A Rib June 2007
This document was uploaded by user and they confirmed that they have the permission to share it. If you are author or own the copyright of this book, please report to us by using this DMCA report form. Report DMCA
Overview
Download & View A Rib June 2007 as PDF for free.
More details
- Words: 4,256
- Pages: 57
Cognitive Radio Research in the U.S.: Overview, Challenges and Directions
ARIB Frequency Resource Development Symposium, June 8, 2007 Narayan B. Mandayam Professor of ECE & Associate Director, WINLAB Rutgers University [email protected] www.winlab.rutgers.edu 1
Cognitive Radio Research A Multidimensional Activity
Spectrum Policy
Theory and Algorithms Fundamental Limits Information & Coding Theory Cooperative Communications Game Theory & Microeconomics
Economics Regulation Legal Business
Hardware/Software Platforms & Prototyping Programmable agile radios GNU platforms
Several Universities
E.g. WINLAB-Rutgers, UC Berkeley, Virginia Tech, etc.
Govt Agencies
E.g. FCC, DARPA, NSF, etc.
Industry – small and big
E.g. Blossom, Shared Spectrum, Vanu, Intel, Alcatel-Lucent, Philips, Qualcomm and a very many more 2
How it all started The Spectrum Debate & Cognitive Radio
3
The Spectrum Debate
Triumph of Technology vs. Triumph of Economics
Open Access (Commons)
[Noam, Benkler, Shepard, Reed …]
Spectrum Property Rights
[Coase, Hazlett, Faulhaber+Farber]
Triumph of Technology Agile wideband radios will dynamically share a commons Success of 802.11 vs 3G
Triumph of Economics Owners can buy/sell/trade spectrum Flexible use, flexible technology, flexible divisibility, transferability A spectrum market will (by the force of economics) yield an efficient solution
What everyone agreed on (a few years ago):
Spectrum use is inefficient FCC licensing has yielded false scarcity
4
Spectrum Management:
In fairness to the FCC, Frequency allocation is complex…
Source: FCC website
5
Spectrum Management: Then again, poor utilization in most bands
Maximum Amplitudes
Heavy Use
Atlanta
Heavy Use
New Orleans
Time
Amplidue (dBm)
Less than 6% Occupancy Sparse Use
Medium Use
San Diego Frequency (MHz)
Frequency
FCC measurement shows that occupancy of approximately 700 MHz of spectrum below 1 GHz is less than 6~10% ~ 13% spectrum opportunities utilized in New York City during 2004 Political Convention to nominate U.S. Presidential Candidate 6
Arguments against Triumph of Technology
Partially developed theory Information theoretic relay channel Information theoretic interference channel Ad hoc network capacity, with/without mobility
Technology Panacea Spread spectrum, UWB, MIMO, OFDM Short range communications Ad hoc multi-hop mesh networks
Infant technology UWB, MIMO antenna arrays Transmitter agility
Technology not separable from user assumptions
Capabilities of technology vary with cooperation 7
Recent Research Developments Research themes that have emerged from mobile ad hoc and/or sensor networks research: Hierarchical Network Architecture wins
Cooperation wins
Achievable rates via information theoretic relay and broadcast channels
“Global” awareness and coordination wins
Capacity scaling, energy efficiency, increases lifetimes, facilitates discovery
Space, time and frequency awareness and coordination beyond local measurements
Efficient operation requires radios that can:
Cooperate Collaborate Discover Self-Organize into hierarchical networks 8
The Spectrum Debate and Cognitive Radio
What everyone agrees on now:
Possible middle ground?
Spectrum use is inefficient FCC licensing has yielded false scarcity
Dynamic spectrum access Short-term property rights Spectrum use driven by both technology and market forces
Cognitive Radios with ability to incorporate market forces?
Microeconomics based approaches to spectrum sharing Pricing and negotiation based strategies
(e.g. Ileri & Mandayam, “Dynamic Spectrum Access Models- Towards an Engineering Perspective in the Spectrum Debate” in IEEE Communications Magazine January 2008) 9
Dynamic Spectrum Management & Cognitive Radio
10
Motivation for Dynamic Spectrum and Cognitive Radio Techniques:
Static allocation of spectrum is inefficient
Spectrum allocation rules that encourage innovation & efficiency
Unlicensed systems need to scale and manage user “QoS”
Density of wireless devices will continue to increase
Free markets for spectrum, more unlicensed bands, new services, etc.
Anecdotal evidence of WLAN spectrum congestion
Slow, expensive process that cannot keep up with technology
~10x with home gadgets, ~100x with sensors/pervasive computing
Interoperability between proliferating radio standards
Programmable radios that can form cooperating networks across multiple PHY’s 11
Spectrum Management: Problem Scope Spectrum Allocation Rules (static)
Spectrum Coordination Server (dynamic)
INTERNET
Auction Server (dynamic)
Dynamic frequency provisioning
Spectrum Coordination protocols
BTS
AP
Short-range infrastructure mode network (e.g. WLAN)
Etiquette policy
Spectrum Coordination protocols Short-range ad-hoc net Wide-area infrastructure mode network (e.g. 802.16)
Dense deployment of wireless devices, both wide-area and shortrange Proliferation of multiple radio technologies, e.g. 802.11a,b,g, UWB, 802.16, 4G, etc. How should spectrum allocation rules evolve to achieve high efficiency? Available options include:
Ad-hoc sensor cluster (low-power, high density)
Agile radios (interference avoidance) Dynamic centralized allocation methods Distributed spectrum coordination (etiquette) Collaborative ad-hoc networks 12
Cognitive Radio: Design Space
Broad range of technology & related policy options for spectrum Need to determine performance (e.g. bps/Hz or bps/sq-m/Hz) of different technologies taking into account economic factors such as static efficiency, dynamic efficiency & innovation premium Unlicensed band + simple coord protocols
Protocol Complexity (degree of coordination)
Ad-hoc, Ad-hoc, Multi-hop Multi-hop Collaboration Collaboration
Internet Internet Server-based Server-based Spectrum Spectrum Etiquette Etiquette Unlicensed Unlicensed Band Band with DCA with DCA (e.g. 802.11x) (e.g. 802.11x) Internet Internet Spectrum Spectrum Leasing Leasing Static Static Assignment Assignment
“cognitive radio” schemes
Radio-level Radio-level Spectrum Spectrum Etiquette Etiquette Protocol Protocol
Reactive Reactive Rate/Power Rate/Power Control Control
Agile Agile Wideband “Open Access” Wideband + smart radios Radios Radios UWB, UWB, Spread Spread Spectrum Spectrum
Hardware Complexity 13
Selected Cognitive Radio Research
14
Cognitive Radio Research Fundamental research and algorithms – based on foundations of:
Information and Coding Theory
Signal Processing
Collaborative signal processing, Signal design for spectrum sharing, Interference avoidance, Distributed sensing algorithms
Game Theory
Relay cooperation, User Cooperation, Coding techniques for cooperation, Collaborative MIMO techniques
Microeconomics and pricing based schemes for spectrum sharing, negotiation and coexistence, Incentive mechanisms for cooperation
MAC and Networking Algorithms
Discovery protocols, Etiquette protocols, Self-organization protocols, Multihop routing
15
Information Theoretic Approaches
Various types of relay cooperation and user cooperation models
Cooperation – nodes share power and bandwidth to mutually enhance their transmissions Can achieve spatial diversity – similar to multiple antennas
Fundamental limits are known in limited cases Primary focus on achievable rates, outage and various cooperative coding schemes, e.g.
