Date post: | 21-Dec-2015 |
Category: |
Documents |
View: | 217 times |
Download: | 1 times |
XG Dynamic Spectrum XG Dynamic Spectrum Experiments, Findings Experiments, Findings
and Plansand PlansPanelPanel
DoD Spectrum Summit 2006DoD Spectrum Summit 2006
Preston Marshall, DARPAPreston Marshall, DARPATodd Martin, STA Todd Martin, STA
Mark McHenry, Shared SpectrumMark McHenry, Shared SpectrumPaul Kolodzy, Kolodzy Consulting Paul Kolodzy, Kolodzy Consulting
2Distribution Statement A – Approved for Public Release – Distribution Unlimited
Panel Structure
Quick Introduction to DARPA XG
Preston Marshall
Legacy Radio Operation and Interference Measurement
Todd Martin
Demonstration Scenario and Results
Mark McHenry
Incremental Transition Paul Kolodzy
Next Steps
Questions and Discussion
Preston Marshall
All
3Distribution Statement A – Approved for Public Release – Distribution Unlimited
All Spectrum May Be Assigned, But…
…Most Spectrum Is Unused!
XG is Developing the Technology and System Concepts for DoD to Dynamically Access All Available
Spectrum
React
Formulate Best Course of Action
ReactReact
Formulate Best Formulate Best Course of ActionCourse of Action
Adapt
Transition network to new emission plan
AdaptAdapt
Transition Transition network to new network to new emission plan emission plan
Characterize
Rapid waveform determination
CharacterizeCharacterize
Rapid waveform Rapid waveform determinationdetermination
Sense
Real time, Low-power, wideband
monitoring
SenseSense
Real time, LowReal time, Low--power, wideband power, wideband
monitoringmonitoring
AutonomousAutonomousDynamic Dynamic SpectrumSpectrumUtilizationUtilization
DARPA XG Program
Goal: Demonstrate Factor of 10 Increase in Spectrum Access
Maximum Amplitudes
Frequency (MHz)
Am
pli
du
e (
dB
m)
Heavy UseHeavy Use
Sparse UseSparse Use
Heavy UseHeavy Use
Medium UseMedium Use
4Distribution Statement A – Approved for Public Release – Distribution Unlimited
Maximum Amplitudes
Frequency (MHz)
Am
pli
du
e (d
Bm
)
Spectrum Allocation & Utilization
Static Spectrum Management is Limited in Its Ability to Improve Spectrum Utilization Efficiencies
Static Spectrum Management is Limited in Its Ability to Improve Spectrum Utilization Efficiencies
• Fixed Spectrum Assignments Lead to Inefficient Spectrum Utilization• Opportunities Exist in Time, Opportunities Exist in Time,
Frequency, and GeographyFrequency, and Geography• Constrain Network AdaptationConstrain Network Adaptation• RF Spectrum Allocated by PolicyRF Spectrum Allocated by Policy
– Allocations, Assignments, Allocations, Assignments, and Incumbents Vary by and Incumbents Vary by CountryCountry
Heavy Use
Sparse Use
•Observations Show Bands of Local Heavy and Sparse Activity Temporal Usage Characteristics
Vary by Band & Service Potential for Usage Dependent on
Incumbent Service & Equipment
Heavy Use
Medium Use
Less than 6% OccupancyLess than 6% Occupancy
Today’s Networks are Heavily Constrained by Lack of Spectrum
Today’s Networks are Heavily Constrained by Lack of Spectrum
5Distribution Statement A – Approved for Public Release – Distribution Unlimited
Transceiver
SystemStrategy
Reasoner
XG Operation
Select
Opportunities
PolicyReasoner
DevelopOptions Process
Request
Determine
Opportunities
Yes/No or Additional
Constraints
AccreditedPolicy
RF Info Acquisition
Sensing Loop
Policy Engine
RadioPlatform
Me
ss
age
Flo
w
RF ResourceRequest
RF Transmit Plan
6Distribution Statement A – Approved for Public Release – Distribution Unlimited
DARPA XG Program Investments
XG PrototypeXG Prototype& Demonstration& Demonstration
Spectrum AwarenessSpectrum Awareness
InterferenceInterferenceEffectsEffectsAssessmentAssessment
BehaviorBehavior
AdaptiveAdaptiveNetworkNetworkOperationOperation
Spectrum Measurements
Sensor Technology Signal Processing Algorithms
Distributed SensingAlgorithms
IEEE 1900IEEE 1900PolicyDescription
