Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
1
Emerging Wireless Technologies and the Future InternetStanford Clean Slate SeminarOct 30, 2007
Prof. D. [email protected]
2
Wireless Applications
3
INTERNETINTERNET
Introduction: Wireless as the key driver for the future Internet
Historic shift from PC’s to mobile computing and embedded devices…
>2B cell phones vs. 500M Internet-connected PC’s in 2005>400M cell phones with Internet capability, rising rapidlyNew types of data devices (blackberry, PDA, iPoD) – distinctions becoming blurrySensor deployment just starting, but some estimates ~5-10B units by 2015
WirelessEdge
Network
WirelessEdge
Network
INTERNETINTERNET
~500M server/PC’s, ~100M laptops/PDA’s
~750M servers/PC’s, >1B laptops, PDA’s, cell phones, sensors
2005 2010
WirelessEdge
Network
WirelessEdge
Network
4
Wireless Applications: Mobile DataMobile data service starting to take off as mass-market serviceCurrent Internet applications on mobile devices via WLAN and 3GNetwork needs to support basic mobility features (roaming and handoff) at scale, multiple radio technologies, efficient TCP, etc.
WLAN Hot Spot
Metro Area 3G Network (EVD0 etc.)
Internet
5
Wireless Applications: Video StreamingVideo broadcast to mobile devices expected to increase in popularityNetwork needs to deal with mobility, multicasting, stream QoS, device heterogeneity, …
Metro Area 3G or Custom Broadcast Network
(e.g. MediaFlo)
Internet
McastVideo
WLANservice
Streaming Video
Video GW
6
Wireless Applications: Broadband AccessWireless being used for both urban and rural broadband servicesNetwork issues: large capacity, wireless backhaul, link quality, end-to-end TCP performance, …
500 node deployment in Portsmouth, UK
7
Wireless Applications: Mobile Content
Delivery of music/video files to mobile devices represents a major new service category
Media content, either on-demand or push (based on user profile)Large file transfer difficult over cellular opportunistic access of short-range wireless caches (“Infostations” concept, also P2P)Network needs to deal with disconnections, opportunistic delivery, content caching, ..
WINLAB’s i-mediaPrototype (2002)
8
Emerging vehicle safety, information and entertainment applicationsNetworking requirements tend to be location-aware
e.g. geocast scenario shown
Network mobilityPotentially high densityAd hoc network formation and disconnectionsV2V and V2I modes
Desired message delivery zone
(Idealized) Broadcast range
Irrelevant vehicles in radio range for few seconds
Passing vehicle,in radio range for tens of seconds
Following vehicle,in radio range for minutes
Wireless Applications: Vehicular Safety
9
Wireless Applications: Physical World with Embedded Sensors/Actuators
Vehicles with Sensors & Wireless
Hospital with Embedded Monitoring
Robotics Application
Smart Public Space
Autonomous Wireless Clusters(“ecosystems”)
Network Connectivity& Computation
Application Management & Control Software
“Human in the Loop”
To Actuators
From Sensors
Virtualized physicalworld object
Content & LocationAware Routers
Computation& Storage
Protocolmodule
Ambient interfaces
Global Pervasive Network(Future Internet)
Multiple radio standards,Cognitive radios
10
Emerging Wireless Technologies
11
Emerging Wireless Technologies: 4G Cellular and WiMax
Next-generation of cellular technology aimed at ~2015
Higher speeds ~100 Mbps+ using MIMO and OFDMA technologiesSimplifications to network architecture more IP oriented: flat network of IP base stationsDecentralized control of radio resourcesSupport for inter BTS mesh network, etc.WiMax has potential for lower cost, commodity equipment model…
12
Emerging Wireless Technologies: Ad-Hoc & Mesh Networks
Multi-hop radio (ad hoc, mesh, vehicular, sensor) technologies now entering the mainstream …
Leverages Moore’s law cost/performance gains of commodity radios such as IEEE 802.xDistributed solution with short-range radios will eventually outperform centralized (cellular) analogous to PC/mainframe evolutionInvolves new routing & discovery protocolsInteractions between lower layers (PHY, MAC) and routing in dense deploymentsProblems with TCP end-to-end model due to changing BW and channel quality
Wired Internet Infrastructure
Mesh GW or AP
Hierarchical Mesh Network
MeshRouter
13
Emerging Wireless Technologies: P2P and DTNP2P and DTN network protocols expected to migrate from niche scenarios to wider usage
Router mobilityNetwork may be disconnected at times …delay tolerant protocolsCaching and opportunistic data delivery …. In-network storageContent- and location- aware protocols
InternetMobile DTN Router
Roadway Sensors
Mobile DTN Router
Ad-HocNetwork
OpportunisticHigh-Speed Link
(MB/s)
Mobile P2P User
Static DTNRouter
14
Emerging Wireless Technologies: Sensor Nets
Mobile Internet (IP-based)
Overlay Sensor Network Infrastructure
Compute & StorageServers
User interfaces forinformation & control
Ad-Hoc Sensor Net A
Ad-Hoc Sensor Net B
Sensor net/IP gateway GW
3G/4GBTS
PervasiveApplication
Agents
Relay Node
Virtualized Physical WorldObject or Event
Sensor/Actuator
Sensor net scenarios involve:Large scaleLimited CPU speed and transmit powerIntermittent connectivity, low-speeds, ad-hoc modesLocation and content-awarenessMay involve closed loop control in real-time
ZigBee,UWB, etc.
