Green Wireless Technology Panel Presentation Professor Sandeep K. S. Gupta IMPACT (Intelligent Mobile Pervasive Autonomic Computing & Technologies) LAB (http://impact.asu.edu) School of Computing, Informatics, Decision Systems Engineering Arizona State University sandeep.gupta@asu.edu 480-965-3806
Sandeep Gupta, IEEE Senior Member Heads @ School of Computing & Informatics Use-inspired, Human-centric research in distributed cyber-physical systems ID Assurance Mobile Ad-hoc Networks Pervasive Health Monitoring Criticality Aware- Systems Thermal Management for Data Centers Intelligent Container BEST PAPER AWARD: Security Solutions for Pervasive HealthCare ICISIP 2006. BOOK: Fundamentals of Mobile and Pervasive Computing, Publisher: McGraw-Hill Dec. 2004 TCP Chair http://www.bodynets.org TCP Co-Chair: GreenCom 07 http://impact.asu.edu/greencom Area Editor Also for IEEE TPDS WINET Email: Sandeep.Gupta@asu.edu; IMPACT Lab URL: http://impact.asu.edu;
Green Technology a Boon or a Curse?
What should be Green?
Two extreme examples of Greenness EEG 1. Embedded networks, e.g. Body Area Networks (BANs), consisting of embedded and inter communicating sensor devices Constrained resources Principally battery powered with limited available energy Sustainability is the main issue Un interrupted operations, e.g. sensing & communication, for long periods Required Approaches Energy efficient algorithm design Scavenging energy from the environment (e.g. human body) BP EKG SpO2 Base Station Motion Sensor Body Area Network (BAN) 2. Large scale networked systems, e.g. data centers, integrating a large number of computing resources for service provisioning Power density increases Circuit density increases by a factor of 3 every 2 years Energy efficiency increases by a factor of 2 every 2 years Effective power density increases by a factor of 1.5 every 2 years [Keneth Brill: The Invisible Crisis in the Data Center] Total Cost of Ownership (TCO) rising DataCenter TCO doubles every three years Cooling the data center can cost up to half of the total electricity bill Power Usage Efficiency has to be reduced Required Approaches [Uptime Institute] Energy efficient thermal aware resource management Coordinated dmanagement of computing and cooling resources Data Center
Green Data Centers System Requirements Equipment Safety Equipment operating temperature should be within a manufacturer specified redline temperature Service Level Agreements (SLAs) The throughput and turn around time should meet the user requirements Energy Issues Computing Energy: Trade off with meeting the SLAs Cooling Energy: Trade off with equipment safety Stems from the data center thermal issues Thermal Issues Heat recirculation Hot air from the equipment air outlets is fed back to the equipment air inlets Hot spots Effect of Heat trecirculation Areas in the data center with alarmingly high temperature Consequence Cooling has to be set very low to have all inlet temperatures within the redline for safety Typical Data Center Design Green data centers call for Coordinated Thermal aware resource management to reduce cooling and computing energy consumption while meeting the equipment safety and SLAs
Ecosystem of Datacenters Different task assignments lead to different power consumption distributions Different power consumption distributions lead to different temperature distributions Different temperature distributions lead to different total energy costs 100 3500 80 3000 100 60 40 20 0 2500 2000 1500 1000 500 0 80 60 40 20 0 S3 1 2 3 4 5 6 7 8 9 10 11 12 S1 S3 S5 1 2 3 4 5 S1 S2 S3 1 2 3 4 S1 S2 Server load Power consumption Temperature distribution distribution distribution Energy cost 7
Coordinated Resource Management for Green Data Centers Energy reduction reduction in data centers has three directions 1. Thermal aware workload management Schedule and place jobs such that low power servers having less impact on the cooling demand are used 2. Server Power Management Operate under utilized servers at low power states (e.g. turn off idle servers) 3. Cooling Management Dynamicallysetthe the coolingto the highesttemperatures temperatures that meets the equipment safety There is a spatio temporal workload schedule that minimizes the total energy (cooling + computing) demand. Find it and perform dynamic cooling management and server power management with it to minimize the total energy consumption while meeting the SLAs and equipment safety! Energy Savings Over Current Practice
Online Data Center Thermal Management Power Characterization Characterize the power consumption of a given workload (CPU, memory, disk etc) on a given server machine Data Center-Level Thermal Models To enable on-line real-time thermal-aware job scheduling fast (analytical, non CFD based) non-evasive (machine-learning) Thermal Management Infrastructure & Services for Data Centers Data Center Monitoring Sensor Data Gathering Service Performance Monitoring Service Model the thermal impact of multicore systems Resource & Server Management Performance Monitoring OS-Level Services Non-Invasive Thermal Evaluation Policy Enforcement Thermal/Power & Performance Correlation Service Thermal Management Policy Enforcement Service Fast Thermal Evaluation Service Thermal Control Policies Thermal-aware Job Scheduling On-line job scheduling algorithm to minimize peak air inlet temperature, thus minimizing the cost of cooling. PI: Sandeep K. S. Gupta Sandeep.gupta@asu.edu Cluster Management Resource Queues Job Scheduling Service Job Queues Facility Management Cooling Control Service Air-flow Control Service http://impact.asu.edu