Decode-and-Forward Compress-and-Forward Amplify-and-Forward
Relay Cooperation
16
Information Theoretic Approaches
Various types of relay cooperation and user cooperation models
Cooperation – nodes share power and bandwidth to mutually enhance their transmissions Can achieve spatial diversity – similar to multiple antennas
Fundamental limits are known in limited cases Primary focus on achievable rates, outage and various cooperative coding schemes, e.g.
User Cooperation
Decode-and-Forward Compress-and-Forward Amplify-and-Forward SN
SN SN
SN
SN
SN
SN
AP
SN SN
Wired
AP SN
SN
SN
SN
SN
SN SN
SN
SN SN
Backbo ne
SN
SN
SN
SN SN
SN
AP
SN SN
SN
SN
SN
17
E.g. Relay Cooperation vs User Cooperation (Sankar-Kramer-Mandayam)
Sector of circle of radius 1 and destination at circle center. Uniform node distribution. Average rate and outage over 100 locations. Path-loss exp. γ = 4. Transmit SNR for all users = P1. Processing cost factor η
0.5 User 1 User 2 Destination Relay
0.4 0.3 0.2 y-coordinate
0.1 0 -0.1 -0.2 -0.3 -0.4 -0.5
0
0.1
0.2
0.3
0.4 0.5 0.6 x-coordinate
0.7
0.8
0.9
1
Compare Outage Probability? How much better than TDMA? What is the effect of processing costs?
18
Amplify-and-Forward Cooperation η = 0.01 Sub-plot 1
0
Sub-plot 2
0
10
10 Coop. 2-hop Relay Pr=0.5P1 Relay Pr=P1 TD-MAC
-1
-2
10
-3
10
-4
-2
10
-3
10
-4
10
10
K=2 Rate R = .25 η = .01
K=2 Rate R = .25 -5
10
-10
Relay Pr=P1 TD-MAC
-1
10 Outage Probability Pout
10 Outage Probability Pout
Coop. 2-hop Relay Pr=0.5P1
-5
-5
0 5 Transmit SNR P1 (dB)
10
10
-10
-5 0 5 10 Total (transmit+proc.) SNR Ptot (dB)
15
Relay and User Cooperation schemes gain over TDMA Relay achieves gains relative to user-cooperation 19
Amplify-and-Forward Cooperation η = 0.5, 1 Sub-plot 1
0
10 K=2 Rate R = .25 η = .5
-1
Outage Probability Pout
10
-2
10
-3
10
Coop. 2-hop Relay Pr=0.5P1
-4
10
-2
10
-3
10
-4
10
Relay Pr=P1 TD-MAC -5
10
K=2 Rate R = .25 η=1
-1
10 Outage Probability Pout
Sub-plot 2
0
10
-5
-5
0 5 10 Total (transmit+proc.) SNR Ptot (dB)
15
10
-5
Coop. 2-hop Relay Pr=0.5P1 Relay Pr=P1 TD-MAC 0 5 10 Total (transmit+proc.) SNR Ptot (dB)
15
Even with η ↑ relay achieves gains relative user cooperation 20
Game Theory Approaches
Negotiation strategies for mediation Pricing and microeconomic strategies to promote cooperation in spectrum sharing Reimbursing costs in cooperation – energy costs, delay costs
Domination strategies for situations of conflict
Spectrum warfare with agile waveforms and competition for spectrum
Coalition formation strategies for cooperation
Coalitional games for receiver and transmitter coalitions in spectrum sharing
Approaches result in algorithms that specify: • Power control • • • •
Rate control Channel selection Cooperation techniques Route selection
21
E.g. Inducing Forwarding through Reimbursement (Ileri-Mandayam)
Each node •Enjoys its own data reaching the access point. •Pays for its outgoing throughput to adjacent devices. •Gets reimbursed for data it relays only if the data reaches the access point. 22
E.g. Inducing Forwarding through Reimbursement 10
19
Aggregate bits/Joule Aggregate bits/Joule, no reimbursement.
Horizontal Trajectory 10
18
ACCESS POINT
10
Radio tower
POSITIONS OF USER 1 IN HORIZONTAL GEOMETRY
5m
5m
10
-
USER 2 (POTENTIAL FORWARDER)
16
10 m
10 -5 m
17
USER 1(NONFORWARDER)
10
15
14
0
1
2
3
4
5
6
7
user 2 distance
• Incentives for forwarding works well when nodes tend to “cluster” • The aggregate bits/Joule in network is higher • Network revenue is also higher 23
PHY & MAC: Reactive Algorithms
Reactive (autonomous) methods used to avoid interference via:
Frequency agility: dynamic channel allocation by scanning Power control: power control by interference detection and scanning Time scheduling: MAC packet re-scheduling based on observed activity Waveform agility: dynamism in signal space
C B
Frequency agility
DD
A&B’s spectrum band
D
C
D
D
Range with Power Control C&D’s spectrum band
A A
Scheduling
Range with Power Control Range without Power Control
Range without Power Control
24
Cognitive Radio: Limitations of Reactive Schemes
Reactive schemes (without explicit coordination protocols) suffer from certain limitations:
Near-far problems possible at the receiver Inability to predict future behavior of other nodes Only detects transmitters, not receivers, but interference is a receiver property
C B
D’s agile radio waveform without coordination protocol
D Coverage
A’s agile radio waveform
A cannot hear D
area of D
A Y
with coordination
Coverage area of A Hidden Terminal Problem
25
Cognitive Approaches: Outlook Fundamental research and algorithms – based on foundations of
Information and coding theory, Game theory, MAC, Networking, Signal processing
all point to the following: Cognitive radio networks require a large of amount of network (and channel) state information to enable efficient Discovery Self-organization Cooperation Techniques
In addition to advances in cognitive radio technology, Network Architectures and Information Aids that support these are required
26
Cognitive Radios need help too!
