Wireless connectivity is vital to mobile computing • But, taxes limited battery capacity of a mobile device Power management can extend battery lifetime -However, it can negatively impact performance Mobi. Com 2003 2 Manish Anand

802. 11 Network Power Management Network interface may be continuously-active (CAM) – Large power cost (~1. 5 Watts) – May halve battery lifetime of a handheld Alternatively, can use power-saving mode (PSM) – If no packets at access point, client interface sleeps – Wakes up periodically (beacon every 100 ms) – Reduces network power usage 70 -80% Mobi. Com 2003 3 Manish Anand

Effect of Power Management on NFS Time to list a directory on handheld with Cisco 350 card PSM-static: • 16 -32 x slower • 17 x more energy PSM-adaptive: • up to 26 x slower • 12 x more energy Mobi. Com 2003 4 Manish Anand

What’s Going On? NFS issues RPCs one at a time …. . RPC requests RPC responses NFS Server Access Point Mobile Client Beacons 50 ms 100 ms Each RPC delayed 100 ms – cumulative delay is large – Affects apps with sequential request/response pairs – Examples: file systems, remote X, CORBA, Java RMI… Mobi. Com 2003 5 Manish Anand

Outline • Motivation • Self Tuning Power Management – Design Principles – Implementation – Evaluation • Related Work and Summary Mobi. Com 2003 6 Manish Anand

Know Application Intent Application: NFS File access • Not enough network traffic to switch to CAM Beacon Period • Data rate is dependent on the power mgt. PSM CAM Best Policy: Use CAM during activity period Mobi. Com 2003 7 Manish Anand

Know Application Intent Application: Stock Ticker that is receiving 10 packets per second • Data rate is not dependent on power mgmt. Beacon Period PSM CAM Best policy: Use PSM STPM allows applications to disclose hints about: - When data transfer are occurring - How much data will be transferred (optional) - Max delay on incoming packets Mobi. Com 2003 8 Manish Anand

Be Proactive Transition cost of changing power mode: 200 -600 ms. Large transfers: use a reactive strategy - If transfer large enough, should switch to CAM - Break-even point depends on card characteristics - STPM calculates this dynamically Many applications (like NFS) only make short transfers: be proactive - Benefit of being in CAM small for each transfer - But if many transfers, can amortize transition cost - STPM builds empirical distribution of network transfers - Switches to CAM when it predicts many transfers likely in future Mobi. Com 2003 9 Manish Anand

Respect the Critical Path Many applications are latency sensitive - NFS file accesses - Interactive applications - Performance and Energy critical Other applications are less sensitive to latency - Prefetching, asynchronous write-back (Coda DFS) - Multimedia applications (with client buffering) - Only energy conservation critical Applications disclose the nature of transfer: foreground or background Mobi. Com 2003 10 Manish Anand

Embrace Performance/Energy Tradeoff Inherent tradeoff exists between performance and energy conservation STPM lets user specify relative priorities using a tunable knob Mobi. Com 2003 11 Manish Anand

Adapt to the operating environment Must consider base power of the mobile computer Consider mode that reduces network power from 2 W to 1 W - Delays interactive application by 10% On handheld with base power of 2 Watts: - Reduces power 25% (from 4 W to 3 W) - Energy reduced 17. 5% (still pretty good) On laptop with base power of 15 Watts: - Reduces power by only 5. 9% - Increases energy usage by 3. 5% - Battery lasts longer, user gets less work done Mobi. Com 2003 12 Manish Anand

Outline • Motivation • Self Tuning Power Management – Design Principles – Implementation – Evaluation • Related Work and Summary Mobi. Com 2003 13 Manish Anand

STPM Architecture User or Energy Aware OS Mobi. Com 2003 14 Manish Anand

Transition to CAM STPM switches from PSM to CAM when: 1. Application specifies max delay < beacon period 2. Disclosed transfer size > break-even size 3. Many forthcoming transfers are likely To predict forthcoming transfers STPM generates an empirical distribution of run lengths Transfers >150 ms Run Mobi. Com 2003 >150 ms Run 15 >150 ms Run Manish Anand

Intuition: Using the Run-Length History A good time to switch Switch when expected # of transfers remaining in run is high Mobi. Com 2003 16 Manish Anand

Expected Time to complete a Run Expected time to execute transfers in PSM mode Expected to execute rest of the transfers in CAM mode Time penalty for making PSM to CAM switch Mobi. Com 2003 17 Manish Anand

Expected Energy to complete a Run • Energy calculation includes base power Mobi. Com 2003 18 Manish Anand

Performance and Energy Tradeoff Calculate expected time and energy to switch after each # of transfers – What if these goals conflict? – Refer to knob value for relative priority of each goal! Mobi. Com 2003 19 Manish Anand