15
Emerging Wireless Technologies: Dynamic Spectrum & Cognitive Radio
Spectrum Policy Server
Maximum Amplitudes
Frequency (MHz)
Am
plid
ue (d
Bm
)
Heavy Use
Sparse Use
Heavy Use
Medium Use
Less than 6% OccupancyLess than 6% OccupancyAtlanta
NewOrleans
SanDiego
Frequency
Time
Dynamic Spectrum ProtocolsFor Coordination
Next-gen wireless devices with dynamic spectrum capability
(fast RF scan, agile, adaptive PHY/MAC)
WiredInternet
Data Signal
Spectrum Coordination
16
Clean Slate Protocol Architecture Projects at WINLAB
17
Clean Slate Project: Cache & Forward Architecture
NSF FIND “postcards from the edge” project involving Rutgers & UMassArchitecture designed to support efficient delivery of content to mobile usersConcept based on hop-by-hop transport, storage and caching in the network
Media Server
PlanetLab Slice
Storage Caches
ORBIT Radio Grid
ORBIT Gateway
Hop-by-hop File TransferHop-by-hop
File Transfer Reliable Link Layer
File sent to multiple destinations
Media file (~10MB-GB)
Wireless AccessNetwork
AP/Gateway(CNF “P.O.”)
Wired Internet withCache & Forward Routers
18
CNF Project: Architecture ConceptsNetwork is optimized for efficient content delivery to mobile devices
Large storage (~TB) at each network nodeSingle- or multi-hop wireless access with occasional disconnectionsStrict hop-by-hop transport (…entire file sent via reliable link protocol)“Post Offices” for opportunistic delivery to mobile clientsIntegration of conventional routing with retrieval of cached content
Receiver’s Post OfficeSender’s
Post OfficeReceiver
Sender
MNMN
MN
CNF NetworkRouters with
Storage
Reliable Link Layer
19
CNF Project: Protocol StackRouting protocol with integrated support of address- and content-based modes of delivery (get “filename”, deliver ”filename” to “address”, etc.)Support for in-network caching of contentMulti-hop wireless access with disconnections and mobility at access edge (..”postbox” concept)
Core
CNF_1 CNF_6
CNF_2
MN
SenderReceiver
MNMN
CNF_3 CNF_4CNF_5
CNC_1
Link ProtocolLink Protocol
Receiver’s Post OfficeSender’s
Post Office
Link Protocol
20
CNF Project: Simulation Results – Network Topology and Simulation Parameters
GT-ITM’s transit-stub modelTransit-transit links
10000Mbps bandwidthUniform (20,23)msec delay
Transit-stub links155Mbps bandwidthUniform (20,23)msec delay
Stub-stub links1000Mbps bandwidthUniform (2,11)msec delay
15 source-destination pairs
ON State: Exponentially distributed with mean 100secOFF State: Exponentially distributed with 3600secAn ON state is always followed by an OFF state (vice versa)Initial state of the source is ONDuring ON State: Fixed file size (50MB), Poisson arrivals
S
S
D
D
S
S D
D
21
CNF Project: Simulation Results (1) – CNF vs. TCP Delay vs. Traffic Type & File Size
TCP outperforms CNF for “always on” traffic and at
low arrival rate
Advantage of TCP over CNF reduces for ON-OFF
traffic
Advantage of TCP over CNF for “always on” traffic is higher at smaller file size
Advantage of TCP over CNF for ON-
OFF traffic is lower at smaller file size
In-network storage in CNF serves to smooth traffic burstsHowever, this must be balanced against loss of pipeline gainOverall, CNF performs well relative to TCP except at low loads
22
CNF Project: Simulation Results (2) –Delay Histograms for Alternative Content Routing/Caching Strategies
“Content Broadcast” scheme has more “low delay” file transfers compared to “Caching and Capture”
