Quantifying the Impact of Edge Computing on Mobile Applications

Slides:



Advertisements
Similar presentations
Context-awareness, cloudlets and the case for AP-embedded, anonymous computing Anthony LaMarca Associate Director Intel Labs Seattle.
Advertisements

Eduardo Cuervo - Duke Aruna Balasubramanian - U Mass Amherst Dae-ki Cho - UCLA Alec Wolman, Stefan Saroiu, Ranveer Chandra, Paramvir Bahl – Microsoft Research.
Cost-Based Cache Replacement and Server Selection for Multimedia Proxy Across Wireless Internet Qian Zhang Zhe Xiang Wenwu Zhu Lixin Gao IEEE Transactions.
Supporting Cooperative Caching in Disruption Tolerant Networks
The case for VM based Cloudlets in Mobile Computing
AMES-Cloud: A Framework of Adaptive Mobile Video Streaming and Efficient Social Video Sharing in the Clouds 作者:Xiaofei Wang, MinChen, Ted Taekyoung Kwon,
Extending the Capacity of Mobile Devices Through Cloud Offloading Francisco Airton – PhD Student 04 of may, 2014 Workshop MoDCS
Asaf Cidon. , Tomer M. London
Ningning HuCarnegie Mellon University1 Optimizing Network Performance In Replicated Hosting Peter Steenkiste (CMU) with Ningning Hu (CMU), Oliver Spatscheck.
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.
GSM.. Onno W. Purbo Trend Global-nya.. GSM  GSM/GPRS  EDGE  WCDMA CDMA  CDMA 1X  CDMA % GSM/GPRS Market Share 2005.
Kiryong Ha ∗, Padmanabhan Pillai†, Wolfgang Richter ∗ Yoshihisa Abe ∗, Mahadev Satyanarayanan ∗ *Carnegie Mellon University, †Intel Labs 6/27/2013 Just-in-Time.
1/30/2015 Just-in-Time Virtual Machine Provisioning for Cloud Offload Kiryong Ha Carnegie Mellon University.
Forwarding Redundancy in Opportunistic Mobile Networks: Investigation and Elimination Wei Gao 1, Qinghua Li 2 and Guohong Cao 3 1 The University of Tennessee,
Efficient and Robust Computation of Resource Clusters in the Internet Efficient and Robust Computation of Resource Clusters in the Internet Chuang Liu,
Choosing Beacon Periods to Improve Response Times for Wireless HTTP Clients Suman Nath Zachary Anderson Srinivasan Seshan Carnegie Mellon University.
The Convergence of Mobile and Cloud Computing Ramesh Govindan University of Southern California 1.
Eyal de Lara Department of Computer Science University of Toronto.
On Exploiting Dynamic Execution Patterns for Workload Offloading in Mobile Cloud Applications Wei Gao, Yong Li, and Haoyang Lu The University of Tennessee,
Bandwidth Measurements for VMs in Cloud Amit Gupta and Rohit Ranchal Ref. Cloud Monitoring Framework by H. Khandelwal, R. Kompella and R. Ramasubramanian.
ThinkAir: Dynamic Resource Allocation and Parallel Execution in Cloud for Mobile Code Offloading Sokol Kosta, Pan Hui Deutsche Telekom Labs, Berlin, Germany.
 Energy Results: Memory Assistant Arcade Game  Performance Results:  Response Time ▪ Memory assistant: 17.3 sec -> 1.5 sec ▪ Arcade game: 6 FPS -> 13.
MOBILE CLOUD COMPUTING
GENI-related research activities of CSE, Aalto Zhonghong Ou Post-doc researcher Department of Computer Science and Engineering.
Characterizing and Modeling the Impact of Wireless Signal Strength on Smartphone Battery Drain Ning Ding Xiaomeng Chen Abhinav Pathak Y. Charlie Hu 1 Daniel.
Early Implementation Experience with Wearable Cognitive Assistance Applications Zhuo Chen, Lu Jiang, Wenlu Hu, Kiryong Ha, Brandon Amos, Padmanabhan Pillai,
Dynamic VM Synthesis for Cloudlet -ISTC Retreat Poster- Kiryong Ha, Padmanabhan S Pillai, Mahadev Satyanarayanan.
Multi-Criteria Routing in Pervasive Environment with Sensors Santhanakrishnan, G., Li, Q., Beaver, J., Chrysanthis, P.K., Amer, A. and Labrinidis, A Department.
Mobile Cloud