Infrastructure that can facilitate cognitive radio networks Coordination mechanisms for coexistence and cooperation
Information aids
Spectrum Coordination Mechanisms
Network architectures
“Spectrum Servers” to advise/mediate sharing
27
Cognitive Radio: Common Spectrum Coordination Channel (CSCC) (Jing-Raychaudhuri)
Common spectrum coordination channel (CSCC) can be used to coordinate radios with different PHY Requires a standardized out-of-band etiquette channel & protocol Periodic tx of radio parameters on CSCC, higher power to reach hidden nodes Local contentions resolved via etiquette policies (independent of protocol) Also supports ad-hoc multi-hop routing associations
CH#N CH#N-1 CH#N-2 : : CH#2
CSCC RX range for X
Ad-hoc net B
Ad-hoc net A
X
Master Node
CH#1 CSCC
Frequency
Ad-hoc Piconet
Y
CSCC RX range for Y 28
CSCC Spectrum Etiquette Protocol
CSCC( Common Spectrum Coordination Channel) can enable
mutual observation between neighboring radio devices by periodically broadcasting spectrum usage information
Service channels
Edge-of-band coordination channel
29
CSCC: Proof-of-Concept Experiments
WLAN-BT Scenario
Different devices with dual mode radios running CSCC d=4 meters are kept constant Priority-based etiquette policy
30
CSCC Results: Throughput Traces
Observations:
WLAN session throughput can improve ~35% by CSCC coordination BT session throughput can improve ~25% by CSCC coordination WLAN = high priority
Bluetooth = high priority 65
4.8
CSCC on CSCC off
60
CSCC on CSCC off
4.4 4.2 4.0 3.8 3.6
Bluetooth Throughput (Kbps)
WLAN Throughput (Mbps)
4.6
55
50
45
40
35 3.4 0
20
40
60
80
100 120 140 160 180 200 220 240
Time (Seconds)
WLAN session with BT2 in initial position
30 0
50
100
150
200
250
300
Time (Seconds)
BT session with BT2 in initial position 31
802.11 & 16 Co-Existence Scenario
802.16a Traffic Type MAC protocol Channel Model
802.11b
UBR (Poisson arrival), UDP packet, 512 Bytes datagram TDMA
IEEE 802.11 BSS mode
AWGN, two ray ground propagation model, no fading
Bandwidth/channels
20 MHz / 4 non-overlapping chs
22MHz / 11 overlapping chs
Bit Rate
13Mbps
2Mbps
Radio parameters
OFDM (256-FFT, QPSK)
DSSS (QPSK)
Background Noise
-174 dBm/Hz
Rx Noise Figure
9 dB
9 dB
Receiver Sensitivity
-80dBm (@BER 10^-6)
-82dBm (@BER 10^-5)
Antenna Height
BS 15m, SS 1.5m
1.5m
Tx Power/Max range
33dBm / 3.2Km
20dBm / 500m
Default channel
Channel 1 : centered at 2412GHz
Channel 1 : centered at 2412GHz
Available channels
4 (non-overlap)
12 (overlapping)
32
802.11 & 16 Co-Existence: Reactive vs. CSCC-based Power Control (Jing-Raychaudhuri) 802.11b Hotspot DSS-AP
100m AP
1km SS
BS
Single 802.11 Hot Spot Case
802.16a Cell 0.7 1.2
0.6 0.5 0.4 0.3
No Coordination CSCC frequency adaptation
0.2 0.1 0.0 802.16 DL
802.11 link
Average Link Throughput (Mbps)
Average Link Throughput (Mbps)
0.8
1.0
0.8
802.16a DL 802.16a DL with CSCC 802.11 link 802.11 link with CSCC Average No Coordination Average with CSCC
0.6
0.4
0.2
802.16 DL and 802.11 link
CSCC frequency adaptation when DSS-AP =200m and traffic load 2Mbps
0.0 0
200
400
600
800
1000
Distance between 802.16 SS and 802.11 hotspot (meters)
Throughputs vs. DSS-AP by using CSCC power adaptation and traffic load 2Mbps
33
802.11 & 16 Co-Existence: Reactive vs. CSCC Power Control 802.16a BS
Multiple 802.11 Hot Spot Case
No Coordination Varying Max Cluster with CSCC Radius with RTPC with TA R16
120000
110000
1km
120000
Average Network Throughput (bps)
802.11b AP 802.11b Client
130000
No Coordination with CSCC with RTPC with TA
110000
100000
90000
80000
70000
100000
0.2
0.4
0.6
0.8
1.0
Clustering Index 90000
Max Hotspot Radius R11
80000
Region (i)
70000 0.2
0.4
0.6
Clustering Index
1 Mbps load
130000
0.8
1.0
Region (ii)
802.16 SS nodes are (i) uniform or (ii) clustered in the cell. clustering index CI = R11/R16
Average Network Throughput (bps)
Average Network Throughput (bps)
600 Kbps load
802.16a SS
No Coordination with CSCC with RTPC with TA
120000
110000
100000
90000
80000
70000 0.2
0.4
0.6
0.8
1.0
Clustering Index
CSCC power adaptation with clustered 802.16a SS in region (ii), with numbers of 802.16a SS : 802.11b nodes = 1:1
34
Cognitive Radio: Spectrum Policy Server
Internet-based Spectrum Policy Server can help to coordinate wireless networks (a “Google for spectrum”)
Needs connection to Internet even under congested conditions (...low bit-rate OK) Some level of position determination needed (..coarse location OK?) Spectrum coordination achieved via etiquette protocol centralized at server Spectrum Policy Server www.spectrum.net
Internet Internet AP1
Access Point (AP2)
WLAN operator A
Etiquette Protocol
AP1: type, loc, freq, pwr AP2: type, loc, freq, pwr BT MN: type, loc, freq, pwr
WLAN operator B Master Node
Wide-area Cellular data service
Ad-hoc Bluetooth Piconet
35
What can a Spectrum Policy Server do?
rate1
rate2
rate4
Spectrum Server
rate3
Spectrum Policy Server facilitates co-existence of heterogeneous set of radios by advising them on several possible issues Spectrum etiquette Interference information Location specific services Many more things ….
36
Scheduling Variable Rate Links with a Spectrum Server (Raman-Yates-Mandayam)
Users share a common frequency band
Wireless network of L directed links Links employ ON-OFF transmission schedule in each time slot
Use constant transmission power in the ON state
Links employ interference-adaptive modulation/coding
Orthogonal signal dimensions = time slots Time domain scheduling is used for channelization
Link rate in each time slot depends on interference, receiver mitigation.
Interference depends on the transmission mode
mode = subset of links that are ON simultaneously
1
1
4
tli = 1, if link l is in mode i
2 3
[0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1] [0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1] [0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1] [0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1]
= 0, otherwise. 3
Transmission mode [1 0 1 0] (one of 24 possible modes) 37
Spectrum Server Mode Matrix Æ Rate Matrix 1
4
1 Glk = link gain from Tx k to Rx l
2 3
3
network with 4 links
Transmission mode [1 0 1 0]
Rate matrix C = [6.6 [0 [0 [0
0 6.6 0 0
0.01 0.06 0 0
0 0 6.6 0
0.56 0 1.0 0
0 1.86 1.86 0
0.01 0 0.06 0 0.83 0 0 6.65
2.05 0 0 0.32
0 0.97 0 0.05
0.01 0 0.06 0 0 0.04 0.04 0.40
0.49 0 0.01] 0 0.77 0.06 ] 0.04 0.04 0.04 ] 0.19 0.05 0.04]
38
Spectrum Server: Optimizing the Mode Schedule
Spectrum server specifies xi = fraction of time mode i is ON
rl = Σi cli
Average rate in link l is
In vector form,
Spectrum server specifies x to optimize an objective such as for example:
xi
r = Cx
Maximum sum rate of the schedule Maximize the common rate on the links Fair scheduling Efficient scheduling
39
Spectrum Server: Scheduling Results
As common rmin increases,
the sum rate decreases Rates of dominant mode links decrease Rates of disadvantaged links increase
40
Spectrum Server: Fairness in Scheduling
41
Hardware & Software Platforms
42
Vanu Software Defined Radio
Vanu Inc. SDR programmable radio based on commodity processors. Supports multiple standards on handheld device.
