DTIC ADA535691: Testing of Policy-Based Dynamic Spectrum Access Radios

0.27 MB·English

by  Defense Technical Information Center

The 2010 Military Communications Conference - Unclassified Program - Waveforms and Signal Processing Track
Testing of Policy-Based Dynamic Spectrum Access
Radios
Jeffrey Boksiner, Kauteng Chang, Yuriy Posherstnik Mike Totaro Trang Mai, Joseph Molnar
Paul Costello, Richard Huck, Sabre Systems Booz Allen Hamilton Naval Research Laboratory
Tim Leising, Michael Zankel Manasquan, NJ Eatontown, NJ Washington, DC
Communications-Electronics Research
Development and Engineering Center
Fort Monmouth, NJ
Abstract—Policy-drivenDynamicSpectrumAccess(DSA)sys- spectrum shortage in the military and civilian environments.
tems are emerging as one of the key technologies to enable However, one of the key issues surrounding deployment and
the Department of Defense (DoD) to meet its increasing re-
continuing development of DSA systems concerns the need
quirements for access to the electromagnetic spectrum. A key
to test and evaluate the performance of DSA in avoiding
openissuesurroundingdeploymentandcontinuingdevelopment
of DSA systems concerns (1) the need to test and evaluate interference to assigned spectrum users as well as avoiding
the performance of DSA in avoiding interference to itself and interference from assigned users to itself and evaluating the
assigned incumbent users and (2) the performance of DSA performanceofDSAnetworkinthepresenceofvarioustypes
networkinthepresenceofvarioustypesofpotentialinterference.
of potential interference. For example, if an assigned user
In this paper we describe test framework and concepts to
begins to transmit on frequency occupied by a DSA system,
characterize performance of DSA-enabled policy-based radios.
Our test framework includes tests to characterize the inherent the times required to identify the user, abandon the channel,
interference-avoidance characteristics of DSA, such as the time andreformthenetworkonanalternatechannel,areimportant
to abandon a channel, as well as tests that address performance parameterswithrespecttointerferenceavoidanceandnetwork
implications of a particular DSA policy. The test framework
performance.
also provides for the ability to inject a relevant electromagnetic
In this paper, we establish an extensible test framework to
environment(EME).Theproposedframeworkisflexibleallowing
forcustomizationoftherelevanttestconditions,suchastheEME, characterize the performance of DSA-enabled PBR networks
and facilitates simulation of typical communications events such that provides a baseline for characterization and testing of
as network formation and fragmentation. the emerging DSA technology. We note that the National
Index Terms—Dynamic spectrum access, policy based radio, Telecommunications and Information Administration (NTIA)
cognitive radio, PBR, DSA, test plan, framework.
has released a test plan to address spectrum sharing and
coexistence between DSA and legacy systems [8]. While
I. INTRODUCTION
it provides a good baseline for characterization of adjacent
Dynamic Spectrum Access (DSA) is emerging as one of channel interference, sensor desensitization, and several other
the key technologies to enable the Department of Defense metrics,itdoesnotaddressnetworkformation,fragmentation,
(DoD) to meet its increasing requirements for access to the collision, coalescence, and other effects defined in this test
electromagneticspectrum[1],[2].Policy-basedradios(PBRs) plan that are also essential metrics for efficient spectrum
with DSA functionality allow for more efficient utilization of sharing. Continuing DSA developments will lead to a more
the available spectrum in comparison to the traditional spec- complete framework encompassing the NTIA tests as well as
trumreservationparadigm[3].Underthespectrumreservation the currently proposed tests.
paradigm,radiocommunicationsystemswithinaspecifiedarea
II. GENERALAPPROACH
are assigned a static set of frequencies and no other systems
are permitted on those frequencies regardless of their actual A. Objective
occupancy. The use of DSA-enabled systems will enable The goal of developing the DSA test framework is to
multiple devices to operate in the same slice of spectrum address testing of features specifically associated with DSA
by using a spectrum coexistence mechanism [4]–[7]. In an and PBR capabilities. There are already well-established test
opportunistic DSA system, the spectrum coexistence mecha- proceduresforvariousradiocharacteristics,suchasfrequency
nismusuallyinvolvesspectrumsensingtoidentifyunoccupied stability or interference immunity [9], [10]. Our approach
frequencies prior to transmission. Coexistence for PBR can is to leverage these existing standards and augment them
also be enforced through the use of radio policies that can with additional tests to evaluate the performance of DSA
prohibit transmission in certain areas, at certain times, or on specificfeatures.Ourproposedframeworkdealswiththeradio
certain frequencies. frequency(RF)aspectsofDSAaswellastheoverallnetwork
DSA represents a promising approach to alleviate the performance of DSA systems.
