Transcript Ad-Hoc Network

WINLAB IAB Meeting
June 10, 2005
Rutgers, The State University of New Jersey
www.winlab.rutgers.edu
Contact: Professor D. Raychaudhuri, Director
[email protected]
1
WINLAB STATUS
UPDATE
2
WINLAB Status Update
WINLAB activity snapshot as of Spring 2005:
~25 faculty/staff (15 academic faculty + 10 research staff/adjunct faculty)
~45 graduate students
~14 companies in corporate sponsor program
25,000 sq-ft in facilities, including new Tech Center II building
Industry funding ~$1M (including both annual sponsorship and focus projects)
$3M+ federal research funding, mostly from NSF
~$500K in NJ State + Rutgers funding (...RU portion increasing in FY05)
Total funding level ~$4.5M in FY’04 (...300% increase over FY’01)
3
Status Update: Faculty List 6/05
Radio/ Modem
Technology
Y. Lu
M. Bushnell
B. Ackland1
P. Spasojevic
L. Greenstein
R. Rajnarayan (Research
Engineer)
K. Wine (Research Engineer)
P. Henry (AT&T Labs)*
Students:
PhD – 5
MS – 3
Radio Resource
Management
&
Wireless Systems
R. Yates
C. Rose
N. Mandayam
D. Frenkiel
Z. Gajic
D. Raychaudhuri
9/04
W. Trappe
I. Seskar (Assoc Dir IT)
R. Siracusa (Research Specialist)1
M. Ott1
3/05
R. Howard1
L. Razoumov (Intel)*
Students:
PhD – 10
MS – 4
Sensor Nets
& Pervasive
Computing
Mobile Network
Architecture &
Protocols
S. Paul (Edgix)*
H. Liu (Thomson)*
A. Acharya (IBM)*
M. Gruteser
B. Nath
H. Hirsh
M. Parashar
Y. Zhang
R. Martin
Students:
PhD – 3
MS – 4
Students:
PhD – 8
MS – 6
* Adjunct Prof
1 Part-time position
4
Status Update: Sponsor Program
Currently ~13 sponsor companies
Recently added 1 new sponsor: US Army CECOM
Target no more than ~10-15 companies, with close engagement
~2-3 industry focus projects currently in progress
MIMO Infostations (STTR for ARL)
3G Security (NICT, Japan)
Carrier ad-hoc networks (under discussion with NTT DoCoMo)
Increasing collaboration with sponsors on large Govt proposals
NSF MIMO project (DAPHNE) - Lucent
ORBIT wireless networking testbed – Thomson, Lucent, IBM
Cognitive radio algorithms and hardware – Lucent
More joint proposals with sponsor/partner companies on key topics
Cognitive radio – Philips, GNU Radio, Raytheon
Pervasive computing, sensor systems – TBD
5
Status Update: Industry Sponsors 6/05
*
Aruba Networks, *
PnP Networks,
Semandex Networks
Mayflower Inc.
Panasonic
US Army CECOM
*Research Partners
6
Status Update: Sponsor Program
Seeking more long-term research collaboration with sponsors
Now that we have invested in critical technology areas, new
labs and a larger, more qualified student pool, we invite
sponsors to work more closely with us:
Specific focus projects on topics of mutual interest
Contributions to existing projects such as ORBIT or NSF “future Internet” project
Joint proposals to future NSF, DARPA, DHS or DoD RFP’s
Visiting researchers, short sabbatical leaves, etc.
Sponsored students, post-docs and student internships
ORBIT facility (~11,000 sq-ft in Rt 1 Tech Center bldg) has
adequate space for research visitors
Also starting to work more actively with early stage
incubations, startup companies and joint ventures...
7
Status Update: WINLAB Activity
Model & Tech Transfer
Sponsor Fees,
& Govt basic
research funds
Core Research Areas
New system concepts, IPR, …
Additional
Project
Support
DARPA
Projects
(e.g.
Infostations)
Major NSF
Projects
(e.g. ORBIT)
NJCST
Project
(e.g. MUSE)
Tech Reports,
Sponsor meetings,
Software tools,
etc.
Focus
Project(s)
with Sponsor
Companies
Corp
R&D
Usually involves partnerships with sponsor companies
And other universities
Pre-commercial technology
Activities
to be carried out at
Tech Center II
Industry, venture
funds, NJCST, …
Technology Transfer Projects
8
Status Update: Research Program 6/05
Research projects in 4 broad areas of wireless technology
radio propagation and modem design
radio research management (RRM)
wireless networks and protocols
mobile computing
