Reducing Network Energy Consumption via Sleeping and Rate Adaptation.

Slides:



Advertisements
Similar presentations
Mobility Increase the Capacity of Ad-hoc Wireless Network Matthias Gossglauser / David Tse Infocom 2001.
Advertisements

Reducing Network Energy Consumption via Sleeping and Rate- Adaption Sergiu Nedevschi, Lucian Popa, Gianluca Iannaccone, Sylvia Ratnasamy, David Wetherall.
A 2 -MAC: An Adaptive, Anycast MAC Protocol for Wireless Sensor Networks Hwee-Xian TAN and Mun Choon CHAN Department of Computer Science, School of Computing.
1 CONGESTION CONTROL. 2 Congestion Control When one part of the subnet (e.g. one or more routers in an area) becomes overloaded, congestion results. Because.
1 MemScale: Active Low-Power Modes for Main Memory Qingyuan Deng, David Meisner*, Luiz Ramos, Thomas F. Wenisch*, and Ricardo Bianchini Rutgers University.
Optimizing Buffer Management for Reliable Multicast Zhen Xiao AT&T Labs – Research Joint work with Ken Birman and Robbert van Renesse.
Rumor Routing in Sensor Networks David Braginsky and Deborah Estrin Presented By Tu Tran 1.
Improving TCP Performance over Mobile Ad Hoc Networks by Exploiting Cross- Layer Information Awareness Xin Yu Department Of Computer Science New York University,
XCP: Congestion Control for High Bandwidth-Delay Product Network Dina Katabi, Mark Handley and Charlie Rohrs Presented by Ao-Jan Su.
Receiver-driven Layered Multicast S. McCanne, V. Jacobsen and M. Vetterli SIGCOMM 1996.
Buffer Sizing for Congested Internet Links Chi Yin Cheung Cs 395 Advanced Networking.
NCKU CSIE CIAL1 Principles and Protocols for Power Control in Wireless Ad Hoc Networks Authors: Vikas Kawadia and P. R. Kumar Publisher: IEEE JOURNAL ON.
Choosing Beacon Periods to Improve Response Times for Wireless HTTP Clients Suman Nath Zachary Anderson Srinivasan Seshan Carnegie Mellon University.
Distributed Priority Scheduling and Medium Access in Ad Hoc Networks Distributed Priority Scheduling and Medium Access in Ad Hoc Networks Vikram Kanodia.
Energy-Efficient Design Some design issues in each protocol layer Design options for each layer in the protocol stack.
1 Emulating AQM from End Hosts Presenters: Syed Zaidi Ivor Rodrigues.
On Multi-Path Routing Aditya Akella 03/25/02. What is Multi-Path Routing?  Dynamically route traffic Multiple paths to a destination Path taken dependant.
Component-Based Routing for Mobile Ad Hoc Networks Chunyue Liu, Tarek Saadawi & Myung Lee CUNY, City College.
Power saving technique for multi-hop ad hoc wireless networks.
1 A State Feedback Control Approach to Stabilizing Queues for ECN- Enabled TCP Connections Yuan Gao and Jennifer Hou IEEE INFOCOM 2003, San Francisco,
Empirical Analysis of Transmission Power Control Algorithms for Wireless Sensor Networks CENTS Retreat – May 26, 2005 Jaein Jeong (1), David Culler (1),
Jennifer Rexford Princeton University MW 11:00am-12:20pm Wide-Area Traffic Management COS 597E: Software Defined Networking.
Cutting the Electric Bill for Internet-Scale Systems Andreas Andreou Cambridge University, R02
Energy Aware Network Operations Authors: Priya Mahadevan, Puneet Sharma, Sujata Banerjee, Parthasarathy Ranganathan HP Labs IEEE Global Internet Symposium.
Not All Microseconds are Equal: Fine-Grained Per-Flow Measurements with Reference Latency Interpolation Myungjin Lee †, Nick Duffield‡, Ramana Rao Kompella†
UCSC 1 Aman ShaikhICNP 2003 An Efficient Algorithm for OSPF Subnet Aggregation ICNP 2003 Aman Shaikh Dongmei Wang, Guangzhi Li, Jennifer Yates, Charles.
EXPLOITING VOIP SILENCE FOR WIFI ENERGY SAVINGS IN SMART PHONES Andrew J. Pyles 1, Zhen Ren 1, Gang Zhou 1, Xue Liu 2 1 College of William and Mary, 2.
Folklore Confirmed: Compiling for Speed = Compiling for Energy Tomofumi Yuki INRIA, Rennes Sanjay Rajopadhye Colorado State University 1.
Fundamental Lower Bound for Node Buffer Size in Intermittently Connected Wireless Networks Yuanzhong Xu, Xinbing Wang Shanghai Jiao Tong University, China.
DaVinci: Dynamically Adaptive Virtual Networks for a Customized Internet Jennifer Rexford Princeton University With Jiayue He, Rui Zhang-Shen, Ying Li,
