1
High Performance Cognitive Radio Platform with Integrated Physical & Network Layer Capabilities
Bryan Ackland, Ivan Seskar WINLAB, Rutgers University bda@winlab.rutgers.edu seskar@winlab.rutgers.edu www.winlab.rutgers.edu
High Performance Cognitive Radio Platform with Integrated Physical - - PowerPoint PPT Presentation
High Performance Cognitive Radio Platform with Integrated Physical & Network Layer Capabilities Bryan Ackland, Ivan Seskar WINLAB, Rutgers University bda@winlab.rutgers.edu seskar@winlab.rutgers.edu www.winlab.rutgers.edu 1 Dynamic
1
Bryan Ackland, Ivan Seskar WINLAB, Rutgers University bda@winlab.rutgers.edu seskar@winlab.rutgers.edu www.winlab.rutgers.edu
2
Large, increasing demand for wireless services Static frequency bands allocated to single service
Inefficient use of spectrum Slow, expensive political process Locally optimized incompatible solutions
FCC exploring alternatives
ISM & U-NII bands
Power and BW limitations to allow co-existence Successful but quickly getting congested
Intelligent or “Cognitive” radios that adapt to local
Improve spectrum efficiency and fairness
3
Programmable radio systems that adapt to:
Changing radio interference Availability of nearby collaborative nodes Changing protocols & standards Application requirements
by modifying
Frequency, power, bandwidth Modulation, coding, MAC Network protocols
and coordinating with other cognitive systems to
4
Tradeoff between flexibility, performance & power “Moore’s Law” improvements in CMOS VLSI
Implement some functions in SW Ultimate goal: software radio?? Reality: some combination of HW, SW and reconfigurable logic
Silicon area efficiency
flexibility speed, power, cost 1 10 100 1000 Microprocessor DSP FPGA ASIC
A/D µP
5
Investigate Cognitive Radio Strategies &
Explore flexible, power efficient wireless
Develop board level platform for system
6
WINLAB, Rutgers University
Bryan Ackland Ivan Seskar
Chris Rose
GEDC, Georgia Institute of Technology
Joy Laskar Stephane Pinel
Wireless Res. Lab., Lucent Bell Laboratories
Tod Sizer Dragan Samardzija
7
Design & build cognitive radio platform that is
High performance HW & SW Programmable Physical, baseband & network layer adaptable Support wide range of spectrum sharing scenarios
Leverage today’s high performance off-the-shelf
Demonstrate architectures and components that
8
Spectrum Allocation Rules (static)
INTERNET BTS
Auction Server (dynamic)
Spectrum Coordination Server (dynamic)
AP
Ad-hoc sensor cluster (low-power, high density) Short-range infrastructure mode network (e.g. WLAN) Short-range ad-hoc net Wide-area infrastructure mode network (e.g. 802.16)
wireless devices, both wide- area and short-range
radio technologies, e.g. 802.11a,b,g, UWB, 802.16, 4G, etc.
allocation rules evolve to achieve high efficiency?
(interference avoidance)
allocation methods
coordination (etiquette)
networks
Etiquette policy Spectrum Coordination protocols Spectrum Coordination protocols Dynamic frequency provisioning
9
Hardware Complexity
“Open Access” + smart radios
Protocol Complexity (degree of coordination)
Reactive Rate/Power Control Reactive Rate/Power Control Agile Wideband Radios Agile Wideband Radios Unlicensed Band with DCA (e.g. 802.11x) Unlicensed Band with DCA (e.g. 802.11x) Internet Server-based Spectrum Etiquette Internet Server-based Spectrum Etiquette Ad-hoc, Multi-hop Collaboration Ad-hoc, Multi-hop Collaboration Radio-level Spectrum Etiquette Protocol Radio-level Spectrum Etiquette Protocol Static Assignment Static Assignment Internet Spectrum Leasing Internet Spectrum Leasing
“Cognitive Radio” schemes
UWB, Spread Spectrum UWB, Spread Spectrum Unlicensed band + simple coord protocols
10
Spectrum scanning & frequency agility Fast physical layer adaptation & power control to
Flexible baseband & MAC switchable on a
Capable of higher layer spectrum etiquette or
Simultaneous heterogeneous radio links Protocol translation & routing to support
11
Separate sub-systems to simplify functional implementation
Flexible RF Flexible RF Flexible RF Flexible Baseband (SDR) Network Processor (MAC+) CR Strategy (host)
local drop
Flexible Antenna