Major NSF projects on future wireless networks (ORBIT), spectrum,
cognitive radio, MIMO, sensors and security/privacy
Numerous smaller projects (both NSF and industry) on topics ranging
from WLAN enhancements and 3G scheduling to network coding and
location services.
Strategic future directions: “wireless ecosystems”, security, nextgeneration Internet and pervasive systems....
9
Status Update: WINLAB Research
Direction
MSC
Internet (IP-based)
Public Switched Network
(PSTN)
Custom
Mobile
Infrastructure
(e.g. GSM, 3G)
Research Themes:
Super-fast short range radios
UWB, MIMO
Sensor devices/SOC
4G radio & next-gen WLAN
Spectrum coordination
Unified mobility protocols
Ad-hoc network RRM , MAC
and routing protocols
Ad-hoc net QoS & security
Sensor net software models
Centralized control 
distributed
etc.
Generic mobile infrastructure
BSC
BTS
WLAN
Access
Point
BTS
Infostation
cache
WLAN
Hot-Spot
VOIP
CDMA, GSM
or 3G radio
access network
Research Themes:
Faster radios
Interference issues
Power control
3G Scheduling
Handoff algorithms
WLAN MAC
3G/WLAN interworking
Security
Mobile content
etc.
Today
Ad-hoc
network
extension
Broadband Media cluster
(e.g. UWB or MIMO)
Future?
VOIP
(dual-mode)
Low-tier clusters
(e.g. low power 802.11 sensor)
10
Status Update: Research Areas
Cognitive Radio
Wireless Sensors
Radio Platforms
Pervasive Computing Application
Agent 2
Agent 1
Agent 3
Wireless Network Testbed
Overlay Network for
Dynamic Agent <-> Sensor
Association
Ad-Hoc Networks
Sensor
Cluster B
Sensor
Cluster A
Resource
Discovery
Ad-hoc
Routing
OS/Process
Scheduling
Run-time
Environment
(network OS)
Wireless/Sensor Net
Software & Security
Packet delivery reliablilty
1
Original 802.11
CR, denisty 1
CR, denisty 2
CR, denisty 4
CR, denisty 8
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0
0.5
1
1.5
2
Normalized distance of AB
System Analysis & Theory
Mobile Computing
11
Status Update: Wireless Roadmap
System
Applications
3G Cellular
WLAN office/home
GSM, GPRS services
Protocols
& Software
3G services
Mobile WLAN services
Cellular handset, BTS
Hardware
Platforms
home media
networks
public WLAN
Broadband Cellular (3G)
Basic
Wireless
Technologies
WLAN (802.11a,b,g)
~ 1 Mbps Bluetooth
2000
Self-Organizing Ad-Hoc
Radio Router
Commodity BTS
Embedded Radio
(wireless sensors)
802.11 Mesh Router*
IP-based Cellular Network (B3G)
WLAN+ (802.11e,n)
~100 Mbps UWB
2005
Multi-standard
Cognitive Radio*
Unified Wireless Access
+ IP-based core network
~100 Mbps OFDM/CDMA
~10 Mbps OFDM
Sensor radios
(Zigbee, Mote)
Sensor Nets
Next-Gen WLAN
(including ad-hoc mesh)
3G Base Station Router
~50 Mbps OFDM
Ad-Hoc & P2P
IP-based Mobile Network
WLAN security, enterprise
ad-hoc/mesh
Pervasive Systems
Mobile Internet Services
& Content Delivery
3G/WLAN interworking
~2 Mbps WCDMA
~11 Mbps QPSK/QAM
Mobile Internet
open systems
Cellular VOIP gateway
802.11 WLAN card/AP
Bluetooth module*
4G Systems
3G/WLAN Hybrid
dynamic
spectrum
sharing
~200 Mbps MIMO/OFDM
~500 Mbps UWB
2010+
12
Status Update: WINLAB R&D map
ORBIT Wireless Network Testbed
System
Prototypes
Protocols
& Software
MUSE System
Prototypes
Infostations Prototypes
(i-media, emergency response)
3G/WLAN
Interworking
Core
Technology
UWB PHY/MAC
SDR
Prototype
802.11e,n
protocols
Wireless security
Multimodal ZnO sensor
Adaptive Radio
Network Prototype
Ad-hoc net with QoS
Self-organizing
Ad-hoc network
Ad-hoc routing Content Routing
in mobile networks
MIMO
Infostation
Spectrum etiquette and adaptive
radio net protocols
Sensor net Privacy
Multimodal sensor-on-silicon
(MUSE) module/chip
Low-power
802.11b
Network-centric
Cognitive radio HW
nx100 Mbps
OFDM Radio
Sensor net models
Interference avoidance, RRM
Algorithms,
Analysis &
Simulation
MIMO networks
3G/4G PHY/MAC (RRM, scheduling, etc.)
Ad-hoc net RRM
OFDM
UWB
2002
Spectrum rights & management
2004
Unlicensed spectrum algorithms
2006
13
Status Update: Major Projects