U.S. Government work not protected by U.S. copyright 182
Report Documentation Page Form Approved
OMB No. 0704-0188
Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,
including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington
VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it

does not display a currently valid OMB control number.
1. REPORT DATE 3. DATES COVERED
NOV 2010 2. REPORT TYPE 00-00-2010 to 00-00-2010
4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
Testing of Policy-Based Dynamic Spectrum Access Radios
5b. GRANT NUMBER
5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S) 5d. PROJECT NUMBER
5e. TASK NUMBER
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION
Army Communications-Electronics Research Development and REPORT NUMBER
Engineering Center,Fort Monmouth,NJ,07703
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)
11. SPONSOR/MONITOR’S REPORT
NUMBER(S)
12. DISTRIBUTION/AVAILABILITY STATEMENT
Approved for public release; distribution unlimited
13. SUPPLEMENTARY NOTES
IEEE Military Communications Conference (MILCOM) 2010 Proceedings, November 2010
14. ABSTRACT
Policy-driven Dynamic Spectrum Access (DSA) systems are emerging as one of the key technologies to
enable the Department of Defense (DoD) to meet its increasing requirements for access to the
electromagnetic spectrum. A key open issue surrounding deployment and continuing development of DSA
systems concerns (1) the need to test and evaluate the performance of DSA in avoiding interference to itself
and assigned incumbent users and (2) the performance of DSA network in the presence of various types of
potential interference. In this paper we describe test framework and concepts to characterize performance
of DSA-enabled policy-based radios. Our test framework includes tests to characterize the inherent
interference-avoidance characteristics of DSA, such as the time to abandon a channel, as well as tests that
address performance implications of a particular DSA policy. The test framework also provides for the
ability to inject a relevant electromagnetic environment (EME). The proposed framework is flexible
allowing for customization of the relevant test conditions, such as the EME and facilitates simulation of
typical communications events such as network formation and fragmentation.
15. SUBJECT TERMS
16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF
ABSTRACT OF PAGES RESPONSIBLE PERSON
a. REPORT b. ABSTRACT c. THIS PAGE Same as 6
unclassified unclassified unclassified Report (SAR)
Standard Form 298 (Rev. 8-98)
Prescribed by ANSI Std Z39-18
While many of the DSA Network Performance tests in the
EME
frameworkassumeamaster-slaveconfiguration,thesametests Simulator
and procedures apply to various architectures. For example, Interference
ad-hoc networks typically have a single control node that Simulation
for test purposes would be the equivalent to a master node.
Likewise,forapeer-peernetwork,anarbitrarynodewouldbe
Coupling Signal
considered the master node. In all cases, the DSA Network Network Recorder
Performance metrics described will be relevant.
B. Test Procedures
The main purpose of DSA is to deal gracefully with the ator ator ator
u u u
perturbations in the electromagnetic environment (EME). Our n n n
e e e
test framework is concerned with the following types of Att Att Att
perturbations:
• Appearance of interferer or assigned user on a frequency
currently occupied by the DSA PBR PBR1 PBR2 PBRN
• Change in position, time, or another environmental pa-
rameter that changes the policy currently in force
• Increase in propagation (path) loss between two groups Aux Aux Aux
of network nodes
• Convergence of network nodes that have been previously
Fig.1. SingleDSAnetworktestconfiguration.
separated due to poor path loss between them.
The first two items listed are generic for any DSA PBR.
The last two items apply specifically to DSA Mobile Ad- required for the tests, although development of scalability
Hoc Networks (MANETs), which are an important subclass tests will require a greater number of nodes. In the following
of DSA systems. To reproduce these effects in the laboratory, test descriptions, we assume that the network uses a master-
we must force adaption of the DSA network in response to a slave configuration. The Master node, which may also be
forced change in EME or policy variable. Thus, we need to referred to as the Base Station node or head node, is the
perform the following: hub of the network. The slave nodes are the spokes of the
network. All network traffic is routed through the master to
1) Configure the DSA system or network
the intended recipients. While the master-slave configuration
2) Load relevant policies
is not required to use the test framework, it is important to
3) Simulate the background EME
identify any privileged or control nodes in advance of the
4) Prepare measuring equipment
test to differentiate interference impacts on these nodes in
5) ApplytheinterferenceorsimulateanotherEMEchange,
comparison to “ordinary” nodes.
such as position change
In addition to the system under test, the essential compo-
6) Measure and record the relevant performance metrics.
nents of the test circuit are the following:
Therefore, a DSA test involves an independent selection of
1) Signal Recorder: The signal recorder must be able
the following elements:
to record the radio-frequency signal in real-time across the
• Configuration and size of the DSA network desired frequency range. It is important to use a device that
• Policy in force during the test recordsthetime-domainsignalcontinuouslyovertherequired
• The simulated EME test period, rather than a scanning device that tunes serially
• Type of interference, e.g. wideband vs. narrowband across the span.
