1

Emerging Wireless Technologies and the Future InternetStanford Clean Slate SeminarOct 30, 2007

Prof. D. Raychaudhuriray@winlab.rutgers.eduwww.winlab.rutgers.edu

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