Several major research and technology transfer projects
currently being carried out at WINLAB
Major government projects






Dynamic spectrum management (NSF ITR, ’02-’05)
Multimodal Sensor-on-Silicon: MUSE (NJCST, ’02-’07)
ORBIT: Open-Access Research Testbed for Wireless Networks (NSF “NRT”
project, ‘03-07) – joint with Columbia, Princeton, Lucent , IBM, Thomson
MIMO networks/DAPHNE (NSF grant, ‘03-06) – joint with Princeton & NJIT
Cognitive Radio hardware & algorithms (NSF NeTS grants, ’04-’07) – joint with
GA Tech and Bell Labs
Privacy and security in sensor nets (NSF NeTS grant, ’04-’07)
Industry supported focus projects
Security in next-generation wireless networks (NICT, Japan ’02-’06)
 MIMO Infostations Prototype for Army (Mayflower/ARL, ’04-05)
 ORBIT Tech Transfer (Intel, DoD, ’05-’06)

14
Status Update: Federal Proposals
Several new proposals submitted or under development for NSF
ITR, NSF NeTS and DARPA, including
Software API & sockets for sensor nets – NSF NeTS NOSS
Spectrum measurements – NSF NeTS ProWIN
Collaborative radio teams (ACERT) –DARPA
Internet spectrum server – NSF NeTS ProWIN
Ad-hoc emergency response networks – DHS (with Columbia U)

Started work on future Internet planning project for NSF – involves
over 20 key networking researchers from various universities and
research labs

Starting work on “wireless ecosystems” ERC focusing on migration
from centralized to distributed systems. Major effort planned for Fall
05 leading to NSF proposal in Nov
15
Status Update: NJ State Projects
NJ State funding for R&D going through major changes:
Emphasizing tech transfer and jobs rather than basic research
MUSE (sensor on silicon) project year 3 funded at 50% level, but
center of excellence program being phased out by NJCST
Tech Center II now in a state “enterprise zone” and thus qualifies
for special programs for incubation and technology transfer
support from NJ EDA
Working on a proposal for a “wireless technology center of NJ”
that would develop technology cores, transfer WINLAB results
and provide specialized services to companies/ventures
Opportunities for co-location of joint venture or wireless activity
at EDA Tech Center Facility
16
Research Highlights
17
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
18
Wireless Architecture: Cognitive Radio Based
Adaptive Networks