Spectrum AdaptiveNetworking
OptimizingStrategies
PolicyReasoning
Capabilityand Affordability
Dynamics
SubnoiseDetection
IncreasedAwarenes
s
PerformanceExperiments
Methodology
Framework &Semantics
EngineeringBasis
Non-InterferingOperation
Tactics
PolicyLanguage
EnforcementImplementation
Assessments
7Distribution Statement A – Approved for Public Release – Distribution Unlimited
Progress In Dynamic Spectrum
• Successful XG Testing Performed at Fort A. P. Hill in August 2006
• Demonstrated the Three Core Principles of XG Program
1. Does Not Interfere with Other Spectrum Users2. Functions in Setting Up and Maintaining Networks3. Creates Capacity Where Spectrum Was Not
Available• Focus in this Demo was Demonstrating non-
Interference to Legacy Systems– Military Network Radios, Military and Civil Voice
• Attendance by DoD (OSD, COCOMs & Services) and Civil Regulatory (FCC and NTIA) Communities
8Distribution Statement A – Approved for Public Release – Distribution Unlimited
NII XG Phase 3bMetric Objectives
• No Harm: Causes no Harmful Interference to Non-XG Systems– Abandon Time: Abandon a Frequency ≤ 500 ms– Interference Limitation: Maintain ≤ 3dB of SNR at a Protected
Receiver.• XG Works: Forms and Maintains Dynamic Connectivity
– Network Formation/Rendezvous Time: Establish XG Network of Six (6) Nodes in ≤ 30 sec.
– Net Join Time: Join a Node to an Existing XG Network in ≤ 5 sec.– Channel Re-Establishment Time: Reestablish XG Network of Six (6)
Nodes ≤ 500 ms• XG Adds Value: Reduces Spectrum Management Setup Time
(Increases Deployment Flexibility) and Increases Spectrum Access (Communications Capacity)– No Pre-assigned Frequencies– ≥ 60% Spectrum Occupancy with XG Network of Six (6) Nodes
Metrics Defined by DARPA & OSD/NII as Threshold for Establishing Early Confidence in Viability of Dynamic Spectrum Access Technologies
Metrics Defined by DARPA & OSD/NII as Threshold for Establishing Early Confidence in Viability of Dynamic Spectrum Access Technologies
9Distribution Statement A – Approved for Public Release – Distribution Unlimited
Legacy Radio Operation and Interference
Measurement
Todd MartinXG Test Director
10Distribution Statement A – Approved for Public Release – Distribution Unlimited
Legacy Radios
Legacy DoD Radio/Test Equipment
PSC-5
Legacy Radios• PRC-117: Frequency Hopping to Force
Dynamics• PSC-5: Narrowband Voice• EPLRS: DoD Networking Radio• Micro-Lite: DoD Networking Radio• ICOM F561: Widely Used in Public Safety
XG Radios(mobile)
Legacy DoD Radio (fixed)
MicrolightPRC-117
11Distribution Statement A – Approved for Public Release – Distribution Unlimited
The XG “Electromagnetic Obstacle Course”
EPLRS
EPLRS Micro-Lite
PRC-117
PSC-5
Night Vision Observation Building
Jammer ICOM
XG drive path
12Distribution Statement A – Approved for Public Release – Distribution Unlimited
Drive Route Spectrum Density
Legacy Node Placement, Terrain, and Propagation Effects Created Dense and Dynamic RF Environment
Legacy Node Placement, Terrain, and Propagation Effects Created Dense and Dynamic RF Environment
• Demo Environment Created Artificially High Spectrum Density to Stress XG
• Some Regions Would Have No Spectrum Available for Multiple XG Nets
• Typical Tactical Density Less Than 6%
13Distribution Statement A – Approved for Public Release – Distribution Unlimited
Legacy Radio Performance Telemetry
4 Legacy Radio 4 Legacy Radio Nets PerformanceNets Performance
PerfectPerfect
InterferenceInterference
• JSC or JHU APL Personnel Oversaw each Radio PairJSC or JHU APL Personnel Oversaw each Radio Pair• Real Time Telemetry of Each Legacy Radio (except for Real Time Telemetry of Each Legacy Radio (except for
Unicom PTT Public Safety)Unicom PTT Public Safety)• Reported Bit Error Rate or Packet Delivery, Depending on Reported Bit Error Rate or Packet Delivery, Depending on
Radio TypeRadio Type• PSC-117 Also Reported Hop FrequencyPSC-117 Also Reported Hop Frequency