Vanu Inc. Software Defined Radio Source: http://www.vanu.com/products.html 43
GNU Hardware/Software Platform
(Blossom)
GNU/USRP boards with GNU Software with API’s for flexible PHY and MAC are currently available for experimentation Various RF front-ends (0-100MHz, 400MHz, 900MHz, 2.4GHZ) with data rates upto 64 MSamples/sec All DSP functions in software on general-purpose CPU
44
WARP Platform
(Rice University)
•Various RF front-ends (2.4GHZ, 5GHz) with baseband of up to 20 MHz •High data rates possible, Programmable PHY & Protocols, Extensible to MIMO •Backed by Xilinx, Nokia
45
Cognitive Radio Networks & Protocols
Discovery strategies
Algorithms and protocols for frequency selection, coordination and cooperation
Multihop strategies
Algorithms for self-organization and routing
E.g. • Cognitive Radio scans for active nodes and executes discovery algorithm • Control protocol for exchange of network information and routing tables End-to-end routed path From A to F Bootstrapped PHY & control link
C
PHY A
B B
DD
PHY C
E
PHY B
A A Control (e.g. CSCC)
F
Multi-mode radio PHY Ad-Hoc Discovery & Routing Capability
Functionality can be quite challenging!
46
WINLAB Cognitive Radio WINLAB’s “network centric” concept for cognitive radio prototype
(..under development in collaboration with GA Tech & Lucent Bell Labs)
Requirements include:
~Ghz spectrum scanning, Etiquette policy processing PHY layer adaptation (per pkt) Ad-hoc network discovery Multi-hop routing ~100 Mbps+
47
Cognitive Radio Networks: “CogNet” Architecture
New NSF FIND project called “CogNet” aimed at development of prototype cognitive radio stack within GNU framework Joint effort between Rutgers, U Kansas, CMU and Blossom Inc
48
Cognitive Radio Networks: “CogNet” Protocol Stack
Global Control Plane (GCP)
Common framework for spectrum allocation, PHY/MAC bootstrap, topology discovery and cross-layer routing
Data plane
Dynamically linked PHY, MAC, Network modules and parameters as specified by control plane protocol
Data Plane Global Control Plane Control Plane
Data Plane
Control Signalling
Application Bootstrap
Disco very
Data Path Establish ment
Naming & Addres sing
Transport Network
Control MAC
MAC
Control PHY
PHY
49
Cognitive Radio Network Experiments Hardware/Software Platforms@WINLAB
ORBIT radio grid testbed currently supports ~10/USRP GNU radios, 100 low-cost spectrum sensors, WARP platforms, WINLAB Cognitive platforms and GNU/USRP2 Each platform will include baseline CogNet stack
Suburban
ORBIT Radio Grid
Current ORBIT sandbox with GNU radio 20 meters 500 meters Office
30 meters
Urban
300 meters
Radio Mapping Concept for ORBIT Emulator
400-node Radio Grid Facility at WINLAB Tech Center Programmable ORBIT radio node
Planned upgrade (2007-08)
URSP2 CR board
50
Concluding Remarks
51
Wireless/Mobile/Sensor Scenarios and the Future Internet – NSF’s GENI Project
Some architectural and protocol implications for the future Internet...
Integrated support for dynamic end-user mobility
Wireless/mobile devices as routers (mesh networks, etc.)
Network topology changes more rapidly than in today’s wired Internet
Significant increase in network scale (10B sensors in 2020!)
New ad hoc network service concepts: sensors, P2P, P2M, M2M,…
Addressing architecture issues – name vs. routable address
Integrating geographic location into routing/addressing
Integrating cross-layer and cognitive radio protocol stacks
Data/content driven networking for sensors and mobile data
Pervasive network functionality vs. broadband streaming
Power efficiency considerations and computing constraints for sensors
Many new security considerations for wireless/mobile
Economic incentives, e.g. for forwarding and network formation 52
NSF GENI Implementation: Wireless Subnets – Overall Wireless Deployment Plan
Five types of experimental wireless networks planned – necessary to support full range of protocol research and to enable new applications
1. Wireless emulation and simulation (repeatable protocol validations) 2. Urban 802.11-based mesh/ad-hoc network (real-world networking experience with emerging short-range radios) 3. Wide-area suburban network with both 3G/WiMax (wide area) and 802.11 radios 4. Sensor networks (…application specific, specific system TBD via proposal process; may include environmental, vehicular, smart spaces, etc.) 5. Cognitive radio network – advanced technology demonstrator (…adaptive, spectrum efficient networks using emerging CR platforms) …also some common network facilities such as location & dynamic binding services
Each network at a different geographic location – new spectrum allocation may be needed at some sites 53
NSF GENI Implementation Wireless Sub-Networks Overview Location Service
Emerging 5 Technologies (cognitive radio)
Advanced Technology Demonstrator (spectrum)
Broadband Services, Mobile Computing
Other GENI services Infrastructure
Ad-Hoc Mesh 2 Network
1 NSF Radio Testbeds
Open API 3 Wide-Area Networks
“Open” Internet Concepts for Cellular devices
Sensor 4 Networks Embedded wireless, Real-world applications Protocol & Scaling Studies
Emulation & Simulation 54
Cognitive Radio in NSF’s GENI Project
Propose to build advanced technology demonstrator of cognitive radio networks for reliable wide-area services (over a ~50 Km**2 coverage 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 proposals Requires enhanced software interfaces for control of radio PHY, discovery and bootstrapping, adaptive network protocols …….. suitable for protocol virtualization FCC experimental license for new cognitive radio band
Cognitive Radio Network Node
Cognitive Radio Client
Spectrum Server
Cognitive Radio Network Node
Cognitive Radio Client
Spectrum Monitors
Connections to GENI Infrastructure Research Focus: 1. New technology validation of cognitive radio 2. Protocols for adaptive PHY radio networks 3. Efficient spectrum sharing methods 4. Interference avoidance and spectrum etiquette 5. Dynamic spectrum measurement 55 6. Hardware platform performance studies
Concluding Remarks
Future wireless networks need ~100-1000x increases in density and bit-rate of radios Æ motivates better spectrum coordination methods Spot shortages of spectrum will occur if present static allocation is continued Æ significant improvement achieved with dynamic allocation Cognitive radio technologies can be characterized in terms of the combination of hardware complexity & level of protocol coordination Promising cognitive radio schemes include
Agile radio with interference avoidance
Spectrum etiquette protocols: spectrum server, CSCC..