Cognitive radio drives consideration of adaptive wireless networks
involving multi-hop collaboration between radio nodes

Needs Internet support similar to ad-hoc network discussed earlier
 Rapid changes in network topology, PHY bit-rate, etc.  implications for routing
 Fundamentally cross-layer approach – need to consider wired net boundary
 High-power cognitive radios may themselves serve as Internet routers…
INTERNET
Bootstrapped PHY &
control link
C
B
B
DD
E
A
A
End-to-end routed path
From A to F
F
19
Cognitive Radio: Hardware Platforms


Next-generation software-defined radio supporting fast spectrum
scanning, adaptive control of modulation waveforms and
collaborative network processing
Facilitates efficient unlicensed band coordination and multi-standard
compatibility between radio devices
Megarray
Connector244 Configurable
I/O pins
XC2V6000
FPGA
TMS320C6701
100BaseT Ethernet
MPC8260
Bell Laboratories Software Defined Radio (Baseband Processor)
Courtesy of Dr. T. Sizer
20
Cognitive Radio: Hardware Platform
Requirements include:

Agile radio
I/O
Software defined modem
Network Processor



Clock Mgmt
Wakeup
Packet
Buffer
DRAM)
A/D
radio
D/A
A/D
radio
radio

~Ghz spectrum scanning,
- Etiquette policy processing
- PHY layer adaptation (per pkt)
- Ad-hoc network discovery
- Multi-hop routing ~100 Mbps+
D/A
Baseband
FPGA
Baseband
Processor Core
(DSP)
Packet
FPGA
A/D
D/A
SRAM
Host
(CR Strategies)
Local ethernet drop
WINLAB’s “network
centric”
for Implementation
cognitive radio prototype
Figure
3. Prototypeconcept
Baseband/MAC
(..under development in collaboration with GA Tech & Lucent Bell Labs)
21
Ad-Hoc Network: Discovery Protocol

Creates efficient ad-hoc network topology just above MAC layer in
order to reduce burden on routing protocol…
Internet
AP
coverage
area
AP
AP
Access Point (AP)
Self-organized
ad-hoc network
Forwarding
Node (FN)
Low-tier access links
(AP/FN Beacons, MN
Associations, Data)
FN
FN
MN
MN
FN
coverage
area
MN
Node
ID
Ad-hoc infrastructure
links between FNs and
APs
(AP/FN Beacons, FN
Associations, Routing
Exchanges, Data)
FN
Low-tier
(e.g. sensor)
Mobile Node (MN)
Broadcast
MAC
MN
MN
MN
Source
MAC
MN
Packet
Type
Beacon Frame Format
Cluster
ID
Sequence
Number
AP
MN
FN
FN
•Scan all channels
•Find minimum delay links to AP
•Set up routes to AP
•Send beacons
Assoc
•Forward SN data
Channel 2
Transmit Power
Hops
Transmit Required: 1mW
Node
To
Type
AP
Power
Beacon
Beacon
Channel 4
Transmit Power
Required: 4mW
SN
•Scan all channels
•Associate with FN/AP
•Send data
22
Ad-Hoc Networks : “SOHAN” Results
Flat
System Parameters:
0.9 sq. km,
Mapping on to ORBIT
20 mobiles/sensors,
Radio grid emulator
4 FNs,
2 APs
802.11a with multiple freqs
AP
FN
MN
Total System Throughput for flat and hierarchical topologies
Hierarchical
50
45
Flat
Hierarchical
• “SOHAN” system evaluated for urban
mesh deployment scenario with ~25 nodes
40
System Throughput (Mbps)
Hierarchical
35
30
• Results show that system scales well
and significantly outperforms flat ad-hoc
routing (AODV)
25
20
15
10
15
20
25
30
35
40
45
50
System offered load (Mbps)
55
60
65
Flat
23
Ad-Hoc MAC: D-LSMA Scheduling
A
B
C
E
RTS
CTS
DATA
D
……
Classified
flows
Scheduler
Upper MAC
D-LSMA
Lower MAC
A
B
C
D
to C
RTS retransmit
E
to C
to E
to C
to E
to C
to E
T
t0