Green Body Area Networks (BANs) System Requirements Sustainability Uninterrupted BANoperations forlong periods essential for medical care Operational Safety Side effects of BAN operations (e.g. heat dissipation) should be within a threshold Security Energy Issues Sensitive medical data collection should be secure as per legal requirements (HIPAA) Secure wireless inter sensor communication is essential Computing and Communication Energy: Trade off with security and sustainability Limited battery power EEG EKG BP Base Station Motion Sensor SpO2 Typical BAN Design Sensors Base Station S i f th h b d i ti l Green BANs call for energy analysis of the computing and communication for potential scavenging to ensure sustainability and security. Scavenging energy from the human body is essential Security adds computation and communication overhead Energy analysis of security primitives important Reduced power consumption in the sensors can enable safety by reducing potential heat dissipation
Sustainable Physiological Value based Security for BANs Physiological i lsignal lbased Key Physiological i l signal Physiological i l signal Agreement (PKA) features Perform key agreement by combining f cryptographic primitives with signal Sender processing key hide features f Un hide Receiver key 0.06 PKA Power Profile (Radio-ON, Vault Size = 5000) 0.05 Receiver(Radio ON) 0.04 004 Sender(Radio ON) 0.03 0.02 0.01 1 Sensing 2 FFT 3 Peak 4 Poly Gen + 5 Add Chaff 6 Vault Tx/Rx 7 Lagrangian 8 Ackn Tx/Rx 9 + Quant Eval Interpolation Scavenging Technique Source The max power required by PKA Body Heat Latent heat of (58 mw) low enough to be vaporization of sustained by prominent energy perspiration scavenging techniques Respiration Chest Expansion while Venkatasubramanian et al Green and Sustainable Security Solutions for BANs, BSN 08 Power Gain 200mW 320mW Respiration Chest Expansion while ~420mW breathing Ambulation Arm & Leg Movement 1.W 1.6W Photovoltaic Cells Photovoltaic Cells 100mW/cm 2
Safety: Minimize Tissue Heating Medicalsensors implanted/worn by human need to be safe. Sensor activity causes heating in the tissue. Heating caused by RF inductive i powering Radiation from wireless communication Power dissipation of circuitry Goal: minimize tissue heating. Two solutions: Communication scheduling for minimizing thermal effects: Rotate cluster leader balance energy usage + distribute heat dissipation Thermal aware routing: route around thermal hotspots Heating Zone Cluster leader Tissue Blow-up
CE UCE Requirement FCC Regulation SAR = σ E 2 / ρ (W/kg) E = induced Electric Field Ρ = tissue density σ = electric conductivity of tissue IEEE Requirement (1g Tissue) Whole Body Average Whole Body Average Solution Random selection may lead to higher SAR = 0.4W/Kg SAR =.08W/Kg temperature rise Similar to Traveling salesman problem but with dynamic metric Heuristic: Leader selection based on sensor location, rotation history 5 1 Peak Local Peak Local 2 4 3 3 (a) Ideal Rotation (b) Nearest Rotation 5 BSN Scheduling ε 1 SAR = 8W/Kg SAR = 1.6W/Kg 2 5 4 4 3 (c) Farthest Rotation Four Approaches 1 Temperature Rise C 2 System Model Consider only one cluster 2D Model Rotate cluster head - dist. energy consump. reduce heating Cluster Leader Medium 2(Body tissue) 2, µ 2, σ 2 Transmitted Wave depth d Control Volume and a cluster of biosensors Medium 1(free space) ε 1, µ 1, σ 1 Incident Plane Wave with power P 0 Reflected Wave RF Powering Source Temperature Rise: Pennes Bio-heat Equation Heat Heat transfer Heat by Heat transfer Heat by power Heat by accumulated by conduction radiation by convection dissipation metabolism Results FDTD + enumeration FDTD + Genetic e Algorithm TSP +enumeration TSP + Genetic Algorithm Comparative Result Optimal Near Optimal Optimal Near Optimal FDTD + enumeration 0.11 Temperature Mean ± Deviation 0.1 FDTD + Genetic Mean 0.09 Algorithm 0.08 TSP + enumeration 0.07 TSP +Genetic 0.06 Worst Dynamic Manual Genetic Optimal Algorithm Q. Tang, N. Tummala, S. K. S. Gupta, and L. Schwiebert, Communication scheduling to minimize thermal effects of implanted biosensor networks in homogeneous tissue, Proc of IEEE Transactions of Biomedical Engineering Temperatu ure 720960 hrs (est.) 100 hrs (est.) 7.6 hrs 5 min Temp rise in sensor surroundings
BAN Development Using Model Based Approach Advantage: Substantial cost reduction for development - through taming complexity reduced time for development etc complexity, etc. leading to profitability and hence sellability of product.
What should be Green? GREEN = safe (i.e. minimum i operational fil failure) + secure (i.e. minimum vulnerability to threats) + sustainable (i.e. uninterrupted operations) + sellable (i.e. both affordable and profitable) while providing the required services
Current and Future Work@IMPACT Tool Development BAND AiDe: Body Area Network Analysis and Design Tool BlueTool: Energy Efficient Data Center analysis anddesigntooldesign Research Issues Designing green systems for critical operations while ensuring resource availability (e.g. providing required energy to the pacemaker from the scavenging sources on detection of heart attack) Investing interdependencies d i bt between safety, ft security sustainability and sellability goals. http://impact.asu.edu