Adaptive networks via ad-hoc collaboration
Early technical results now available for some of these methods, but very different complexity factors and market implications… 56
Concluding Remarks
Future research areas in cognitive radio include:
New concepts and algorithms for agile radio and spectrum etiquette protocols
Architecture and design of adaptive wireless networks based on cognitive radios
Detailed evaluation of large-scale cognitive radio systems using alternative methods
Spectrum measurement and field validation of proposed methods
Cognitive radio hardware and software platforms
User-level field trials of emerging cognitive radios and related algorithms/protocols may also be useful to gain experience
Controlled testbed experiments comparing different co-existence methods
Large-scale “spectrum server” trial for 802.11x coordination
Experimental deployments in proposed US FCC cognitive radio band
Success with cognitive radio technologies should lead to major improvements in spectrum efficiency, performance and interoperability 57
ARIB Frequency Resource Development Symposium, June 8, 2007 Narayan B. Mandayam Professor of ECE & Associate Director, WINLAB Rutgers University [email protected] www.winlab.rutgers.edu 1
Cognitive Radio Research A Multidimensional Activity
Spectrum Policy
Theory and Algorithms Fundamental Limits Information & Coding Theory Cooperative Communications Game Theory & Microeconomics
Economics Regulation Legal Business
Hardware/Software Platforms & Prototyping Programmable agile radios GNU platforms
Several Universities
E.g. WINLAB-Rutgers, UC Berkeley, Virginia Tech, etc.
Govt Agencies
E.g. FCC, DARPA, NSF, etc.
Industry – small and big
E.g. Blossom, Shared Spectrum, Vanu, Intel, Alcatel-Lucent, Philips, Qualcomm and a very many more 2
How it all started The Spectrum Debate & Cognitive Radio
3
The Spectrum Debate
Triumph of Technology vs. Triumph of Economics
Open Access (Commons)
[Noam, Benkler, Shepard, Reed …]
Spectrum Property Rights
[Coase, Hazlett, Faulhaber+Farber]
Triumph of Technology Agile wideband radios will dynamically share a commons Success of 802.11 vs 3G
Triumph of Economics Owners can buy/sell/trade spectrum Flexible use, flexible technology, flexible divisibility, transferability A spectrum market will (by the force of economics) yield an efficient solution
What everyone agreed on (a few years ago):
Spectrum use is inefficient FCC licensing has yielded false scarcity
4
Spectrum Management:
In fairness to the FCC, Frequency allocation is complex…
Source: FCC website
5
Spectrum Management: Then again, poor utilization in most bands
Maximum Amplitudes
Heavy Use
Atlanta
Heavy Use
New Orleans
Time
Amplidue (dBm)
Less than 6% Occupancy Sparse Use
Medium Use
San Diego Frequency (MHz)
Frequency
FCC measurement shows that occupancy of approximately 700 MHz of spectrum below 1 GHz is less than 6~10% ~ 13% spectrum opportunities utilized in New York City during 2004 Political Convention to nominate U.S. Presidential Candidate 6
Arguments against Triumph of Technology
Partially developed theory Information theoretic relay channel Information theoretic interference channel Ad hoc network capacity, with/without mobility
Technology Panacea Spread spectrum, UWB, MIMO, OFDM Short range communications Ad hoc multi-hop mesh networks
Infant technology UWB, MIMO antenna arrays Transmitter agility
Technology not separable from user assumptions
Capabilities of technology vary with cooperation 7
Recent Research Developments Research themes that have emerged from mobile ad hoc and/or sensor networks research: Hierarchical Network Architecture wins
Cooperation wins
Achievable rates via information theoretic relay and broadcast channels
“Global” awareness and coordination wins
Capacity scaling, energy efficiency, increases lifetimes, facilitates discovery
Space, time and frequency awareness and coordination beyond local measurements
Efficient operation requires radios that can:
Cooperate Collaborate Discover Self-Organize into hierarchical networks 8
The Spectrum Debate and Cognitive Radio
What everyone agrees on now:
Possible middle ground?
Spectrum use is inefficient FCC licensing has yielded false scarcity
Dynamic spectrum access Short-term property rights Spectrum use driven by both technology and market forces
Cognitive Radios with ability to incorporate market forces?
Microeconomics based approaches to spectrum sharing Pricing and negotiation based strategies
(e.g. Ileri & Mandayam, “Dynamic Spectrum Access Models- Towards an Engineering Perspective in the Spectrum Debate” in IEEE Communications Magazine January 2008) 9
Dynamic Spectrum Management & Cognitive Radio
10
Motivation for Dynamic Spectrum and Cognitive Radio Techniques:
Static allocation of spectrum is inefficient
Spectrum allocation rules that encourage innovation & efficiency
Unlicensed systems need to scale and manage user “QoS”
Density of wireless devices will continue to increase
Free markets for spectrum, more unlicensed bands, new services, etc.
Anecdotal evidence of WLAN spectrum congestion
Slow, expensive process that cannot keep up with technology
~10x with home gadgets, ~100x with sensors/pervasive computing
Interoperability between proliferating radio standards
Programmable radios that can form cooperating networks across multiple PHY’s 11
Spectrum Management: Problem Scope Spectrum Allocation Rules (static)
Spectrum Coordination Server (dynamic)
INTERNET
Auction Server (dynamic)
Dynamic frequency provisioning
Spectrum Coordination protocols
BTS
AP
Short-range infrastructure mode network (e.g. WLAN)
Etiquette policy
Spectrum Coordination protocols Short-range ad-hoc net Wide-area infrastructure mode network (e.g. 802.16)
Dense deployment of wireless devices, both wide-area and shortrange Proliferation of multiple radio technologies, e.g. 802.11a,b,g, UWB, 802.16, 4G, etc. How should spectrum allocation rules evolve to achieve high efficiency? Available options include:
Ad-hoc sensor cluster (low-power, high density)
Agile radios (interference avoidance) Dynamic centralized allocation methods Distributed spectrum coordination (etiquette) Collaborative ad-hoc networks 12
Cognitive Radio: Design Space
Broad range of technology & related policy options for spectrum Need to determine performance (e.g. bps/Hz or bps/sq-m/Hz) of different technologies taking into account economic factors such as static efficiency, dynamic efficiency & innovation premium Unlicensed band + simple coord protocols
Protocol Complexity (degree of coordination)
Ad-hoc, Ad-hoc, Multi-hop Multi-hop Collaboration Collaboration
Internet Internet Server-based Server-based Spectrum Spectrum Etiquette Etiquette Unlicensed Unlicensed Band Band with DCA with DCA (e.g. 802.11x) (e.g. 802.11x) Internet Internet Spectrum Spectrum Leasing Leasing Static Static Assignment Assignment
“cognitive radio” schemes
Radio-level Radio-level Spectrum Spectrum Etiquette Etiquette Protocol Protocol
Reactive Reactive Rate/Power Rate/Power Control Control
Agile Agile Wideband “Open Access” Wideband + smart radios Radios Radios UWB, UWB, Spread Spread Spectrum Spectrum
Hardware Complexity 13
Selected Cognitive Radio Research
14
Cognitive Radio Research Fundamental research and algorithms – based on foundations of:
Information and Coding Theory
Signal Processing
Collaborative signal processing, Signal design for spectrum sharing, Interference avoidance, Distributed sensing algorithms
Game Theory
Relay cooperation, User Cooperation, Coding techniques for cooperation, Collaborative MIMO techniques
Microeconomics and pricing based schemes for spectrum sharing, negotiation and coexistence, Incentive mechanisms for cooperation
MAC and Networking Algorithms
Discovery protocols, Etiquette protocols, Self-organization protocols, Multihop routing
15
Information Theoretic Approaches
Various types of relay cooperation and user cooperation models
Cooperation – nodes share power and bandwidth to mutually enhance their transmissions Can achieve spatial diversity – similar to multiple antennas
Fundamental limits are known in limited cases Primary focus on achievable rates, outage and various cooperative coding schemes, e.g.