OPTIMAL SERVER PROVISIONING AND FREQUENCY ADJUSTMENT IN SERVER CLUSTERS Presented by: Xinying Zheng 09/13/ XINYING ZHENG, YU CAI MICHIGAN TECHNOLOGICAL.
Enhanced power efficient sleep mode operation for IEEE e based WiMAX Shengqing Zhu, and Tianlei Wang IEEE Mobile WiMAX Symposium, 2007 IEEE Mobile.
Wei Gao1 and Qinghua Li2 1The University of Tennessee, Knoxville
Maintaining Performance while Saving Energy on Wireless LANs Ronny Krashinsky Term Project
A Distributed Energy Saving Approach for Ethernet Switches in Data Centers Weisheng Si 1, Javid Taheri 2, Albert Zomaya 2 1 School of Computing, Engineering,
BMAC - Versatile Low Power Media Access for Wireless Sensor Networks.
Energy Efficient Digital Networks Rich Brown Lawrence Berkeley National Laboratory Presentation to DOE State Energy Advisory Board Meeting August 14, 2007.
Mohamed Hefeeda 1 School of Computing Science Simon Fraser University, Canada Video Streaming over Cooperative Wireless Networks Mohamed Hefeeda (Joint.
1 Power-Aware Routing in Mobile Ad Hoc Networks S. Singh, M. Woo and C. S. Raghavendra Presented by: Shuoqi Li Oct. 24, 2002.
ACN: RED paper1 Random Early Detection Gateways for Congestion Avoidance Sally Floyd and Van Jacobson, IEEE Transactions on Networking, Vol.1, No. 4, (Aug.
Computer Networks with Internet Technology William Stallings
High-speed TCP  FAST TCP: motivation, architecture, algorithms, performance (by Cheng Jin, David X. Wei and Steven H. Low)  Modifying TCP's Congestion.
An Adaptive, High Performance MAC for Long- Distance Multihop Wireless Networks Presented by Jason Lew.
An Energy Efficient MAC Protocol for Wireless LANs Eun-Sun Jung Nitin H. Vaidya IEEE INFCOM 2002 Speaker :王智敏 研二.
An optimal power-saving class II for VoIP traffic and its performance evaluations in IEEE e JungRyun Lee School of Electrical and Electronics Eng,Chung-Ang.
Queueing and Active Queue Management Aditya Akella 02/26/2007.
Minimizing Energy Consumption in Sensor Networks Using a Wakeup Radio Matthew J. Miller and Nitin H. Vaidya IEEE WCNC March 25, 2004.
ELECTIONEL ECTI ON ELECTION: Energy-efficient and Low- latEncy sCheduling Technique for wIreless sensOr Networks Shamim Begum, Shao-Cheng Wang, Bhaskar.
Multi-Power-Level Energy Saving Management for Passive Optical Networks Speaker: Chia-Chih Chien Advisor: Dr. Ho-Ting Wu Date: 2015/03/25 1.
An Energy Efficient MAC Protocol for Wireless LANs, E.-S. Jung and N.H. Vaidya, INFOCOM 2002, June 2002 吳豐州.
Loss-Bounded Analysis for Differentiated Services. By Alexander Kesselman and Yishay Mansour Presented By Sharon Lubasz
Adaptive Sleep Scheduling for Energy-efficient Movement-predicted Wireless Communication David K. Y. Yau Purdue University Department of Computer Science.
A+MAC: A Streamlined Variable Duty-Cycle MAC Protocol for Wireless Sensor Networks 1 Sang Hoon Lee, 2 Byung Joon Park and 1 Lynn Choi 1 School of Electrical.
GreenCloud: A Packet-level Simulator of Energy-aware Cloud Computing Data Centers Dzmitry Kliazovich ERCIM Fellow University of Luxembourg Apr 16, 2010.
Smart Sleeping Policies for Wireless Sensor Networks Venu Veeravalli ECE Department & Coordinated Science Lab University of Illinois at Urbana-Champaign.
CprE 458/558: Real-Time Systems (G. Manimaran)1 CprE 458/558: Real-Time Systems Energy-aware QoS packet scheduling.
Toward Reliable and Efficient Reporting in Wireless Sensor Networks Authors: Fatma Bouabdallah Nizar Bouabdallah Raouf Boutaba.
William Stallings Data and Computer Communications
Routing Protocols to Maximize Battery Efficiency
Energy Aware Network Operations
Topics discussed in this section:
Green cloud computing 2 Cs 595 Lecture 15.
A Study of Group-Tree Matching in Large Scale Group Communications
CONGESTION CONTROL.
Ultra-Low Duty Cycle MAC with Scheduled Channel Polling
Amogh Dhamdhere, Hao Jiang and Constantinos Dovrolis
High Throughput Route Selection in Multi-Rate Ad Hoc Wireless Networks
Reducing Network Energy Consumption via Sleeping and Rate Adaptation
Reducing Total Network Power Consumption
Chih-Hsun Chou Daniel Wong Laxmi N. Bhuyan
Presentation transcript:

Reducing Network Energy Consumption via Sleeping and Rate Adaptation

2 Authors: Sergiu Nedevschi UC Berkeley & Intel Research Lucian Popa (UC Berkeley) Sylvia Ratnasamy (Intel Research) Gianluca Iannaccone (Intel Research) David Wetherall (U Washington & Intel Research) My Name: Anand Seetharam

3 Motivation Network energy consumption a growing issue – Equipment increasingly power-hungry (power density) – Rising energy costs (significant fraction of TCO) – Environmental concerns Energy Efficient Ethernet Taskforce (IEEE az) – Focuses on saving network energy for Ethernet

NetworkUtilization AT&T switched voice 33% Internet Links 15% Private line networks 3-5% LANs 1% “Data networks are lightly utilized, and will stay that way” A. M. Odlyzko, Review of Network Economics, 2003 Networks are provisioned for peak-load – phone network needs to work on 1 st JAN, at 12AM Average utilization is low: Opportunity

5 Energy consumption proportional to capacity, not actual utilization!! – Idle energy consumption is high – For example, a Cisco GSR linecard draws: [Chabarek etal, INFOCOM08] ~ 80W idle ~ 90W fully loaded Most energy consumed by networks is wasted! Goal: Make network energy consumption reflect utilization levels, not peak provisioning

6 Idea Key Idea: Let network equipment sleep for brief periods or slow down when lightly loaded to save energy Inspiration: Use of sleep and performance states in PCs, processors Rationale: E ~= p idle T idle + p active T active Assumptions: We assume support for sleep/performance states in NICs, linecards, switches, and routers and consider how to best use them Depend on: – Type/extent of hardware support for sleep and performance states – Careful use of these states to protect performance and maximize savings Sleeping reduces idle energy Slowing down reduces both

7 Outline 1.Key questions and method 2.Sleeping 3.Rate adaptation (slowing down) 4.Sleep vs. Rate adaptation

8 1. Key questions and method How much energy can we save without compromising performance? Can we realize these savings with practical schemes? Methodology: 1.Model hardware support for sleep and rate adaptation 2.Evaluate savings/performance with simulations (ns) Abilene and Intel topologies and their traffic workloads 3.Look for (unrealistic) bounds as well as practical schemes

9 Model Single sleep state with power p sleep << p idle δ: transition period (ms) Timer or activity-driven wakeup Interfaces sleep independently Metrics Energy savings in % time asleep Performance in loss and max delay 2. Sleeping states time power p idle p sleep δ (sleep) (idle)

10 Packets over a link: sleep time depends on: Buffer and burst: When can a link sleep? time δ Transition time Periods of sleep δδ δ δ time Sleep

11 Making sleep gaps on links with buffer & burst (B&B) Basic idea: use limited buffering at ingress to create predictable and useful sleep gaps (>2δ); do once, adds bounded delay t=3 t=B+3 t=2B+3 2 ms 5 ms 20 ms t=1 t=B+1 t=8 t=B+8 t=28 t=B+28 t=2B+28 R1 R2R3