Autonomic scheduling of tasks from data parallel patterns to CPU/GPU core mixes Published in: High Performance Computing and Simulation (HPCS), 2013 International.
Reducing Energy Consumption in Human- centric Wireless Sensor Networks The 2012 IEEE International Conference on Systems, Man, and Cybernetics October.
Dual-Region Location Management for Mobile Ad Hoc Networks Yinan Li, Ing-ray Chen, Ding-chau Wang Presented by Youyou Cao.
Embedded System Lab. 정범종 A_DRM: Architecture-aware Distributed Resource Management of Virtualized Clusters H. Wang et al. VEE, 2015.
Eduardo Cuervo – Duke University Aruna Balasubramanian - University of Massachusetts Amherst Dae-ki Cho - UCLA Alec Wolman, Stefan Saroiu, Ranveer Chandra,
Power Controlled Network Protocols for Multi- Rate Ad Hoc Networks Pan Li +, Qiang Shen*, Yuguang Fang +, and Hailin Zhang # +: EE, Florida University.
Shuo Deng, Ravi Netravali, Anirudh Sivaraman, Hari Balakrishnan
Project Topics ECE 591. Project 1: Localization through Wi-Fi and Wireless Camera WIFI localization: Wireless Camera: Goal: Understand RF based localization.
Simplifying Cloud Connectivity for Your Clients Presenter: Tom SharkeyTom Sharkey December 8,
Reliability of Wireless sensors with code attestation for intrusion detection Ing-Ray Chen, Yating Wang, Ding-Chau Wang Information Processing Letters.
Application-Aware Traffic Scheduling for Workload Offloading in Mobile Clouds Liang Tong, Wei Gao University of Tennessee – Knoxville IEEE INFOCOM
Software-Defined Architecture for Mobile Clouds in Device-to-Device Communication Muhammad Usman; Anteneh A. Gebremariam; Fabrizio Granelli; Dzmitry Kliazovich.
On the Statistical Multiplexing Gain of Virtual Base Stations Pools Lewis (Jingchu) LIU, Sheng ZHOU, Jie GONG, Zhisheng Niu, Tsinghua University, Beijing,
Resource Allocation in Mobile Cloud Computing. Motivation ›Apart from offloading, resource provisioning has emerged to be an important issue. › Resource.
Kiryong Ha ∗, Padmanabhan Pillai†, Grace Lewis‡, Soumya Simanta‡, Sarah Clinch§, Nigel Davies§, and Mahadev Satyanarayanan ∗ ∗ Carnegie Mellon University,
A Hierarchical Edge Cloud Architecture for Mobile Computing IEEE INFOCOM 2016 Liang Tong, Yong Li and Wei Gao University of Tennessee – Knoxville 1.
A NOVEL PREFETCHING METHOD FOR SCENE- BASED MOBILE SOCIAL NETWORK SERVICE 作者 :SONG LI, WENDONG WANG, YIDONG CUI, KUN YU, HAO WANG 報告者 : 饒展榕.
Edge Computing ——vision, challenges and promise. 物联网云计算.
Enterprise Network Sourcing & Price: Cost Considerations for WAN Design Brianna Boudreau Senior Analyst.
Lecture 1: Getting Ready
Cost Effectively Deploying of Relay Stations (RS) in IEEE 802
Huber Flores Social-aware Hybrid Mobile Offloading A contribution for edge and fog computing? Huber Flores
Architecture and Algorithms for an IEEE 802
Introduction to Edge Computing
Mark Claypool, Feng Li and Jae Chung
Mobile edge computing Report by Weiqing huang.
Bandwidth Measurements for VMs in Cloud
Di Zhang, Yuezhi Zhou, Xiang Lan, Yaoxue Zhang, Xiaoming Fu
Speaker: I-LUN LEE ADVISOR: DR. HO-TING WU
Cross-Layer Optimization for State Update in Mobile Gaming
Authors: Ing-Ray Chen; Yating Wang Present by: Kaiqun Fu
Energy Efficient Scheduling in IoT Networks
Speaker: Jin-Wei Lin Advisor: Dr. Ho-Ting Wu
Congestion Control in SDN-Enabled Networks
Congestion Control in SDN-Enabled Networks
Iterative Optimization of Registration and Paging Policies
SD-WAN PoC Final Demo Aitor Zabala (TELCA) Date.
© 2016 Global Market Insights, Inc. USA. All Rights Reserved Ethernet Storage Market Size Growth During Forecast Period.
Ultra Low Networking Latency
No-Jump-into-Latency in China's Internet
Presentation transcript:

Quantifying the Impact of Edge Computing on Mobile Applications 25 min including Q&A Wenlu Hu, Ying Gao, Kiryong Ha, Junjue Wang, Brandon Amos, Zhuo Chen, Padmanabhan Pillai (Intel), Mahadev Satyanarayanan

Different Names of Edge Computing Mobile Edge Computing Fog Computing Micro Data Centers Cloudlets 8/4/2016

One-hop wireless network What is a cloudlet? Mobile Device Cloud Cloudlet One-hop wireless network Internet Median RTT (ms) Cloud Cloudlet WiFi 75 1 LTE 121 15 What if an application is both compute-intensive and latency-critical? Average RTT to optimal cloud = 74ms (measured from 260 global vantage points) [1] [1] A. Li, X. Yang, S. Kandula, and M. Zhang. Cloudcmp:comparing public cloud providers. In Proceedings of the 10th annual conference on Internet measurement, page 1-14. ACM, 2010. Latency Bandwidth Scalability 95% RTT (ms) Cloud Cloudlet WiFi 92 11 LTE 135 21 8/4/2016

Of course cloudlets help. But by how much?

Our Contribution Quantified the difference between cloudlet and cloud with a variety of applications with WiFi and LTE network. 8/4/2016

List of Evaluations WiFi offloading – response time LTE offloading – response time WiFi offloading – energy LTE offloading – energy tradeoff Effects of interactivity Effects of radio link management 8/4/2016

List of Evaluations WiFi offloading – response time LTE offloading – response time WiFi offloading – energy LTE offloading – energy tradeoff Effects of interactivity Effects of radio link management 8/4/2016

Experimental Setup - Application Compute-intensive, latency-critical 3 pre-partitioned applications Face Recognition Mobile Augmented Reality Fluid Simulation 3 dynamically partitioned applications Linpack CPU Benchmark PI Benchmark 8/4/2016

Experimental Setup - Network EC2 West EC2 Europe EC2 East EC2 Asia Internet No offload WiFi cloudlet WiFi cloud LTE cloud LTE cloudlet Cloudlet Cloudlet In-lab Base Station WiFi Access Point Commercial Cell Tower LTE (cloud) WiFi (cloudlet and cloud) LTE (cloudlet) Mobile Device 8/4/2016

List of Evaluations WiFi offloading – response time LTE offloading – response time WiFi offloading – energy LTE offloading – energy tradeoff Effects of interactivity Effects of radio link management 8/4/2016

WiFi Offloading – Response Time Cumulative Distribution Function (CDF) Typical cloud, EC2-West 8/4/2016 (FACE)

WiFi Offloading – Response Time Valid across applications MAR Fluid Linpack CPU Benchmark PI Benchmark 8/4/2016

List of Evaluations WiFi offloading – response time LTE offloading – response time WiFi offloading – energy LTE offloading – energy tradeoff Effects of interactivity Effects of radio link management 8/4/2016

LTE Offloading – Response Time (FACE) (FACE) 8/4/2016

List of Evaluations WiFi offloading – response time LTE offloading – response time WiFi offloading – energy LTE offloading – energy tradeoff Effects of interactivity Effects of radio link management 8/4/2016

WiFi Offloading - Energy Energy consumed by the mobile device Offloading to cloud usually saves energy Offloading to cloudlet saves much more Offload None Cloudlet Cloud Face (J/query) 12.4 6.1 Fluid (J/query) 0.8 0.9 MAR (J/query) 5.4 4.3 Linpack (J/run) 40.3 16.9 CPU (J/run) 9.6 5.8 PI (J/run) 129.7 107.6 Offload None Cloudlet Cloud Face (J/query) 12.4 2.6 6.1 Fluid (J/query) 0.8 0.3 0.9 MAR (J/query) 5.4 0.6 4.3 Linpack (J/run) 40.3 13.0 16.9 CPU (J/run) 9.6 5.7 5.8 PI (J/run) 129.7 53.9 107.6 8/4/2016

More Details in the Paper LTE offloading – energy tradeoff Effects of interactivity Effects of radio link management 8/4/2016

Conclusion Choice of offloading site is important Cloudlet wins significantly over cloud Valid across applications and networks New genre of applications compute-intensive and latency-critical Cloudlet 8/4/2016

Thank you! Wenlu Hu Ying Gao Kiryong Ha Junjue Wang Brandon Amos Zhuo Chen Padmanabhan Pillai (Intel) Mahadev Satyanarayanan 8/4/2016