Decode-and-Forward Compress-and-Forward Amplify-and-Forward
Relay Cooperation
16
Information Theoretic Approaches
Various types of relay cooperation and user cooperation models
Cooperation – nodes share power and bandwidth to mutually enhance their transmissions Can achieve spatial diversity – similar to multiple antennas
Fundamental limits are known in limited cases Primary focus on achievable rates, outage and various cooperative coding schemes, e.g.
User Cooperation
Decode-and-Forward Compress-and-Forward Amplify-and-Forward SN
SN SN
SN
SN
SN
SN
AP
SN SN
Wired
AP SN
SN
SN
SN
SN
SN SN
SN
SN SN
Backbo ne
SN
SN
SN
SN SN
SN
AP
SN SN
SN
SN
SN
17
E.g. Relay Cooperation vs User Cooperation (Sankar-Kramer-Mandayam)
Sector of circle of radius 1 and destination at circle center. Uniform node distribution. Average rate and outage over 100 locations. Path-loss exp. γ = 4. Transmit SNR for all users = P1. Processing cost factor η
0.5 User 1 User 2 Destination Relay
0.4 0.3 0.2 y-coordinate
0.1 0 -0.1 -0.2 -0.3 -0.4 -0.5
0
0.1
0.2
0.3
0.4 0.5 0.6 x-coordinate
0.7
0.8
0.9
1
Compare Outage Probability? How much better than TDMA? What is the effect of processing costs?
18
Amplify-and-Forward Cooperation η = 0.01 Sub-plot 1
0
Sub-plot 2
0
10
10 Coop. 2-hop Relay Pr=0.5P1 Relay Pr=P1 TD-MAC
-1
-2
10
-3
10
-4
-2
10
-3
10
-4
10
10
K=2 Rate R = .25 η = .01
K=2 Rate R = .25 -5
10
-10
Relay Pr=P1 TD-MAC
-1
10 Outage Probability Pout
10 Outage Probability Pout
Coop. 2-hop Relay Pr=0.5P1
-5
-5
0 5 Transmit SNR P1 (dB)
10
10
-10
-5 0 5 10 Total (transmit+proc.) SNR Ptot (dB)
15
Relay and User Cooperation schemes gain over TDMA Relay achieves gains relative to user-cooperation 19
Amplify-and-Forward Cooperation η = 0.5, 1 Sub-plot 1
0
10 K=2 Rate R = .25 η = .5
-1
Outage Probability Pout
10
-2
10
-3
10
Coop. 2-hop Relay Pr=0.5P1
-4
10
-2
10
-3
10
-4
10
Relay Pr=P1 TD-MAC -5
10
K=2 Rate R = .25 η=1
-1
10 Outage Probability Pout
Sub-plot 2
0
10
-5
-5
0 5 10 Total (transmit+proc.) SNR Ptot (dB)
15
10
-5
Coop. 2-hop Relay Pr=0.5P1 Relay Pr=P1 TD-MAC 0 5 10 Total (transmit+proc.) SNR Ptot (dB)
15
Even with η ↑ relay achieves gains relative user cooperation 20
Game Theory Approaches
Negotiation strategies for mediation Pricing and microeconomic strategies to promote cooperation in spectrum sharing Reimbursing costs in cooperation – energy costs, delay costs
Domination strategies for situations of conflict
Spectrum warfare with agile waveforms and competition for spectrum
Coalition formation strategies for cooperation
Coalitional games for receiver and transmitter coalitions in spectrum sharing
Approaches result in algorithms that specify: • Power control • • • •
Rate control Channel selection Cooperation techniques Route selection
21
E.g. Inducing Forwarding through Reimbursement (Ileri-Mandayam)
Each node •Enjoys its own data reaching the access point. •Pays for its outgoing throughput to adjacent devices. •Gets reimbursed for data it relays only if the data reaches the access point. 22
E.g. Inducing Forwarding through Reimbursement 10
19
Aggregate bits/Joule Aggregate bits/Joule, no reimbursement.
Horizontal Trajectory 10
18
ACCESS POINT
10
Radio tower
POSITIONS OF USER 1 IN HORIZONTAL GEOMETRY
5m
5m
10
-
USER 2 (POTENTIAL FORWARDER)
16
10 m
10 -5 m
17
USER 1(NONFORWARDER)
10
15
14
0
1
2
3
4
5
6
7
user 2 distance
• Incentives for forwarding works well when nodes tend to “cluster” • The aggregate bits/Joule in network is higher • Network revenue is also higher 23
PHY & MAC: Reactive Algorithms
Reactive (autonomous) methods used to avoid interference via:
Frequency agility: dynamic channel allocation by scanning Power control: power control by interference detection and scanning Time scheduling: MAC packet re-scheduling based on observed activity Waveform agility: dynamism in signal space
C B
Frequency agility
DD
A&B’s spectrum band
D
C
D
D
Range with Power Control C&D’s spectrum band
A A
Scheduling
Range with Power Control Range without Power Control
Range without Power Control
24
Cognitive Radio: Limitations of Reactive Schemes
Reactive schemes (without explicit coordination protocols) suffer from certain limitations:
Near-far problems possible at the receiver Inability to predict future behavior of other nodes Only detects transmitters, not receivers, but interference is a receiver property
C B
D’s agile radio waveform without coordination protocol
D Coverage
A’s agile radio waveform
A cannot hear D
area of D
A Y
with coordination
Coverage area of A Hidden Terminal Problem
25
Cognitive Approaches: Outlook Fundamental research and algorithms – based on foundations of
Information and coding theory, Game theory, MAC, Networking, Signal processing
all point to the following: Cognitive radio networks require a large of amount of network (and channel) state information to enable efficient Discovery Self-organization Cooperation Techniques
In addition to advances in cognitive radio technology, Network Architectures and Information Aids that support these are required
26
Cognitive Radios need help too!