• Network locations affected by the interference 2) EME Simulator: The EME simulator simulates the
• Any other variablesthat can affect policy decisions,such background EME. Sec. III-F describes the various types of
as position, altitude, etc. simulated EME that may be of interest. The EME simulator
may be an arbitrary signal generator or it may be composed
III. FRAMEWORKDETAILS
of one or more actual emitters.
A. Single-Network Setup
3) Interference Simulator: The interference simulator sim-
The specific steps for setting up a DSA network are largely ulates the interference or appearance of the assigned (incum-
dependent on the manufacturer of the system. However, the bent) device. Interference may be injected at various points
generalapproachtothetestshouldbeindependentofthePBR in the network as shown in Fig. 1 using dashed lines. The
type. Fig. 1 illustrates a general test configuration. interference signal is part of the EME and has the same
The setup in Fig. 1 illustrates testing of a generic N- properties as the general EME signal described in Sec. III-F.
node network. For networks capable of multiple nodes, we The chief difference is that the interference is intended to be
deem a four-node network to be the minimum network size thetriggerforchangesinthestateofthePBR,whiletherestof
183
Aux Aux Aux Aux Aux Aux
PBR PBR PBR
WND1 WND2 WNDM
k+1 k+2 N
or or or or or or
at at at EME at at at
u u u u u u
n n n Simulation n n n
e e e e e e
Att Att Att Att Att Att
Attenuator
EME Coupling Signal Coupling Coupling
Simulation Network Recorder Network Network
or or or or or or
at at at at at at Signal
u u u u u u
n n n n n n Recorder
e e e e e e
Att Att Att Att Att Att
PBR1 PBR2 PBRN PBR1 PBR2 PBRk
Aux Aux Aux Aux Aux Aux
Fig. 2. Single DSA network test configuration with a network of legacy Fig. 3. Test configuration to simulate fragmentation or coalescence of a
wirelessnetworkdevices. single N-node DSA network where the network fragments have size k and
N−k.ThisarrangementcanalsosimulatecollisionoftwoDSAnetworks.
EMEprovidesthebackgroundenvironmentwherethechanges
occur. Of course, it is possible that the EME may cause PBR C. DSA Network Performance
state changes subsequent to the initiation of the interference.
The DSA network performance is characterized by aban-
In many cases, the same piece of test equipment can perform
donment time, initial network formation time, network join
the EME simulation and the interference simulation.
time, network migration time, initial network reconnect time,
Animportanttestcaseconcernstheuseofalegacysystem,
final network reconnect time, network fragmentation time,
such as legacy wireless network devices (WNDs), as the
network collision time, network coalescence time, and the
interference simulator. This allows the testing of the spectrum
ability for network roaming. Our test framework allows for
coexistenceofDSAandnon-DSAlegacysystem.Fig.2shows
the characterization in an ideal lab environment, simulated
the test configuration for this situation.
real-world environment, and field test environment. Charac-
4) Auxiliary Equipment (Aux): The auxiliary equipment
terization parameters, such as frequency, bandwidth, frame
provides the network loading and network measurement ca-
length,modulationscheme,EME,powerlevels,etc.,shouldbe
pabilities. It may also include computers, spectrum analyzers
specified within the framework. Fig. 4 illustrates the relevant
and other test equipment, microphones, video cameras, etc.
performance metrics when an interfering signal is introduced.
5) Coupling Network: The coupling network provides the
The following describes these and other performance metrics.