12
Flexible RF Flexible RF Flexible RF Flexible Baseband (SDR) Network Processor (MAC+) CR Strategy (host) Flexible Antenna A/D/A A/D/A A/D/A
Baseband & Network Processor Board (Rutgers & Lucent) Antenna & RF Board (Georgia Tech.) A/D/A Board (Rutgers)
13
Tri-band operation:
700-800 MHz 2.40-2.48 GHz ISM band 5.15-5.825 GHz ISM and UN-II bands
2 Transmit + 2 Receive channels for data + spectrum
20 MHz bandwidth on each channel tunable over band
Narrow band selection performed at baseband
100mW transmit power (variable) per channel Sensitivity & linearity to meet 802.11a
14
To baseband A/D’s tri-band antenna tri-band VGA
tri-band RX 20 MHz BW IF filter Power detection Standard identification Tri-band Sensing /Monitoring Unit
tri-band RX 20 MHz BW IF filter
tri-band RX 20 MHz BW IF filter
SW
M A T R I x 800 MHz 2.4 GHz 5.2 GHz 800 MHz 2.4 GHz 5.2 GHz AgileTriband LNA + Agile High Q matching network
tri-band antenna
SOC SOP
Low-IF ~150 MHz I Q I Q Channel 1 Channel 2 I Q I Q I Q
15
2 3 4 5 6 7 0.25 0.3 0.35 0.4 0.45 0.5 0.55 Band-I Band-II
Oscillation Frequency (in GHz) Vtune (in V)
Switched-L Frequency Agile VCO
16
Triple-Broadband Antenna for handheld terminals
PCB
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 1 2 3 4 5
VSWR Frequency (GHz)
Frequency Range (MHz): 810-1000 1600-2500 4000-6000 VSWR: ≤1.5 Pattern (azimuth plane) : Omni-directional Non-omni Peak Gain (azimuth plane) : 0 dBi 3 dBi Polarization: Mixed Antenna dimensions: 50 mm ×50 mm ×0.2 mm
17
FR-4 Organic high density multi-layer Reconfigurable CMOS RFIC Multi- band/wideband antenna. RF Tx / Rx RF-MEMS Switch Flexible baseband
L 1 O u t V D D C 1 R 3 R Vg ain C 4 C 5 V b Q 1 GN D G N D G N D Q 3 R G N D C 2 C 3 Q 2 R 2 G N D V D D R 1 L(active) V D D 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 1 2 3 4 5
VSWR Frequency (GHz)
Reconfigurable CMOS RFIC RF-MEMS Switch & Multi-band Antenna
18
Interface to multiple radio channels Real time spectral analysis Support comparison of HW & SW baseband
MAC, protocol conversion, SAR, routing Data rates (total) up to 100 Mb/s Support novel reconfigurable architectures in
Clean partitions between Baseband, NP and CR Simple programming environment (not DSP) Fast reconfiguration time (~µs)
19
MPC8260 TMS320C6701 XC2V6000 FPGA Ethernet
6M gates programmable logic 2.5 Megabits DPRAM in
FPGA
144 dedicated multipliers 1 GFLOPS TMS320C6701 280 MIPS MPC8260 244 configurable I/O pins
Megarray Connector- 244 Configurable I/O pins
20
(XC2V3000) each with 256K x 18 ZBT SRAMs
External Interface
Configuration Control
and four 100 MHz 12-bit D/A channels
programmable clock
FLASH
21
Baseband FPGA Virtex-4 94K logic cells 160 DSP slices
PowerPC (RTOS)
Network FPGA Virtex-4 94K logic cells Soft RISC cores SRAM (4MB) SDRAM (128MB) PowerPC PowerQuick III 600 MHz (LINUX) DRAM (64MB)
64
EEPROM (config)
Data, control & sensing to/from RF front-end Gig-E USB-II
22
Packet Scheduler (RISC) Header Buffer Packet Buffer (DRAM) Packet Processor (RISC/reconfig) Packet Processor (RISC/reconfig) Packet Processor (RISC/reconfig) Local I&D Local I&D External DRAM Local I&D
to/from baseband to/from CR host
23
Need efficient multi-user, multi-proc. compile &
Short learning curve for student SW developers
Linux OS with Gnu tool chain
Open source Modular: I/O drivers can be installed without kernel
User friendly development environment
Simulink models compiled to VHDL and/or C
24
2-3 Cognitive Radio Scenario Demos*
System Baseband & Network Proc. RF Front-end
Integrated SIP/SOC agile tri-band radio Y3 Q4 Y3 Q3
Y3 Q2 Y3 Q1 System Prototype based on:
Prototype boards available Y2 Q4 Prototype Software Dev. Env. Y2 Q3 Agile prototype – mainly off the shelf – some custom components – full functionality Y2 Q2 Y2 Q1 Proof of concept system prototype based on:
schematics.
Initial prototype – off the shelf components – limited flexibility Y1 Q4 Detailed Architecture Specification Y1 Q3 Result of HW (FPGA) and SW implementation studies Detailed performance & interface specs (12/04) Y1 Q2 Y1 Q1 *Note: Further release of Cognitive Radio Boards to community contingent on separate funding