14Distribution Statement A – Approved for Public Release – Distribution Unlimited
XG Demonstration and Quantitative Results
Mark McHenryXG Principle Investigator
15Distribution Statement A – Approved for Public Release – Distribution Unlimited
XG Radio System
GPP with 802.16 modem
GPP with XG algorithms
RF Enclosure
Display showing XG operational stateRockwell Sensor RF Power Amp
225-600 MHz RF Transceiver (located under shelf)
16Distribution Statement A – Approved for Public Release – Distribution Unlimited
Components
• Trusted dynamic policy control architecture• Scheduler (detector to share the RF chain)• Group sensing (use distributed measurements made
by individual nodes and fused across a collection of nodes)
17Distribution Statement A – Approved for Public Release – Distribution Unlimited
Test scenario
18Distribution Statement A – Approved for Public Release – Distribution Unlimited
objective
• No harm ( avoid interference)• Works (forms and maintains connected nw)• Adds value (spectrum efficiency)
19Distribution Statement A – Approved for Public Release – Distribution Unlimited
No harm
• Channel abandonment time <500ms– Time between one of the nodes detect a non-cooperative
signal and the time the XG radios ceases transmission– Sends an alert to nw members and renegotiate band– Challenges:
• Propagation loss causes some packets to be lost• XG nodes have different observation of spectrum availability• RF chain differences (detection sensitivities, WiMAX receiver
sensitivities, power level)
• Interference to Noise Ratio– Interference at primary receivers– Metric: whether a XG radio can cease transmission in the
presence of (weak) primary signals– Note: there are assumptions on relative locations of PM and
XG.
20Distribution Statement A – Approved for Public Release – Distribution Unlimited
XG works
• Network join time– Detect the signal signature of the XG system, (sweeping
existing bands)– Declare its presence– Handshake and join the existing nw.
• NW re-establish time– Time to reestablish a channel after abandoning its existing
channel– ? Use rendezvous channels? – Allow existing users to empty queues
• No pre-assigned frequencies– No infrastructure, no dedicated control channel for network
startup
• Link uptime
21Distribution Statement A – Approved for Public Release – Distribution Unlimited
Adds value
• White space fill factor– Time factor.– How about spatial factor?
• Success in channel use– Limiting factors:
• No channel available (too many secondary users)• Cannot find channel (sensing failure)• Link not usuable because of propagation loss
22Distribution Statement A – Approved for Public Release – Distribution Unlimited
Discussions
• Various metrics defined in this study --- a set of reasonable expectations at the current stage
• Simple schemes designed to achieve the target• Research opportunities:
– Other metrics? – Better protocols to achieve a set of targets
• how to set a startup protocol with no prior info. on other users’ locations, cell size, frequency, intf., etc.
• ESCAPE protocol for channel evacuation – Performance analysis (both protocol-wise and fundamental)
• given the required success in channel use, how many users can N channels support? Can a centralized protocol achieve it? How about a (simple) robust distributed one?
• INR analysis w.r.t. locations• Given collision probability, the ultimate capacity bound (Senhua’s work)
– Tradeoffs among a set of targets• Tradeoff between utilization (fill factor) and QoS (success in channel use)• Aggressive in sensing (sensing threshold) vs. INR (intf. to noise ratio)
– Application-aware/context-aware schemes• A VoIP may choose a band that is narrower, but more reliable (e.g.,
occupancy factor)• A large file download may choose a high-bandwidth high risk channel. • How to quantify such heuristics?