RF connectivity for the PBR network. A functional coupling
1) Abandonment Time (T ): Abandonment time character-
network may be achieved in various ways, and it is up to the a
izes the time required for the DSA network to abandon a
usertodeterminethepropernetworkingcomponents.Thenet-
channel after an assigned spectrum user, or an interference
work may consist of directional couplers, splitters/combiners,
signal, start transmitting on said channel. This is a critical pa-
RF switches, and many other components. The user must
rametersinceanassignedusermustknowthespecificamount
ensure all components of the coupling network meet the
of interference a system will be receiving to decide whether
frequency requirements of the DSA PBRs under test.
to enable DSA in that spectrum. While the abandonment time
B. Multi-Network Setup is a system specification that is guaranteed by design, this
While the setup illustrated in Fig. 1 is useful for tests performanceparametermustbetestedindependentlyfromthe
requiringasingleDSAnetwork,Fig.3illustratesthesetupfor manufacturer.
two or more DSA networks required in such characterizations 2) Base Network Migration Time (T ): Base network mi-
b
asnetworkfragmentation,collision,andcoalescence.InFig.3, grationtimecharacterizesthetimeittakesforthePBRmaster
an adjustable attenuator simulates variable path loss. node to find an unoccupied frequency and start transmitting
184
Tm of continuous, uninterrupted communications among users.
Tf 9) Network Collision: Network collision characterizes the
Tb Ti performance of the DSA network when two separate co-
locatedDSAnetworksoperateononefrequency.Eventhough
ncyf2 1 2 N theuseofDSAshouldpreventthetwonetworksfromcolliding
e
qu MigratedDSA on the same frequency for an extended period of time, it is
Fre Ta networkemissions importanttocharacterizetheDSAnetworkbehaviortoensure
uninterrupted communications among users.
f1
Interferingorlegacy 10) Network Coalescence: Network coalescence charac-
DSAnetwork usersignal
terizes the time required for two separate DSA networks
emissions
approachingonthesamefrequencytocoalesceintoonesingle
Time
network. Even though adjacent networks generally should not
Fig.4. IllustrationofDSAnetworktimingunderinterferencecondition. coalesceonDSA-enabledPBRnetworks,networkcoalescence
may be desired in certain tactical situations. It is especially
importanttocharacterizetheDSAnetworkbehaviortoensure
on a newly selected operating frequency, after abandoning the uninterrupted communications among users during unusual or
previous channel. unintended scenarios.
3) Initial Network Reconnect Time (Ti): Initial network 11) Network Roaming: Network roaming characterizes the
reconnect time characterizes the time required for the first ability of a slave node to roam among several DSA networks.
slave node to reconnect to the master node after base network ThistestisimportanttocharacterizetheDSAnetworkbehav-
migration. ior in a mobile environment.
4) Final Network Reconnect Time (T ): Final network
f
reconnect time characterizes the time required for the last
D. Policy-Based Control
slave node to reconnect to the master node after base network
PBRs require policies to enable use of DSA on commu-
migration.
5) FullNetworkMigrationTime(T ): Fullnetworkmigra- nication networks. Policies dictate channels and/or behaviors
m
which are allowed or disallowed by the system, depending
tiontimecharacterizesthetimeittakesfortheDSAnetworkto
on certain operational or environmental criteria. The goal of
abandonachannel,andhavethemasternodeandallconnected
our test framework for policy performance characterization is
slave nodes form a network and start communicating on a
to test the fidelity of the PBR with respect to executing the
new channel. This is the sum of abandonment time, base
policy. We suggest the following parameters to be evaluated:
network migration time, and final network reconnect time
(T = T +T +T ). The purpose of subdividing the full 1) Dynamic Detector Threshold: The detector threshold
f a b f
migration time into constituent time components is to obtain specifies the levels, in dBm, at which the PBR can identify
abetterunderstandingofthescalingofthefullmigrationtime an assigned user or an interference signal causing the DSA
as a function of the number of nodes, EME, etc. network to abandon an operating channel. It is important
6) Initial Network Formation Time: Network formation to verify the capability of the PBR to dynamically adjust
time characterizes the time it takes one or more DSA-enabled the detector threshold if it is too sensitive or not sensitive
PBRs to connect to the master node, form a network, and enough, depending on the operating RF environment of the
start transmitting data across the network. This parameter is DSA network. This characteristic is unique to systems em-
tomeasuretheformationofanewDSAnetwork,afterafresh ployingenergysensingalgorithms.NotetheDynamicDetector
radio powerup, without any channel abandonment or network Threshold is a test of the capability of the DSA PBR only,
migration. The formation time may vary according to the RF not a test of a mode of operation. If dynamic thresholds are
network type (e.g. WiFi, WiMax, etc.) and network size. employed, a related system policy must also be in place to
7) NetworkJoinTime: Networkjointimecharacterizesthe adjust transmitter parameters to ensure coexistence.
time for one slave node to join a previously formed DSA 2) Power Control: Transmitter power is an important char-
network of two or more nodes. The time will vary depending acteristic that determines if network communication will be
on the parameters of the test (e.g. noise, interference, link successful. This test is used to characterize the effectiveness
attenuation, size of network, etc.). The slave node should not of a policy to dynamically control transmitter power levels of
have prior information regarding the network’s geolocation, the PBR.
