Transcript IEEE DySPAN 2010 Demonstrations Cavin Wang, IDA, Singapore Ser Wah Oh, I2R, Singapore Przemysław Pawełczak, UCLA, USA.

IEEE DySPAN 2010 Demonstrations
Cavin Wang, IDA, Singapore
Ser Wah Oh, I2R, Singapore
Przemysław Pawełczak, UCLA, USA
IEEE DySPAN 2010 Demos: Stats
• 12 Demos submitted, 10 accepted, 9
presented
• Number of IEEE DySPAN demos stays relatively
constant over the years
– Dublin (9), Chicago (13)
• Location
– EU (5 - !), USA (2), Canada (1), Singapore (1)
Software Defined Radio (SDR)
Implementation of Spectrally
Modulated Spectrally Encoded (SMSE)
Based Overlay Cognitive Radio (CR)
Ruolin Zhou, WSU
Clifton Bullmaster, AFRL
Cognitive Radio
Overlay CR – utilizes the white
space (unused spectrum)
Underlay CR – UWB
B  500MHz
Hybrid Overlay/Underlay CR –
utilizes both the white space and
the gray space. Do not need UWB
for underlay
V. Chakravarthy, Z. Wu, M. Temple, and F. Garber, “Novel Overlay/Underlay Cognitive Radio Waveforms Using SD-SMSE Framework to
Enhance Spectrum Efficiency - Part I: Theoretical Framework and Analysis in AWGN Channel," IEEE Transactions on Communications,
vol. 57, no. 12, pp. 3794-3804, December 2009
SMSE Framework
M. L. Roberts, M. A. Temple, R. A. Raines, R. F. Mills, and M. E. Oxley, “Communication Waveform Design Using an Adaptive Spectrally Modulated
Spectrally Encoded (SMSE) Framework,” IEEE Journal of Selected Topics in Signal Processing, June 2007
Demonstration
• Flexibly generates SMSE based non-contiguous OFDM, MC-CDMA,
CI/MC- CDMA, and TDCS waveforms to take advantage of multiple
spectrum holes
•Adaptively avoids interference from and to the primary users, and
intelligently provides coexistence
•Future Work – “SD-SMSE Based Hybrid Overlay/Underlay CR”
Cognitive Radio for Home
Networking
Vladimir Atanasovski (Faculty of Electrical Engineering and Information
Technologies - Skopje, MK); Daniel Denkovski (Faculty of Electrical
Engineering and Information Technologies, MK); Tim Farnham (Toshiba
Research Europe Ltd., UK); Liljana Gavrilovska (Faculty of Electrical
Engineering and Information Technologies, MK); Alain Gefflaut (European
Microsoft Innovation Center, DE); Vinay Kolar (Carnegie Mellon
University, QA); Petri Mähönen (RWTH Aachen University, DE); Elena
Meshkova (RWTH Aachen University, DE); Benjamin Motz (Toshiba
Research Europe Ltd., UK); Jad Nasreddine (RWTH Aachen University, DE);
Valentina Pavlovska (Faculty of Electrical Engineering and Information
Technologies, Skopje, MK); Marina Petrova (RWTH Aachen University,
DE); Sadia Quadri (Toshiba Research Europe Ltd., UK); Krisakorn Rerkrai
(RWTH Aachen University, DE)
What has this man to do with…
… wireless home networking
ARAGORN…
http://www.ict-aragorn.eu/
– A cooperation project between 4 universities and 4 companies:
– RWTH Aachen University, CFR, UCL and Univ. Ss. Cyril and Methodius
– Microsoft, Toshiba, ST Microelectronics and Huawei
– Develops Cognitive Radio and DSA solutions with learning
capabilities for low-cost commercial applications.
– Highlighting in DySPAN 2010:
– Cognitive Resource Management Architecture
– Cross-Layer Optimization and Interference Management
– Policy Management and Application Priorities through hierarchical policy servers
Welcome to our demo!
Decomposable MAC Framework for
Highly Flexible and Adaptable MAC
Realizations
Junaid Ansari, Xi Zhang, Andreas Achtzehn,
Marina Petrova, Petri Mähönen
Institute for Networked Systems
RWTH Aachen University, Germany
Concept
• Decomposition of MAC protocols into fundamental functional blocks
based on the commonalities among different MACs.
• Realization of a particular MAC solution by binding the blocks together
appropriately through a Wiring Engine.
• On-the-fly composition and reconfiguration of MAC protocols with high
degree of code reuse.
• A key enabling technology for implementing and prototyping Cognitive
Radios and dynamic wireless devices.
MAC 1
MAC 2
Design and Implementation
• Granular MAC blocks are
implemented with flexible APIs on
WARP boards.
• A MAC Description Language eases
implementation effort for users.
• Interpreter translates user inputs
into executable instructions.
• Wiring Engine coordinates data and
control flow between blocks and
allows run-time configuration by
block insertion and removal
through a set of dependency tables.