Infrastructure that can facilitate cognitive radio networks Coordination mechanisms for coexistence and cooperation
Information aids
Spectrum Coordination Mechanisms
Network architectures
“Spectrum Servers” to advise/mediate sharing
27
Cognitive Radio: Common Spectrum Coordination Channel (CSCC) (Jing-Raychaudhuri)
Common spectrum coordination channel (CSCC) can be used to coordinate radios with different PHY Requires a standardized out-of-band etiquette channel & protocol Periodic tx of radio parameters on CSCC, higher power to reach hidden nodes Local contentions resolved via etiquette policies (independent of protocol) Also supports ad-hoc multi-hop routing associations
CH#N CH#N-1 CH#N-2 : : CH#2
CSCC RX range for X
Ad-hoc net B
Ad-hoc net A
X
Master Node
CH#1 CSCC
Frequency
Ad-hoc Piconet
Y
CSCC RX range for Y 28
CSCC Spectrum Etiquette Protocol
CSCC( Common Spectrum Coordination Channel) can enable
mutual observation between neighboring radio devices by periodically broadcasting spectrum usage information
Service channels
Edge-of-band coordination channel
29
CSCC: Proof-of-Concept Experiments
WLAN-BT Scenario
Different devices with dual mode radios running CSCC d=4 meters are kept constant Priority-based etiquette policy
30
CSCC Results: Throughput Traces
Observations:
WLAN session throughput can improve ~35% by CSCC coordination BT session throughput can improve ~25% by CSCC coordination WLAN = high priority
Bluetooth = high priority 65
4.8
CSCC on CSCC off
60
CSCC on CSCC off
4.4 4.2 4.0 3.8 3.6
Bluetooth Throughput (Kbps)
WLAN Throughput (Mbps)
4.6
55
50
45
40
35 3.4 0
20
40
60
80
100 120 140 160 180 200 220 240
Time (Seconds)
WLAN session with BT2 in initial position
30 0
50
100
150
200
250
300
Time (Seconds)
BT session with BT2 in initial position 31
802.11 & 16 Co-Existence Scenario
802.16a Traffic Type MAC protocol Channel Model
802.11b
UBR (Poisson arrival), UDP packet, 512 Bytes datagram TDMA
IEEE 802.11 BSS mode
AWGN, two ray ground propagation model, no fading
Bandwidth/channels
20 MHz / 4 non-overlapping chs
22MHz / 11 overlapping chs
Bit Rate
13Mbps
2Mbps
Radio parameters
OFDM (256-FFT, QPSK)
DSSS (QPSK)
Background Noise
-174 dBm/Hz
Rx Noise Figure
9 dB
9 dB
Receiver Sensitivity
-80dBm (@BER 10^-6)
-82dBm (@BER 10^-5)
Antenna Height
BS 15m, SS 1.5m
1.5m
Tx Power/Max range
33dBm / 3.2Km
20dBm / 500m
Default channel
Channel 1 : centered at 2412GHz
Channel 1 : centered at 2412GHz
Available channels
4 (non-overlap)
12 (overlapping)
32
802.11 & 16 Co-Existence: Reactive vs. CSCC-based Power Control (Jing-Raychaudhuri) 802.11b Hotspot DSS-AP
100m AP
1km SS
BS
Single 802.11 Hot Spot Case
802.16a Cell 0.7 1.2
0.6 0.5 0.4 0.3
No Coordination CSCC frequency adaptation
0.2 0.1 0.0 802.16 DL
802.11 link
Average Link Throughput (Mbps)
Average Link Throughput (Mbps)
0.8
1.0
0.8
802.16a DL 802.16a DL with CSCC 802.11 link 802.11 link with CSCC Average No Coordination Average with CSCC
0.6
0.4
0.2
802.16 DL and 802.11 link
CSCC frequency adaptation when DSS-AP =200m and traffic load 2Mbps
0.0 0
200
400
600
800
1000
Distance between 802.16 SS and 802.11 hotspot (meters)
Throughputs vs. DSS-AP by using CSCC power adaptation and traffic load 2Mbps
33
802.11 & 16 Co-Existence: Reactive vs. CSCC Power Control 802.16a BS
Multiple 802.11 Hot Spot Case
No Coordination Varying Max Cluster with CSCC Radius with RTPC with TA R16
120000
110000
1km
120000
Average Network Throughput (bps)
802.11b AP 802.11b Client
130000
No Coordination with CSCC with RTPC with TA
110000
100000
90000
80000
70000
100000
0.2
0.4
0.6
0.8
1.0
Clustering Index 90000
Max Hotspot Radius R11
80000
Region (i)
70000 0.2
0.4
0.6
Clustering Index
1 Mbps load
130000
0.8
1.0
Region (ii)
802.16 SS nodes are (i) uniform or (ii) clustered in the cell. clustering index CI = R11/R16
Average Network Throughput (bps)
Average Network Throughput (bps)
600 Kbps load
802.16a SS
No Coordination with CSCC with RTPC with TA
120000
110000
100000
90000
80000
70000 0.2
0.4
0.6
0.8
1.0
Clustering Index
CSCC power adaptation with clustered 802.16a SS in region (ii), with numbers of 802.16a SS : 802.11b nodes = 1:1
34
Cognitive Radio: Spectrum Policy Server
Internet-based Spectrum Policy Server can help to coordinate wireless networks (a “Google for spectrum”)
Needs connection to Internet even under congested conditions (...low bit-rate OK) Some level of position determination needed (..coarse location OK?) Spectrum coordination achieved via etiquette protocol centralized at server Spectrum Policy Server www.spectrum.net
Internet Internet AP1
Access Point (AP2)
WLAN operator A
Etiquette Protocol
AP1: type, loc, freq, pwr AP2: type, loc, freq, pwr BT MN: type, loc, freq, pwr
WLAN operator B Master Node
Wide-area Cellular data service
Ad-hoc Bluetooth Piconet
35
What can a Spectrum Policy Server do?
rate1
rate2
rate4
Spectrum Server
rate3
Spectrum Policy Server facilitates co-existence of heterogeneous set of radios by advising them on several possible issues Spectrum etiquette Interference information Location specific services Many more things ….
36
Scheduling Variable Rate Links with a Spectrum Server (Raman-Yates-Mandayam)
Users share a common frequency band
Wireless network of L directed links Links employ ON-OFF transmission schedule in each time slot
Use constant transmission power in the ON state
Links employ interference-adaptive modulation/coding
Orthogonal signal dimensions = time slots Time domain scheduling is used for channelization
Link rate in each time slot depends on interference, receiver mitigation.
Interference depends on the transmission mode
mode = subset of links that are ON simultaneously
1
1
4
tli = 1, if link l is in mode i
2 3
[0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1] [0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1] [0 0 1 1 0 0 1 1 0 0 1 1 0 0 1 1] [0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1]
= 0, otherwise. 3
Transmission mode [1 0 1 0] (one of 24 possible modes) 37
Spectrum Server Mode Matrix Æ Rate Matrix 1
4
1 Glk = link gain from Tx k to Rx l
2 3
3
network with 4 links
Transmission mode [1 0 1 0]
Rate matrix C = [6.6 [0 [0 [0
0 6.6 0 0
0.01 0.06 0 0
0 0 6.6 0
0.56 0 1.0 0
0 1.86 1.86 0
0.01 0 0.06 0 0.83 0 0 6.65
2.05 0 0 0.32
0 0.97 0 0.05
0.01 0 0.06 0 0 0.04 0.04 0.40
0.49 0 0.01] 0 0.77 0.06 ] 0.04 0.04 0.04 ] 0.19 0.05 0.04]
38
Spectrum Server: Optimizing the Mode Schedule
Spectrum server specifies xi = fraction of time mode i is ON
rl = Σi cli
Average rate in link l is
In vector form,
Spectrum server specifies x to optimize an objective such as for example:
xi
r = Cx
Maximum sum rate of the schedule Maximize the common rate on the links Fair scheduling Efficient scheduling
39
Spectrum Server: Scheduling Results
As common rmin increases,
the sum rate decreases Rates of dominant mode links decrease Rates of disadvantaged links increase
40
Spectrum Server: Fairness in Scheduling
41
Hardware & Software Platforms
42
Vanu Software Defined Radio
Vanu Inc. SDR programmable radio based on commodity processors. Supports multiple standards on handheld device.