When nodes exchange “content caching” information with neighbors Average Delay in locating content reduces and the reduction is more when caching is limited to fewer nodes
0.6590Broadcast, h=Max
28.5%0.92206Caching only, h=max
0.575Broadcast, h=2
23.8%0.75455Caching only, h=2
0.54013Broadcast, h=1
22.3%0.69539Caching only, h=1
ImprovementAverage DelayExperiments
23
CNF Project: TCP vs. CNF link layer
Instantaneous received rate of TCP, CLAP, UDP (noisy WLAN example)
0
1
2
3
4
5
6
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Time sequence (seconds)
Rec
eive
d M
bps
UDPTCPCLAP
Cro
ss-la
yer i
nfor
mat
ion
Conceptual Timing Diagram of CLAP
Cro
ss-la
yer i
nfor
mat
ion
Conceptual Timing Diagram of CLAP
Cross-layer aware transport protocols achieve major performance gains for wireless videoFor example, CLAP [Gopal, 2007] provides for significantly faster video on-demand downloads
CLAP filetransfer
completed
24
ORBIT Radio Grid
Mobile node trajectory
AP2
AP1
Location aware routing protocol
Clean Slate Project: Geometric Stack for Location Aware Networks
Location-aware protocol architecture project funded under NSF FINDIntended to study future Internet architecture impact of locationEvaluation of alternative methods, e.g. overlays vs. integrated layer 3, etc.
Geometric Stack: Geo Cooperative Forwarding
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
−76
−74
−72
−70
−68
−66
−64
−62
−60
−58
−56
Time (Sec)
RS
SI (
dBm
)
Highly-time varying channels (shadow fading)
Traditional routing selects next hop prior to transmission
too slow for fast link changesNeed more opportunistic protocols not based on static link assumption
Geobackoff for selecting next-hop from successfully received nodes: Receiver Diversity, and cooperative ARQ
26
Clean Slate Project: “CogNet”Architecture
“CogNet” protocols project under NSF FIND (Rutgers, KU, CMU, Blossom)Intended to develop a baseline protocol stack for cognitive radiosBased on a global control plane for bootstrapping CR subnetworks with adaptive MAC/PHY capabilitiesEnd-to-end integration of control across the Internet also being studied
Control PHYControl MAC
Boot-strap
Discovery
PHY
MACNetworkTransport
Application
Control Plane Data Plane
Global Control Plane
Data Plane
Control Signalling
Data Path
Establishment
Naming&
Addressing
Radios withProgrammable
PHY/MACWiNC2R platform
Adaptive Networks of Cognitive Radios
PHY1/MAC1
PHY2/MAC2
PHY3/MAC3
Protocol Stack with GCP
to the Internet
27
CogNet Project: ns2 Protocol Simulation
Evaluation by ns-2 simulationsBootstrap/Discovery: network setup time, overhead, theoretical end-to-end rateDPE: joint F/P/R allocation success ratio, overheadNaming/addressing: uniqueness of IP/Name
Ad hoc network – nodes randomly boot up
Control Interface
(802.11b)
Data Interface
(generic OFDM radio parameters)
28
CogNet Project: ns2 Simulation Results
Maximum and average network setup time (BSB interval 2sec, LSA interval 5sec, nodes randomly start [0, 4]sec)
Control overheadTheoretical max end-to-end rate averaged over the network
29
Key Technologies for Experimental Networks
30
Experimental Platforms: Programmable Wireless Devices
Single wireless GENI node architecture that covers different wireless network element needs:
Standard set of CPU platforms with different size/performanceMultiple radio cards as “plug-in” – easy to change radios, upgradeLinux OS with appropriate “open API” drivers for each radioExternal control module for remote management (reboot, etc.)GMC control module interfaced to uniform radio API
Plug-In radio modules(802.11, WiMax, 3G,..)