23Distribution Statement A – Approved for Public Release – Distribution Unlimited
Phase 3 Metrics & Results Summary
Metric Threshold ResultsXG Causes No Harm
Abandon Time 500 msec 100% in 465 msec
Interference Limit 3 dB Mean: 0.16dB, Max: 0.49dB
XG WorksNet Formation 30 sec w/ 6 Nodes 90%: 3.6s; 100%: 8.68s
Net Join 5 sec 90%: 2.07s; 100%: 4.36s
Net Re-Establish 500 msec 100%: 165msec
XG Adds ValueSpectrum Occupancy 60% w/ 6 Nodes 85% Occupancy at 83%
Confidence
XG Demonstrated Reliable Networking Without Harming Legacy Nodes In Dense Spectrum Environments
XG Demonstrated Reliable Networking Without Harming Legacy Nodes In Dense Spectrum Environments
24Distribution Statement A – Approved for Public Release – Distribution Unlimited
XG Demo Results Instant Replay
4 Legacy Radio 4 Legacy Radio Nets PerformanceNets Performance
Three XG Radio Nets PerformanceThree XG Radio Nets Performance
PerfectPerfect
InterferenceInterference
Circles Are
Legacy Radio Nets
Dots Are XG Radio Nets
Colors Indicate Current Frequency (ex. EPLRS is Orange)
PLAY
25Distribution Statement A – Approved for Public Release – Distribution Unlimited
Phase 3 Metrics & Results Summary
Metric Threshold ResultsXG Causes No Harm
Abandon Time 500 msec 100% in 465 msec
Interference Limit 3 dB Mean: 0.16dB, Max: 0.49dB
XG WorksNet Formation 30 sec w/ 6 Nodes 90%: 3.6s; 100%: 8.68s
Net Join 5 sec 90%: 2.07s; 100%: 4.36s
Net Re-Establish 500 msec 100%: 165msec
XG Adds ValueSpectrum Occupancy 60% w/ 6 Nodes 85% Occupancy at 83%
Confidence
XG Demonstrated Reliable Networking Without Harming Legacy Nodes In Dense Spectrum Environments
XG Demonstrated Reliable Networking Without Harming Legacy Nodes In Dense Spectrum Environments
26Distribution Statement A – Approved for Public Release – Distribution Unlimited
No Harm: XG INR at Non-XG Radios
XG Produced Marginal INR at Non-XG Radios in All CasesXG Produced Marginal INR at Non-XG Radios in All Cases
INR Histogram and CDF
0%
2%
4%
6%
8%
10%
12%
14%
-0.1
6-0
.12
-0.0
9-0
.06
-0.0
30.
010.
040.
070.
100.
140.
170.
200.
230.
270.
300.
330.
360.
400.
430.
460.
49
Measured INR
Occ
urr
ence
Pro
bab
ility
(%
)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Cu
mu
lati
ve P
rob
abili
ty (
%)
Measured Data
Cumulative Probability
Mean Value < 0.1 dB !
27Distribution Statement A – Approved for Public Release – Distribution Unlimited
Channel Re-Establishment Time
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 20 40 60 80 100 120 140 160 180
Channel Re-Establishment Time (ms)
Cu
mu
lati
ve P
rob
abili
ty (
%)
2 of 6 Nodes
3 of 6 Nodes
4 of 6 Nodes
5 of 6 Nodes
6 of 6 Nodes
XG Works: Re-Establish Time
XG Re-Established Networks in < 500 msec. in All CasesXG Re-Established Networks in < 500 msec. in All Cases
28Distribution Statement A – Approved for Public Release – Distribution Unlimited
XG Spectrum Access vs. Network Connectivity- Phase 2 Predicted and Phase 3 Demonstrated -
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0% 20% 40% 60% 80% 100%
% XG Bandwidth Utilization
% X
G C
on
necti
vit
y
Adds Value: Spectrum Occupancy
XG Achieved > 60% Spectrum Occupancy for Networks of 6 Nodes: 85% Access Confidence at 83% Occupancy
XG Achieved > 60% Spectrum Occupancy for Networks of 6 Nodes: 85% Access Confidence at 83% Occupancy
Phase 2 Simulations
Phase 3 Field Data
4 Node NW
6 Node NW 0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0% 20% 40% 60% 80% 100%
25-500 kHz Mix
50 kHz - 1 MHz Mix
100 kHz - 2 MHz Mix
150 kHz - 3 MHz Mix
200 kHz - 4 MHz Mix
300 kHz - 6 MHz Mix
Measured Tactical Occupancy
29Distribution Statement A – Approved for Public Release – Distribution Unlimited
XG Transition Strategies and Regulatory Needs
Paul KolodzyEx-DARPA PM and FCC Spectrum
Policy Task Force
30Distribution Statement A – Approved for Public Release – Distribution Unlimited
• Anticipate Incremental Adoption on a Not to Interfere Basis (NIB)– Military on Military (10x Greater Packing of Radios)– Coordinated Sharing (Military
with Coordinated Users)– Opportunistic (Widespread
NIB Operation)
• Incremental Rollout Enables Near-Term Deployment as Appliqué Into Existing Systems– Add Protocols and Adaptation
Software to Digital Networking Radios
– Add Spectrum Sensing Algorithms
XG Program – Transition