Outline • Motivation • Self Tuning Power Management – Design Principles – Implementation – Evaluation • Related Work and Summary Mobi. Com 2003 20 Manish Anand

Evaluation Client: i. PAQ handheld with Cisco 350 wireless card Evaluate STPM vs. CAM, PSM-static, and PSM-adaptive: – NFS distributed file system – Coda distributed file system – XMMS streaming audio – Remote X (thin-client display) Run DFS workload to generate access stats for STPM – Use Mummert’s file system trace (SOSP ’ 95) – File system operations (e. g. create, open, close) – Captures interactive software development Mobi. Com 2003 21 Manish Anand

Results for Coda Distributed File System Workload: 45 minute interactive software development activity Energy (Joules) Time (Minutes) STPM: 21% less energy, 80% less time than 802. 11 b power mgmt. Mobi. Com 2003 22 Manish Anand

Results for Coda on IBM T 20 Laptop Same workload as before: effect of base power on power mgmt strategies Time (Minutes) Energy (Joules) PSM-Static and PSM-Adaptive use more energy than CAM! Mobi. Com 2003 23 Manish Anand

Results for XMMS Streaming Audio Workload: 128 Kb/s streaming MP 3 audio from an Internet server Effect of knowing application intent Power (Watts) XMMS buffers data on client: • App not latency sensitive • PSM uses least power STPM: 2% more power usage than PSM-static – no dropped pkts Mobi. Com 2003 24 Manish Anand

Related Work – Lu, Y. H. , Benini, L. , AND Micheli, G. D. Power-aware operating systems for interactive systems. IEEE Trans. on VLSI (April 2002) – Simunic, T. , Benini, L. , Glynn, P. and Micheli, G. D. Dynamic Power Management for Portable Systems. Mobile Computing and Networking (2000) – Kravets, R. , and Krishnan, P. Application-driven power management for mobile communication. ACM Wireless Nets. (2000) – Shih’s Wake on wireless: (MOBICOM '02) – Krashinsky’s BSD Protocol: (MOBICOM '02) Mobi. Com 2003 25 Manish Anand

Summary STPM adapts to: – Base power of mobile computer – Application network access patterns – Relative priority of performance and energy conservation – Characteristics of network interface Compared to previous power management policies, we perform better and conserve more energy Mobi. Com 2003 26 Manish Anand

Self-Tuning Wireless Network Power Management Manish Anand Edmund B. Nightingale Jason Flinn Department of Electrical Engineering and Computer Science University of Michigan 27 Manish Anand

Expected Time to complete a Run Expected time to execute transfers in PSM mode Expected to execute rest of the transfers in CAM mode Time penalty for making PSM to CAM switch Consider the case of switching before the 3 rd transfer: Mobi. Com 2003 28 Manish Anand

Results for tuning performance/energy Same workload as before: effect of tuning relative priorities CAM knob=100 • Decreasing the knob value PSM-static knob=95 knob=90 PSM-adaptive knob=80 knob=0 -70 Mobi. Com 2003 29 never yields increased energy usage • Increasing the knob value never yields reduced performance Manish Anand

Self Tuning Power Management STPM adapts to: – Base power of mobile computer – Application network access patterns – Relative priority of performance and energy conservation – Characteristics of network interface Compared to previous power management policies, we perform better and conserve more energy Mobi. Com 2003 30 Manish Anand

Results for Non Hinting Applications Running Mummert’s purcell trace on Coda Time (Minutes) Energy (Joules) STPM without hints: 16% less energy, 72% less time than 802. 11 b Power Management Mobi. Com 2003 31 Manish Anand

Results for executing a web trace Result of executing a 45 minute BU web trace TIME ENERGY • CAM performs only 0. 8% better than PSM-static while expending 62% more energy • STPM behaves like PSM-static when conserving energy and like CAM in presence of abundant energy Mobi. Com 2003 32 Manish Anand

Results for Remote X (No Think Time) Energy (Joules) Time (Minutes) STPM uses less energy than CAM if think time > 6. 5 seconds Mobi. Com 2003 33 Manish Anand

Managing Other Devices with STPM well-suited for power management when: – Performance / energy conservation tradeoff exists – Transition costs are substantial Consider disk power management: – Web browser, DFS, mobile DB cache data locally – Hard drive spins down for power saving – Significant transition cost to resume rot. latency – Faster, less energy to read small object from server – But, if many accesses, want to spin-up disk For what other devices can STPM be applied? Mobi. Com 2003 34 Manish Anand

Expected Cost Calculation Mobi. Com 2003 35 Manish Anand

STPM as a wireless power management strategy • Holistic solution – Application intent through hints – Proactive solution using run histogram – Nature of network transfer : foreground or background – Performance/Energy tradeoff with a tunable knob – Operating Environment: base power Mobi. Com 2003 36 Manish Anand