Performance
Plots
Flowchart
Visualization
Interpreter
Host PC
Wiring
Engine
Code
Execution
Performance
Statistics
WARP Board
Implementation Modules
Demonstration and Visualizations
• Users can interactively
compose and modify MACs at
runtime through flowcharts.
• Corresponding autogenerated MAC code is
shown.
• Live performance statistics of
the MAC is displayed.
• A spectrum-agile MAC
developed using the
framework is shown to
reconfigure based on the user
controlled interferences.
Shared Spectrum
WARP
WARP
Primary user
Ethernet
PC
PC
Demonstration of Sequence Detection
Algorithms for Dynamic Spectrum
Access Networks
Zhanwei Sun, Glenn J. Bradford and J. Nicholas
Laneman
Department of Electrical Engineering, University
of Notre Dame, USA
Sequence Detection Algorithms for
Dynamic Spectrum Access Networks
• Energy Detection does not consider the PU’s channel access
pattern.
•
Sequence Detection
- Based upon hidden Markov
model, integrating memory
into spectrum sensing
- Different cost factors for missed
detections and false alarms
- Minimizing detection risk
Network Setup
• A PU pair and a SU pair operate
at the same frequency band,
with video streaming for each
user
• Primary transmitter accesses the
channel in a Markov chain
• Secondary transmitter accesses
the channel opportunistically on
detecting spectrum hole
Demonstration
Cognitive, Radio-Aware, Low-Cost
(CORAL) Research Platform
John Sydor, Siva Palaninathan, Bernard Doray,
David Roberts, Muhmudar Rahman, Li Pan,
Jiangsin Hu, Amir Ghasemi, Wayne Brett, Larry
Stone
Communications Research Centre, Canada
What is CORAL*?
A Wi-Fi® router with a cognitive radio control shell around it, thus creating the WIFI_CR unit
 WIFI_CR: has IEEE 802.11g PHY attributes. However with the CR_NMS control system we implement a cognitive
radio as defined by the ITU….which uses environment knowledge, dynamically & autonomously adjusts, learns…
It implements all the functionality of CR: Radio Environment sensing, virtual environment memory, cognitive engines,
control channel, undertakes network and terminal re-configurability, and can be used create numerous wireless topologies:
Mesh, Pico-cell networks, Femtocells, P-MP/P-P, relays, etc..
CORAL is a CR development platform allowing implementation of Cognitive Networks in the ISM band…where
interference, fallow spectrum, primary users, and poor propagation are the norm….If Cognitive Radio can solve wireless
problems in the ISM band, it will probably solve them in other, less demanding band…like the TV bands
 Will give developers fresh approaches to wireless…especially in the ISM band which uses a technology ( WIFI/IEEE
802.11) that is not spectrum efficient in high interference and is in need of improvement after 15 years of the same old
access algorithms…
How about a cognitive ISM band MIMO router in the home that shares spectrum with its neighbors..and acts as a
femtocell for cellular? New approaches to old wireless concepts.
A demonstration of CORAL’s CR capabilities..
for Dyspan 2010
(1) Creation of a Radio Environment Awareness MAP
A virtual representation of the Radio Environment
is required for learning and decision making by
the Cognitive Engines. We will show how CORAL:
Captures full ISM band WI-Fi interference by
providing occupancy information, interference
power, identity, IP Link associations; undertakes
spectrum analysis; can incorporate specific sensors,
Can ‘Sniff’ specific sectors capturing interference
that is spatially dependent,
Collects throughput and channel utilization data by
the members of the CRN to aid in bandwidth allocation,
Creates a map of the interference and occupancy attributes of the
CRN that can be searched by time, space, spectrum, RSSI, identity,
IP link, occupancy, etc.
A demonstration of CORAL’s CR capabilities..
for Dyspan 2010
(2) Dynamic Spectrum Assignment
Using the REAM and Sensor information, the CORAL CRN ( AP with 3 clients)
selects the most appropriate ISM channel based on occupancy, interference.
power level, duration, and user terminal’s bandwidth (fairness) requirements.
Dynamically moves to alternate channel when interference environment
changes.
(3) Primary User detection-alternative channel move
On detection of a Mimicked Primary User that appear on-channel, CORAL moves
to an alternative channel…mimicking TV band/radar detection type actions.
(4) Spatial Selection for Interference Mitigation
 Demonstration of how CORAL system can change its reception pattern, allowing
selection of direction and sectors less prone to interference.
Demonstration of CORAL’s TDD/TDMA Wi-Fi capability; per packet directional
switching that can be used by cognitive engines with spatial interference
knowledge.
OFDM Pulse-Shaped Waveforms for
Dynamic Spectrum Access Networks
Paul Sutton, Barış Özgül, Irene Macaluso, Linda Doyle
CTVR at University of Dublin, Trinity College, Ireland
OFDM Pulse-Shaped Waveforms for Dynamic Spectrum
Access Networks
Paul Sutton, Barış Özgül, Irene Macaluso, and Linda Doyle
CTVR at University of Dublin, Trinity College, Ireland
BACKGROUND:
•
•
OFDM is the modulation scheme preferred in many wireless communication systems (DSA
networks, DVB, ISDB, variants of Wifi, Wimax, LTE, LTE-advanced ...)