12 Coordination among ingresses Basic idea: align bursts/gaps on links in networks by adjusting relative timing phase of different ingresses 8 ms 3 ms t+5, t+5+B,… t, t+B,… coordinate burst times to align in the network R I1 I2

13 Potential for savings with sleep (optB&B) “ perfect” coordination not generally possible 1ms 2ms 15ms 20ms t1 t2 Upper bound (optB&B): Global search to find ingress transmission times that maximize network-wide sleep I1 R1 R2 t1 + 1ms = t2 + 20ms t1 + 15ms = t2 + 2ms

14 Potential benefits of sleeping A little shaping can get most of the utilization gain Abilene, transition time=1 ms, B=10 ms Upper bound without buffering/shaping Upper bound for any scheme idle (bound) WoA (pareto) WoA (CBR) optB&B(CBR) Upper bound with buffering/shaping

15 Practical sleeping algorithm (practB&B) 1.Ingress buffers and transmits packets in a bunch every Bms 2.Within bunch, packets are organized by egress 3.Router interfaces wake to process bursts 4.Router interfaces sleep if start of next burst is >2δ ms away

16 No coordination (practB&B) Practical algorithm realizes most of the benefit Abilene, transition time=1 ms, B=10 ms

17 Impact of sleeping on delay No added loss; added delay ~ bounded by B ms Abilene, transition time=1 ms 98 th percentile delay (ms)

18 Impact of sleep: Any Losses? No additional losses are incurred until utilizations come close to saturating some links. Losses greater than 0.1% occur at Abilene, network utilization=5% SchemeUtilization Default41% B = 10ms38% B = 25ms36%

19 Impact of sleep transition time Quick transitions (preferably < 1 ms ) needed Abilene, network utilization=5%

20 Outline 1.Key questions and method 2.Sleeping 3.Rate adaptation (slowing down) 4.Sleep vs. Rate adaptation

21 3. Rate adaptation states Model N performance states Rates r 1, …, r n and p i < p i+1 δ : transition period (ms) Interfaces can rate-adapt independently Metrics Energy savings in average rate reduction Performance in loss and max delay time power p i+1 pipi δ (1G) (100M)

22 Using performance states Optimal algorithm: ideal service curve follows shortest Euclidean distance. bytes arriving at router bytes leaving router service rate Basic idea: decrease rate as much as possible without introducing more than than d ms per hop

23 Practical rate adaptation (practRA) Idea: lower rate if doing so will maintain minimal queuing delay (of at most d ms); aggressively increase rate to clear buildup Algorithm: r f : estimated arrival rate as average (EWMA) of past arrivals q: current queue size d: target maximum queuing delay r i : current link operating rate Rules: 1.increase to r i+1 iff (q/r i > d) OR (δr f +q)/r i+1 > (d- δ) 2.decrease to r i-1 iff (q = 0) AND (r f < r i-1 ) –duration since last rate change > k δ (k=4) Leave headroom for transition time Avoid frequent changes

24 Benefits of rate adaptation Abilene, transition time δ =1 ms, d =3 ms Upper bound for any scheme Practical rate adaptation close with uniform rates Far with exponential rates Added delay < d * (#hops) No observed packet loss

25 Outline 1.Key questions and method 2.Sleeping 3.Rate adaptation (slowing down) 4.Sleep vs. Rate adaptation

26 Models of future power profiles p active = C + fn(rate) p idle = C + β fn(rate) p sleep = μ p idle (r max ) Fraction of power that doesn’t scale with rate Idle/Active Workload Ratio Rate scaling function fn(rate) ~ rate frequency scaling fn(rate) ~ rate 3 dynamic voltage scaling Power reduction using sleep

27 Sleeping and rate adaptation (DVS-r 3 )

28 Sleeping and rate adaptation (Frequency Scaling -r)

29 Observations The authors say “Hence to avoid complex interactions, we consider that the whole network, or at least well-defined components of it, run either rate adaption or sleep” But both schemes can be combined to give better results. For eg: In rate adaptation one can try to put the links to sleep instead of keeping them in the idle state.

30 Observations When rate adaptation is done using frequency scaling the authors themselves say that for values (C=0.3 and β =0.1) and (C=0.3 and β =0.8) the savings obtained are poor and add little additional information. My observation is that rate adaptation (frequency scaling) gives no gain in terms of energy.

31 Thank you. Questions?