t1
t2
Link scheduling to allow parallel transmissions, solves “exposed node”  useful for QoS on
ad-hoc FN-FN infrastructure in hierarchical systems
Distributed scheduling algorithm (upper MAC), using 802.11-based lower MAC
24
Wireless Architecture: Sensor Nets and
Pervasive Systems
Compute & Storage
Servers
Pervasive
Application
Agents
User interfaces for
information & control
Mobile Internet (IP-based)
Overlay Pervasive Network Services
3G/4G
BTS
Sensor net/IP gateway
GW
Relay Node
Ad-Hoc Sensor Net A
Sensor/
Actuator
Ad-Hoc Sensor Net B
Virtualized Physical World
Object or Event
25
Pervasive Systems: Key Technologies
Application
Agents
IP Network
Content-Based Routing
Content
Router
Caching, Dynamic Binding
IP Routing
Application
Server
IP Network Gateway
Application
Ad-Hoc Net Protocols
Caching, Dynamic Binding
Wireless Access Point
Content-Based Routing
Ad-Hoc Net Protocols
Infostation
(wireless cache)
Radio Forwarding Node
Application
Caching, Dynamic Binding
Adaptive CR Net Protocols
Content-Based Routing
PHY Adaptation
Ad-Hoc Net Protocols
CR Software Platform
TinyOS
Wireless Sensors
Future Cognitive Radio
26
Sensor Hardware: Multimodal ZnO
device

“Tunable” ZnO sensor prototype
developed:



2DEG Sensing device with chemically
mesa
selective receptor coating
Can be “reset” to increase sensitivity, e.g. in
liquids or gas
Dual mode (acoustic and UV optic)
Applicable to variety of sensing needs
Mixer
Sensor
output
REF.
SAW
IDT
Gate voltage
input
2DEG
2DEG
Ground
mesa
Courtesy of: Prof Y. Lu,
Rutgers U
27
Pervasive Applications: Highway Safety

Sensors in roadway interact with sensor/actuator in cars

Opportunistic, attribute-based binding of sensors and cars
 Ad-hoc network with dynamically changing topology
 Closed-loop operation with tight real-time and reliability constraints
28
Pervasive Systems: Software Model

Sensor net scenarios require a fundamentally new software
model (…not TCP/IP or web!!):




Large number of context-dependent sources/sensors with unknown IP address
Content-driven networking (…not like TCP/IP client-server!)
Distributed, collaborative computing between “sensor clusters”
Varying wireless connectivity and resource levels
Pervasive Computing Application
Agent 2
Agent 3
Agent 1
Sensor Net
Software
Model
Overlay Network for
Dynamic Agent <-> Sensor
Association
Sensor
Cluster B
Sensor
Cluster A
Resource
Discovery
Ad-hoc
Routing
Run-time
Environment
(network OS)
OS/Process
Scheduling
29
ORBIT Testbed: Radio Grid
64-node radio grid prototype at Busch Campus (8/04)
400-node radio grid system at Tech Center II
(under construction 5/05)
30
ORBIT: Field Trial System
Lucent “Base Station Router”
with IP interface
“Open API” 802.11a,b,g
ORBIT radio node
31
Web Sites for More Information:
 WINLAB:
www.winlab.rutgers.edu
 ORBIT: www.orbit-lab.org
32