OFDM has flexibility to support adaptive bit/power loading, embedded signatures, noncontiguous transmissions, and pulse shaping
DEMO:
•
GOAL: Suppressing out-of-band radiation of an OFDM signal through shaping
–
•
for coexistence of more signals in a limited frequency band
SCENARIO: A high-power OFDM-based secondary transmission at a frequency adjacent to an
OFDM-based primary system
–
–
–
Secondary Tx is positioned next to the primary Rx
Baseband Tx/Rx chains run on our highly reconfigurable Iris 2.0 software radio platform. USRP is the
RF front-end
Primary system transmits audio over air
– Primary RX cannot receive and play audio due to adjacent channel
interference, when secondary Tx applies no shaping
PAPERS in DySPAN 2010: 1) “Experiences from the Iris Testbed in Cognitive Radio and Dynamic Spectrum Access
Experimentation” – Thu, 14:15-15:45, Room: Ocean 1 , 2) “Dynamic Block-Edge Masks (BEMs) for Dynamic Spectrum
Emission Masks (SEMs)” – Thu, 14:15-15:45, Room: Ocean 3
• SETUP: 3 pairs of “1 Laptop+1 USRP” for interfering Tx,
primary Tx and primary Rx
– Laptops run baseband TX and RX chains implemented on Iris 2.0
– Baseband samples transferred to/received from USRP over USB
– USRP transmits/receives signal over air
Suppression of adjacent channel
interference through shaping
No shaping → Harmful adjacent
interference, no audio
OFDM-based Dynamic Spectrum
Access
Milan Zivkovic, Dominik Auras, Rudolf Mathar
RWTH Aachen University
Demonstration scenario
detectscoexistence
of unused
spectrum
and
adapts
its transmission
 SU
Thesystem
Interference-free
performance
ofparts
PU transmission
of two
OFDM
is not affected
systems
by
within
SU
a
parameters
(used subchannels,
rate and power allocation) satisfying
common frequency
communication
band
given requirements (constrained total power, required rate, BER)
fc
PU
f
500 kHz
SU
USRP
USRP
USRP
USRP
2 MHz
Lehrstuhl für Theoretische
Informationstechnik
f
28
System architecture
 Reconfigurable continuous one-way
transmission of OFDM symbol frames
 Baseband signal processing is
implemented in GNU Radio
 Blocks for adaptive (de)mapping and
power loading allows for capacity
achieving functionality
 The backbone of the system is realized
over local Ethernet network by CORBA
event service
 The central control unit (resource
manager) determines optimal
transmission parameters for given
requirements
 Resource manager can be easily configured for different DSA scenarios
Lehrstuhl für Theoretische
Informationstechnik
29
Digital and Analog Solution for
Low-power Multi-band Sensing
Sofie Pollin, Eduardo Lopez, Anthony Antoun,
Peter Van Wesemael, Lieven Hollevoet, Andre
Bourdoux, Antoine Dejonghe, Liesbet Van der
Perre
IMEC
A single reconfigurable analog front-end: Scanning from
100MHz to 6GHz
Prototype demonstrating sensing
capabilities of IMEC Scaldio2B RFIC
500 MHz
• 40 nm RFIC
• On chip ADC
• 100 MHz -> 6 GHz
imec/restricted 2010
31
2.5 GHz
A sensing enabled digital front-end:
Further band selection and FFT processing
imec confidential 2009
An algorithm for multi-band sensing:
Iterative Leakage Removal
Transmitted
Signal:
Many small
spectrum holes
Filtered
Received Signal
After FFT
After leakage
removal
Holes
Identified
33
TV White Space Video Streaming
Demo
Ser Wah Oh, Yonghong Zeng, Weiqiang Zhang,
Syed Naveen A. A., Francois Chin
Institute for Infocomm Research (I2R), A*STAR
Imagination to Reality
http://www.i2r.a-star.edu.sg
Goals and Architecture
• Goals
– Testing spectrum sensing in real-world environment
– Showcasing opportunistic utilization of unoccupied
spectrum for communication
• Architecture
TV WHITE-SPACE SPECTRUM SENSING PROTOTYPE
DIGITAL HARDWARE PLATFORM
PC PLATFORM
TV Band
UHF RF
Monopole
Antenna
RF Front End
Digital Front End
RF Receiver Broadband
Tuner
ADC & DDC
44MHz IF
Covariance
Algorithm
5.5MHz IF
Spurious
Detection
Jun 7, 10
Graphical User
Interface
Spectrum
Manager
Imagination to Reality
http://www.i2r.a-star.edu.sg
Demo 1
Jun 7, 10
Imagination to Reality
http://www.i2r.a-star.edu.sg
Demo 2
Antenna
DTV
Modulator
Antenna
Tunable RF
Attenuator
DTV
Demodulator
TV WHITE SPACE DEVICE
Frequency: 512 – 698 MHz
Bandwidth: 6 MHz
Jun 7, 10