Multitier Cloud, 5G, Smart Cities · Multitier Cloud Computing, 5G Networks & Smart Cities Alberto...
Transcript of Multitier Cloud, 5G, Smart Cities · Multitier Cloud Computing, 5G Networks & Smart Cities Alberto...
Multitier Cloud Computing, 5G Networks & Smart Cities
Alberto Leon-Garcia and Hadi Bannazadeh
University of Toronto, Canada [email protected]
Acknowlegments: Eric Lin, Petros Spachos, Ali Tizghadam
5G is Different! 5G is about Applications at the Edge • Yes, Higher speed & capacities, but also • Things, sensors, actuators, computing at the edge • Lower latency • Heterogeneity • Hugh scale • Adaptability, flexibility, programmability • Virtualization
Multitier cloud computing for E2E applications • Situational awareness • Sensing, Intelligence, Open Interfaces • Essential for Smart City applications
What are SAVI and CVST? • Smart Applica,ons on Virtual Infrastructures
– NSERC Strategic Research Network – “Canada’s GENI” – Partnership of industry, academia, research, and educaAon networks – Research and Testbed on Applica,on Pla;orms – Supports research of 100 graduate / postgraduate students, 70 interns – Applica,on pla;orm converging cloud compu,ng, SDN, SDR
• Connected Vehicles and Smart Transporta,on – Ontario Research Fund – ApplicaAon PlaJorm for Smart TransportaAon – Situa,onal Awareness for Greater Toronto Area Traveller Network
– runs on top of SAVI testbed using a Future Internet protocol
SAVI’s View of Future Clouds • MulA-‐Tiered Cloud: Core, Smart Edges, Access
Multitier Cloud supports applications • Core Datacenter: massive data centers • Smart Edge: Agile data centers residing closer to end users
contains heterogeneous resources & virtualizes access • vCPE/Sensors: virtual Customer Premises Edge (vCPE)
resources support Internet of Things • Application Platform: orchestrates E2E applications on
virtualized Software-Defined-Infrastructure (SDI) leveraging situational awareness and intelligence
SAVI Testbed: Smart Edge • Smaller scale datacentres close to end-‐users
• For services requiring low-‐latency, local processing, caching • Virtualizing the access networks
• Virtualiza,on of heterogeneous resources • e.g. GPUs, FPGAs, Access Points, Sensors, etc.
• OpenBTS, virtualized WIFI, virtualized IDS/IPS
SAVI vCPE (SvCPE) • Close Proximity: Host applications serve a small number
of users in very close proximity • Virtualized: Provide virtualized resources to support
different applications • Continuous Operation: continue functioning
independently in the event of network disruptions
SAVI Testbed Network
CORE
BCNet/CANARIE/ORION
• >100 graduate students • ~1000 Cores • 12+ FPGA; 6+ GPU Systems
• 50+ TB storage • 10/1 GE fabric OpenFlow • 1 GE Layer 2 Backbone
SAVI SDI Architecture
Converged Heterogeneous Resources
Resource Controller A Resource Controller B
External Entities
Resource Type A
Resource Controller C
SDI Resource Management
System
Resource Type B Resource Type C Physical resource
Virtual resource
SDI Manager Topology Manager Monitoring & Analytics Control & Management
Open Interfaces
• Each resource type controlled by specialized controllers • Each controller communicates with logically central C&M framework
• SDI Manager, Topology Manager, and Monitoring and Analy,cs
• Exposes open interfaces for external users and enAAes
MonARCH Monitoring system • Data Collection
• Multi-layer monitoring • Open Infrastructure
• Data Storage • Monitoring as a Service (MaaS)
• Analytics • Batch and streaming analytics • Graph analysis / machine learning with
monitoring data and resources relationship
Collection
StorageAnalysis
MonArch Demo
• monarch.savitestbed.ca
SAVI and vCPE SAVI Testbed
SDI Manager and Resource Controllers
Customer Premise
SAVI Very Small Edge
SAVI Small Edge
Control
Smart Edge
SAVI vCPE Use Cases • Collection point for Internet of Things devices • Smart home and office
• sensor management • power management • security enforcement point
• Web acceleration and service delivery point • NFVs such as proxy, firewall, IDS/IPS and VPN services
Small SAVI vCPE • Supports compute and networking
virtualization • Able to host multiple applications • Connected to the SAVI Smart
Edge with VPN technology
• Has all the capabilities provided from the virtualized system in SAVI (e.g. tenant isolation)
• Capable of leveraging advanced Features of SAVI SDI (e.g. NFV Service Chaining)
Janus&SDI&Manager&OpenStack&and&OpenFlow&
Customer&Premise&
SAVI&Very&Small&Edge&
SAVI&Small&Edge&
Control&
Smart&Edge&
Monitoring CO2
• Sensor nodes: consists of a carbon dioxide sensor and radio module
• Relay nodes: responsible for forwarding any received packet toward the destination
• Control Room: destination of sensor data and data aggregation point
Sensor Deployment on SAVI vCPE • Connect the destination node of the wireless sensor
network to the vCPE machine • Connected to monitoring and analytics manager, called
MonArch • Run the control room software and the MonArch data
collection Agent in the vCPE
SAVI Smart Edge SAVI vCPE
Wireless Sensor Network
MonArch Agent
Supporting Public & Private Providers
Public App Provider:·∙ ITS Services
Private App Provider:·∙ Traveler Assistance·∙ Personalized Routing
·∙ Fleet Management
Content-based Routing (Publish/Subscribe)
Content-based Router
Clients (publisher/subscriber)
Server Farm
Computers
Computers
Workstation
Laptops
Switch
Database
BV2V
Cluster
A
C
Truck
RoadSensors
Sensing
Intelligence
Control
CVST Conceptual Pyramid
• Transportation applications • Recommendations (Routing, City planning) • What-If Analysis, Impact analysis
Application Development
• Real-time streaming analytics • OD Demands • Trending / Forecasting • KPI analysis
Intelligent Transportation Management
• Data collection • Data Integrity, cleansing • Data anonymity • Publish / subscribe
Data Dissemination
• Sensors • V2V, V2R, V2I • Cloud resource
management • Horizontal / vertical scaling
Connected Vehicles & Mobile Computing Cloud
CVST GTA Portal: Real-Time Situational Awareness
• www.portal.cvst.ca
Feb 2, 2015 10am
Vision-as-a-Service
Next: Smart City Platforms • SAVI multitier cloud as foundation for smart city platforms • Edge computing is a key research challenge
• Smart Edge; IoT; 5G wireless capabilities • Smart Applications are enabled by multitier clouds
• Measurement and Monitoring • Intelligence and value-add through analytics • Smart application ecosystem through open interfaces to
enabling services • Economies of scale and synergies in interconnected
management systems: • Energy, Water, Waste, Air Quality, ICT • Transportation, Safety, Health, Buildings, Home, Industry
International Partnerships • Global Environment for Network Innovations (GENI) • White House $160 million Smart Cities Initiative • NIST-USIgnite Global Smart City Challenge • USIgnite Smart and Connected Cities project
• 15 smart cities • GENI racks, new fiber/wireless deployments • applications creation environment
Thank you Q&A