Vanu Inc. Software Defined Radio Source: http://www.vanu.com/products.html 43
GNU Hardware/Software Platform
(Blossom)
GNU/USRP boards with GNU Software with API’s for flexible PHY and MAC are currently available for experimentation Various RF front-ends (0-100MHz, 400MHz, 900MHz, 2.4GHZ) with data rates upto 64 MSamples/sec All DSP functions in software on general-purpose CPU
44
WARP Platform
(Rice University)
•Various RF front-ends (2.4GHZ, 5GHz) with baseband of up to 20 MHz •High data rates possible, Programmable PHY & Protocols, Extensible to MIMO •Backed by Xilinx, Nokia
45
Cognitive Radio Networks & Protocols
Discovery strategies
Algorithms and protocols for frequency selection, coordination and cooperation
Multihop strategies
Algorithms for self-organization and routing
E.g. • Cognitive Radio scans for active nodes and executes discovery algorithm • Control protocol for exchange of network information and routing tables End-to-end routed path From A to F Bootstrapped PHY & control link
C
PHY A
B B
DD
PHY C
E
PHY B
A A Control (e.g. CSCC)
F
Multi-mode radio PHY Ad-Hoc Discovery & Routing Capability
Functionality can be quite challenging!
46
WINLAB Cognitive Radio WINLAB’s “network centric” concept for cognitive radio prototype
(..under development in collaboration with GA Tech & Lucent Bell Labs)
Requirements include:
~Ghz spectrum scanning, Etiquette policy processing PHY layer adaptation (per pkt) Ad-hoc network discovery Multi-hop routing ~100 Mbps+
47
Cognitive Radio Networks: “CogNet” Architecture
New NSF FIND project called “CogNet” aimed at development of prototype cognitive radio stack within GNU framework Joint effort between Rutgers, U Kansas, CMU and Blossom Inc
48
Cognitive Radio Networks: “CogNet” Protocol Stack
Global Control Plane (GCP)
Common framework for spectrum allocation, PHY/MAC bootstrap, topology discovery and cross-layer routing
Data plane
Dynamically linked PHY, MAC, Network modules and parameters as specified by control plane protocol
Data Plane Global Control Plane Control Plane
Data Plane
Control Signalling
Application Bootstrap
Disco very
Data Path Establish ment
Naming & Addres sing
Transport Network
Control MAC
MAC
Control PHY
PHY
49
Cognitive Radio Network Experiments Hardware/Software Platforms@WINLAB
ORBIT radio grid testbed currently supports ~10/USRP GNU radios, 100 low-cost spectrum sensors, WARP platforms, WINLAB Cognitive platforms and GNU/USRP2 Each platform will include baseline CogNet stack
Suburban
ORBIT Radio Grid
Current ORBIT sandbox with GNU radio 20 meters 500 meters Office
30 meters
Urban
300 meters
Radio Mapping Concept for ORBIT Emulator
400-node Radio Grid Facility at WINLAB Tech Center Programmable ORBIT radio node
Planned upgrade (2007-08)
URSP2 CR board
50
Concluding Remarks
51
Wireless/Mobile/Sensor Scenarios and the Future Internet – NSF’s GENI Project
Some architectural and protocol implications for the future Internet...
Integrated support for dynamic end-user mobility
Wireless/mobile devices as routers (mesh networks, etc.)
Network topology changes more rapidly than in today’s wired Internet
Significant increase in network scale (10B sensors in 2020!)
New ad hoc network service concepts: sensors, P2P, P2M, M2M,…
Addressing architecture issues – name vs. routable address
Integrating geographic location into routing/addressing
Integrating cross-layer and cognitive radio protocol stacks
Data/content driven networking for sensors and mobile data
Pervasive network functionality vs. broadband streaming
Power efficiency considerations and computing constraints for sensors
Many new security considerations for wireless/mobile
Economic incentives, e.g. for forwarding and network formation 52
NSF GENI Implementation: Wireless Subnets – Overall Wireless Deployment Plan
Five types of experimental wireless networks planned – necessary to support full range of protocol research and to enable new applications
1. Wireless emulation and simulation (repeatable protocol validations) 2. Urban 802.11-based mesh/ad-hoc network (real-world networking experience with emerging short-range radios) 3. Wide-area suburban network with both 3G/WiMax (wide area) and 802.11 radios 4. Sensor networks (…application specific, specific system TBD via proposal process; may include environmental, vehicular, smart spaces, etc.) 5. Cognitive radio network – advanced technology demonstrator (…adaptive, spectrum efficient networks using emerging CR platforms) …also some common network facilities such as location & dynamic binding services
Each network at a different geographic location – new spectrum allocation may be needed at some sites 53
NSF GENI Implementation Wireless Sub-Networks Overview Location Service
Emerging 5 Technologies (cognitive radio)
Advanced Technology Demonstrator (spectrum)
Broadband Services, Mobile Computing
Other GENI services Infrastructure
Ad-Hoc Mesh 2 Network
1 NSF Radio Testbeds
Open API 3 Wide-Area Networks
“Open” Internet Concepts for Cellular devices
Sensor 4 Networks Embedded wireless, Real-world applications Protocol & Scaling Studies
Emulation & Simulation 54
Cognitive Radio in NSF’s GENI Project
Propose to build advanced technology demonstrator of cognitive radio networks for reliable wide-area services (over a ~50 Km**2 coverage 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 proposals Requires enhanced software interfaces for control of radio PHY, discovery and bootstrapping, adaptive network protocols …….. suitable for protocol virtualization FCC experimental license for new cognitive radio band
Cognitive Radio Network Node
Cognitive Radio Client
Spectrum Server
Cognitive Radio Network Node
Cognitive Radio Client
Spectrum Monitors
Connections to GENI Infrastructure Research Focus: 1. New technology validation of cognitive radio 2. Protocols for adaptive PHY radio networks 3. Efficient spectrum sharing methods 4. Interference avoidance and spectrum etiquette 5. Dynamic spectrum measurement 55 6. Hardware platform performance studies
Concluding Remarks
Future wireless networks need ~100-1000x increases in density and bit-rate of radios Æ motivates better spectrum coordination methods Spot shortages of spectrum will occur if present static allocation is continued Æ significant improvement achieved with dynamic allocation Cognitive radio technologies can be characterized in terms of the combination of hardware complexity & level of protocol coordination Promising cognitive radio schemes include
Agile radio with interference avoidance
Spectrum etiquette protocols: spectrum server, CSCC..
Adaptive networks via ad-hoc collaboration
Early technical results now available for some of these methods, but very different complexity factors and market implications… 56
Concluding Remarks
Future research areas in cognitive radio include:
New concepts and algorithms for agile radio and spectrum etiquette protocols
Architecture and design of adaptive wireless networks based on cognitive radios
Detailed evaluation of large-scale cognitive radio systems using alternative methods
Spectrum measurement and field validation of proposed methods
Cognitive radio hardware and software platforms
User-level field trials of emerging cognitive radios and related algorithms/protocols may also be useful to gain experience
Controlled testbed experiments comparing different co-existence methods
Large-scale “spectrum server” trial for 802.11x coordination
Experimental deployments in proposed US FCC cognitive radio band
Success with cognitive radio technologies should lead to major improvements in spectrum efficiency, performance and interoperability 57