GENI M&C(“GMC”)
Linux OS
Control Module(out-of-band channel if available)
Processor Chassis with appropriate size/performance(sensor GW, mobile node, ad hoc router, AP, BTS…)
End-user Wireless Devices(commercial sensors, phones, PDA’s, laptops)
Linux OS w/GENI drivers and “thin GMC client”
Typical Experimental Wireless Network Platform
All components also to be available as “GENI wireless kits”
Radio API
31
Experimental Platforms: Open API Wireless AP’s and Routers
Radio Router
CPU board
AP/BTS/Forwarding Node/Sensor GW
or
Wired I/F
Radio Module
Radio Module
CPU board
Radio Card
Wired I/FSwitch fabric
Example implementation:ORBIT dual-radio node
Cognitive radio
3G/WiMax radio
802.11
Typical wireless network elements in GENIProgrammable Access Points (AP), Base Stations (BTS), Forwarding Nodes & Sensor Gateways, typically with 2 wired or wireless portsMore general n-port multi-radio router platform with hardware support for routing/switching
Radio Moduleoptions
Example implementation:4x4 switch + plug-in radio modules
Remote ManagementInterface (for reboot, etc.)
32
ORBIT Multi-Radio Node (v1.0)with integrated Chassis Manager
Experimental Platforms: Example of Open API Radio Node Implementation
COTS ORBIT Node (v2.0)With GPS & GPRS control
Libmac
Ioctl and /procInterfaces
Libnet and Libpcap
Device drivers (lower MAC, PHY)
Click802.11 Interface parameters
UserSpace
KernelSpace
Application Layer
ORBIT Node Software
33
Control plane(radio network controller)
Call Processing/ Signaling(GGSN)
Air Interface(basestation)
Functions
UMTS IP
UMTSNode B
UMTSNode B
O
GGSN
O
SGSN
O
RNC
ATMIP
ATMIP
Wireless Network
Lucent IP Base Station Router at WINLAB
Open API BTS needed for wide-area mobile services
UMTS BSR prototype with IP interface availableFuture work on WiMax BTSEquipment vendor collaboration required for implementation
Experimental Platforms: Example of Open API Base Station (UMTS, WiMax)
Courtesy: Lucent Bell Labs
34
Experimental Platforms: Vehicular NodesSystem leverages urban mesh infrastructure for Internet connectivityCampus/city-wide deployment (~100s mobile nodes)
private cars, taxi, campus shuttles, busesOn board equipment:
Radios: conventional (WiFi, 802.11p, Bluetooth, WiMax); next generation (MIMO, cognitive radios, etc); Sensors: GPS; video cameras; acoustic sensors; on board sensors..Data server, harvester: classify and store events; P2P applications
Example Vehicular Deployment (ORBIT outdoor)
Sample RSSImeasurements
Support urban-scale measurement and diverse applicationsLinux-based embedded PCs with 802.11a/b/g and 802.15.4
Mounted on top of light poles, buildings, etc.Possible sensor types include meteorological, environmental, pollution, etc.
Web-based interface for job scheduling, debugging, profilingOpen resource for the sensor network community
A
B
C
D
E
F
G
H
I
J
K
L
M
A
B
C
D
E
F
G
H
I
J
K
L
M
Photocell (Power)
CitySense Node goes here
Photocell (Power)
CitySense Node goes here
Metrix embedded PC
Harvard/BBN CitySensedeployment plan
Vaisala meterologicalsensor
Experimental Platforms: Sensor Net Deployments
Courtesy: Matt Welsh, Harvard U
36
Experimental Platforms: Programmable and Cognitive Radios
Various experimental programmable radio platforms under development for wireless network research…
WARP programmable radio, GNU radio, KU agile radio & near-future cognitive radios, ….Key issue: open software API’s and protocol stacks for full control of physical and link/MAC layers
.