Not Necessary to Establish New RegulatoryNot Necessary to Establish New RegulatoryFramework, Either Nationally or InternationallyFramework, Either Nationally or Internationally
31Distribution Statement A – Approved for Public Release – Distribution Unlimited
Next Steps
Preston MarshallXG Program Manager
32Distribution Statement A – Approved for Public Release – Distribution Unlimited
XG Program Next Steps• What’s Done
– Demonstrated Ability to Avoid Interference to Other Radios• 225-600 MHz
– Developed Waveform for Spectrum Agility using Dynamic PHY and Wideband Sensing Integrated into MAC
– Validated Core Components of Spectrum Access Logic and Algorithm Needs• What’s Left to Do
– Integrate XG into Network Technology• Enable Variable Network Topologies• Establish Load Balancing to Assure High Confidence
– Develop and Demonstrate Scalability• Increased DoD Radio Applications up to 2.5 GHz• Greater XG-XG Network Size, Density, and Interaction
– Address Broader Class of Signals• Sub-noise Detection and Wideband Signals• Data Fusion for False Alarm and Detection Confidence
– Extend Spectrum Access Logic and Algorithms to Cover the Range and Complexity of DoD Operational Needs
• Early Transition– Investigate Operational Benefits in a Jamming Environment– Investigate Immediate Transition into an Existing Military Network Radio
Phase 3b Investments Provide Cornerstone for Phase 3c Development of Fieldable Technologies
Phase 3b Investments Provide Cornerstone for Phase 3c Development of Fieldable Technologies
33Distribution Statement A – Approved for Public Release – Distribution Unlimited
XYZ.COM
Future Wireless Program is Attacking All Aspects of Networking
MigratingMigratingCentralCentralServersServers
DistributingCache
Servers
UnknownTopology
No Frequency No Frequency & Network & Network PlanningPlanning
Enter SearchDistributed Distributed
Index Index ServicesServices
No Fiber &Wires
No Cell Towers FiniteFinite
EnergyEnergy
34Distribution Statement A – Approved for Public Release – Distribution Unlimited
WNAN/WANN Adaptive Radio Uses All Network Layers to Resolve Issues
MIMOMIMO
BeamBeamFormingForming NullingNulling
TopologyTopologyPlanningPlanning
SpectrumSpectrumPlanningPlanning
DeviceDeviceSpurs, …Spurs, …
RelocateAround
Spur
SpectrumToo Tight
Re-planAcross
Network
Re-planTopology
UnavoidableStrongSignal
NeedMore Range
Each Technology Can Throw “Tough” Situations to other More Suitable Technologies without Impact on User QOS
No Good MIMO Paths
Network-Wide
Radio Device
Link
Move to New Preselector
BandStrongStrong
NeighborNeighborSignalSignal
DynamicDynamicSpectrumSpectrum
Dynamic Spectrum Key to Adaptive
Networking
35Distribution Statement A – Approved for Public Release – Distribution Unlimited
WANN/WNAN Hardware Platform
• Single RF Processing Slice Replicated to form 1, 2 and 4 channel MIMO/XG/ Beamforming Capable Radios
• Reverse of Standard STO Approach– Build Early H/W and Incrementally Add Network
Capability– Have Early Demonstrator of DARPA Philosophy
and Technology
• Approach:– Develop early Prototypes By Leveraging
Available Commercial Chips (TV-Tuners and Others)
• Use Cost Pressure to Force Innovation for Lower Cost/Higher Performance
• Contribution from MTO New Analog Signal Processing (MEMs Filter Program) Essential
Frequency 900 MHz to 6 GHz
Power 36 dBm
SFDR 60 dB
IP3 What it is!
Peak 10 Mbps
Control- Based MANET
NewTechnology
NewTechnology
DynamicSpectrum
(XG)
MIMO(MnM)
COTS Chip Set
$ 500 per 4 Channel Node, Spectrally Adaptive, MIMO, Beamforming, Member of Four Simultaneous Subnetworks, Ultra Low Latency
36Distribution Statement A – Approved for Public Release – Distribution Unlimited
WNAN Networks Achieve Reliability through Diverse Paths and Frequencies
Mesh or MANET WNAN
• Limited Scalability• Bandwidth Constrained by Mutual
Interference – More Nodes do Not Create More Capacity
• Low Reliability Due to Single Link Routes
• Large Number of Nodes on Single Frequencies
• Unconstrained Scalability• Additional Nodes Create Additional
Networks – Enabling More Capacity • Diversity in Frequency Avoids
Interference• Multiple Routes Reduce Link
Dependency• Dynamic Spectrum Can Use Network
to “Make Before Break”
Color Depicts all radios on the same frequency Color Depicts sub-net
frequencies -- MIMO Mode Not Depicted