802.11 AP
GNU USRP Software Radio
KU Agile RadioWINLAB/Lucent Cognitive Radios
Rice “WARP” board
37
Key Technologies: Network Virtualization
Support concurrent experiments to increase capacitySupport both short-term and long-term service experiments Virtualization in wireless networks complicated by PHY/MAC interactions between nodesSeveral techniques being investigated:
Virtual MAC (VMAC)Space divisionFrequency divisionTime division
38
Wireless Network Virtualization: VMAC
VirtualAccessPoint 1
Essid:1
Exp. 1 Exp. 2
Essid:2
Ch. y
AccessPoint
Essid:1
Exp. 1 Exp. 2
Ch. y
AccessPoint
Essid:2Ch. x
VirtualAccessPoint 2
Sliver 1 Sliver 2
39
Wireless Network Virtualization: VMAC Experimental Results
40
Wireless Network Virtualization: FDMATwo concurrent experiments can coexist using the same hardware via multiple radio cards/frequencies
Exp. 1
Exp. 2 Exp. 2
Exp. 1
Host OS
UML 1 UML 2
Wireless card 1 Wireless card 2
Virtual Device Virtual Device
41
UML Virtualization: FDMA slicesUDP experiments with 4 clients and 1AP
Measurements of throughput, delay and delay jitter as function of offered load
Throughput tracks quite well, but need to be careful about delay
42
Key Technologies: ORBIT Radio Grid as an Example of Large-Scale Programmable Net
ORBIT radio grid testbed currently supports networks with ~100’s of radio nodes (both end-points and routers)Integration with wired network testbeds available (PlanetLab, VINI)GNU radio for programmable MAC/PHY beyond open API features
Urban
300 meters
500 meters
Suburban
20 meters
ORBIT Radio Grid
Office
30 meters
Radio Mapping Concept for ORBIT Emulator
400-node Radio Grid Facility at WINLAB Tech Center
ProgrammableORBIT radio node
URSP2CR board
Planned upgrade(2007-08)
Current ORBIT sandbox with GNU radio
43
User’s Workstation
Store
Developer’s Workstation
Node
Grid Controller
Experimenter Code
NodeHandler
NodeAgent
Deployed Application
OML
Upload
Install
Spawn
Application Description
Code Generator
DBExperiment
Package
Control
Execution HarnessOperator Console
Experiment Description
Key Technologies: ORBIT Software Components
44
## Define nodes used in experiment#nodes([4,3], 'sender') {|node|
node.image = nil # Use default disk image
# experiment property spacenode.prototype("test:proto:sender", { # use prototype "sender"
'destinationHost' => '192.168.5.4', # Set it's property "destinationHost"'packetSize' => Experiment.property("packetSize"), 'rate' => Experiment.property("rate") # bind the remaining properties to defaults
}) # Can be overridden laternode.net.w0.mode = "master"node.net.w0.type = 'b'node.net.w0.essid = "helloworld” # Set wireless parametersnode.net.w0.ip = "%192.168.%x.%y"node.net.w0.rate = "11m"
}## Define nodes used in experiment#nodes([4,3], 'sender') {|node|
node.image = nil # Use default disk image
# experiment property spacenode.prototype("test:proto:sender", { # use prototype "sender"
'destinationHost' => '192.168.5.4', # Set it's property "destinationHost"'packetSize' => Experiment.property("packetSize"), 'rate' => Experiment.property("rate") # bind the remaining properties to defaults
}) # Can be overridden later node.net.w0.mode = "master"node.net.w0.type = 'b'node.net.w0.essid = "helloworld” # Set wireless parameters node.net.w0.ip = "%192.168.%x.%y"node.net.w0.rate = "11m"
}
ConsoleSupport services
Experiment Script
START
NodeAgents
NodeHandler
Key Technologies: ORBIT Execution Script
45
ConsoleMeasurementsSupport services
NodeAgent
Sliver running NodeAgent
Link bandwidth for sliver
Reserve testbed
NodeAgent
NodeAgent
NodeAgent
Planetlab – ORBITGateway NodeHandler UI
Nodeagent running on radio nodes in the ORBIT grid
Internet
Add Planetlabnode via
ORBIT slice
Key Technologies: Integrating PlanetLab with Wireless Testbeds – PL slice for ORBIT users
46
ConsoleMeasurementsSupport services
Internet
PlanetlabNode
Sliver
ORBIT grid = PlanetLab sliver
Link bandwidth for sliver
Add testbed to slice
Planetlab – ORBITProxy
NodeHandler UI
Nodeagent running on all nodes in the testbed
Key Technologies: Integrating PlanetLab with Wireless Testbeds – ORBIT Proxy for PL Users
47
Wired PlanetLab Wireless Orbit
Japan
Washington
Key Technologies: Planet-Lab ORBIT Proof-of-Concept Example 1
Example video streaming experiment from wired server node on PL to wireless receiver node on ORBIT
48
Key Technologies: Planet-Lab ORBIT Proof-of-Concept Example 2
Example with 3 virtual network slices on same network, with and without resource management (traffic shaping) for slice “QoS”
Exp. 1PlanetLab Wired ORBIT
wireless
Channel x
VAP
Exp. 2
Exp. 3
Exp. 1
Exp. 2
Exp. 3
0 30 60 90 120 1500
2
4
6
8
10
12
Clie
nt
Th
rou
gh
pu
t (M
bp
s)
Time (secs)
Throughput from Flow of experiment 1
Throughput from Flow of experiment 2
Offered load on each experiment
Experimentflows aremanuallyrate limited
49
Key Technologies: VINI-ORBIT Integration Example
Example mobile handoff experiment using VINI and ORBIT nodes
Video Handoff:PlanetLab NodesRunning VINI
CALIFORNIA TECH
BERKELEY
MIT
OpenVPNEthernetTunnel
AP1
VideoServer
Mobile Client172.16.0.1/30
AP2
OpenVPNEthernetTunnel
OpenVPNEthernetTunnel
Mobility Emulation
50
GENI Implementation Plan for Wireless
51
GENI Implementation: Open API Urban Ad-Hoc Mesh
Ad-hoc wireless network providing full coverage of high-density urban area ~ 10 Km**2
Enables experimentation with mesh network protocols & broadband mobile applicationsDual-radio forwarding node as building blockOpen API 802.11 with soft MAC, virtualization by frequency or spaceServices for running expts, data collection, frequency assignment and spectrum meas
Dual-radio ad-hoc router(includes wired interface for
AP sites)
RadioNodes
~50-100 mspacing
Ad-hocRadiolinks Access Point (wired)
Ad-Hoc Radio Node
Spectrum Monitor
Research Focus:1. Ad-hoc routing2. Self-organization & discovery3. Cross-layer optimizations4. MAC layer enhancements5. Security with ad-hoc routing6. Broadband QoS7. Impact of mobility8. Real-world application studies
52
GENI Implementation: Open API Wide Area Mobile Network
Open API wide-area wireless network to explore alternatives to cellular, hybrids with WLAN, Infostations, new mobile applications…
Suburban coverage ~50 Km**2 using ~10 wide-area BTS’s + ~100 short-range AP’sOpen API 3G or WiMax BTS and dual-radio 802.11 node as building blocks
Open API 3G/WiMax BTS
802.11 relay node or AP
WiMax or 3G Base Station Router
802.11 Relay NodePlatform
Connections toGENI Infrastructure
Research Focus:1. Internet transport for 3G/cellular2. Mobility support in future Internet3. Hybrid 3G/WLAN handover, etc.4. Multicasting5. Transport layer for wireless6. Security in future 3G/4G7. Information caching and multimedia
53
GENI Implementation: Urban Mesh + Sensor Network
2-3 sensor network projects to be selected via proposal process for integration into urban mesh deployment
Sensor network experiments will leverage 802.11 mesh or 3G wide area infrastructure in items 2,3Provide “user deployment kit” with platforms including sensor nodes and sensor/WLAN or sensor/3G gateway
Dual-radio ad-hoc router(includes wired interface for
AP sites)
RadioNodes
~50-100 mspacing
Ad-hocRadiolinks Access Point (wired)
Ad-Hoc Radio Node
Spectrum Monitor
Sensor Net Area
Sensor Nodes
Sensor Gateway
802.11 Access Pointor Relay Node
802.11 radio link
Short-range sensor radio link
54
GENI Implementation: Cognitive Radio Technology Demonstrator
Advanced technology demonstrator of cognitive radio networks for reliable wide-area services (over a ~50 Km**2 area) with spectrum sharing, adaptive networking, etc.
Basic building block is a cognitive radio platform, to be selected from competing research projects now in progress and/or future proposalsRequires enhanced software interfaces for control of radio PHY, discovery and bootstrapping, adaptive network protocols, etc. – suitable for protocol virtualizationNew experimental band for cognitive radio (below 1 Ghz preferable)
Cognitive Radio Network Node
Cognitive Radio Client
Cognitive Radio Network Node
Cognitive Radio Client
Connections to GENIInfrastructure
Spectrum MonitorsSpectrum Server
Research Focus:1. New technology validation of cognitive radio2. Protocols for adaptive PHY radio networks3. Efficient spectrum sharing methods4. Interference avoidance and spectrum etiquette5. Dynamic spectrum measurement6. Hardware platform performance studies
55
Concluding Remarks
56
Concluding Remarks: The Way Forward
Future Internet research ….….:
Identification of
57
Web Sites for More Information:
WINLAB: www.winlab.rutgers.eduORBIT: www.orbit-lab.orgGENI: www.geni.net