cell size, or operating frequency. 3) Selective Frequency Map: A selective frequency map
8) Network Fragmentation Time: Network fragmentation specifies the operating range of the PBR, as well as specific
time characterizes the time for one formed DSA network frequency channels that are prohibited for use by DSA net-
to fragment and form two smaller DSA networks. This is works(asmaybespecifiedbyachannelplanorotherspectrum
an important parameter as it predicts performance of DSA use regulations). This test characterizes the effectiveness of a
network in field situations. Allows for characterization of policy to specify frequency bands on which the DSA network
performance of the DSA network to validate the requirement may and may not transmit.
185
4) GeospatialOperation: Geospatialoperationrequiresthe F. Electromagnetic Environments (EME)
DSA network to be aware of its geospatial positioning. This
Radio frequencies are never used in an environment devoid
awareness may be achieved through built-in or commercially
of other electromagnetic activity. To simulate real-world con-
available GPS simulators, which also allow for the simulation
ditionsofRFenvironmentsandcharacterizetheireffectonthe
ofmovingorvaryinglocations.Regulatoryandsystempolicies
radio systems, electromagnetic signals need to be introduced
maybecreatedthatdefinegeospatialoperationwithinacertain
withpredefinedcharacteristics.Theseincludeadjacentchannel
distance of points, lines, or contours. This test characterizes
narrowband, wideband, wideband noise, frequency sweep,
the effectiveness of a policy to control the operation of the
frequency hop, and high power signals. This portion of the
DSA network based on geospatial awareness.
frameworkallowsforrealistictestinginalabenvironment,and
5) Time-Based Performance: There are instances in which
allows for the creation of a baseline performance metric for
spectrumrestrictionsexistduringcertaindaysofthemonth,or
a DSA-enabled PBR network. Comparison of baseline results
certain times of day. This test characterizes the effectiveness
with the EME-stressed results will provide a good indication
ofapolicytocontroltheoperationoftheDSAnetworkbased
of the performance of DSA in the field.
on parameters of time.
EME conditions are crucial to the operation of DSA PBR
6) Dense/Noisy Environment Performance: With an in-
networks. Simulation of an EME for a testbed should be
creased number of legacy systems being fielded, a decreased
as realistic as possible to accurately measure the success of
amountofspectrumavailable,andactivemilitaryengagements
the DSA technology. Realistic EME may be achieved by
all over the world, intentional and unintentional jamming is
survey, modeling and simulation of a real-world environment
a major concern. This test characterizes the effectiveness of
in the band of interest. Without such a process, real-world
a policy to maintain communication while sustaining heavy
electromagnetic characteristics, including natural and man-
interference or high-power jamming.
made elements, may not be reproduced for testing against
7) Threshold Sensing Accuracy: As described for the Dy-
a realistic environment. The simulated realistic environment
namicDetectorThresholdtestinSecIII-D1,thethresholdisa
could serve as the electromagnetic ambient and other EME
vitalparameterthatenablesefficientuseofDSAonthePBRs.
effects could be added to create different testing conditions.
This test characterizes the sensing accuracy of the detector to
1) Adjacent Channel Narrowband Signal: Adjacent chan-
ensure the PBR is sensing the proper signal levels that are
nel narrowband signal is used to characterize the performance
expected.
of the DSA network in the presence of an adjacent channel
8) Composite Policy Performance: In a real field environ-
narrowband signal. The definition of narrowband depends on
ment,acombinationofanyoftheaforementionedpoliciesmay
the frequency band of interest. For example, a common def-
be running on the PBR at the same time. To ensure uninter-
inition of narrowband signal in UHF/VHF range is occupied
ruptedcommunicationsandoperationofthePBRsasexpected,
bandwidth ≤25 kHz.
this test characterizes the performance of the DSA network
2) WidebandSignal: Widebandsignalisusedtocharacter-
with multiple policies loaded and running simultaneously.
ize the performance of the DSA network in the presence of
9) Cooperative Sensing: Characterizes the ability of mul-
a wideband signal. A wideband signal is defined as a signal
tiple DSA-enabled PBRs to cooperatively detect and commu-
having similar power and spectral density as the DSA radio
nicate the presence of a hidden node. This is an important
signal.
characteristic to solve the hidden node problem in multi-node
3) Wideband Noise-Like Signal: Wideband noise-like sig-
networks.
nal is used to characterize the performance of the DSA
network in the presence of a wideband noise signal. Based
E. General Network Metrics
onourtesting,wedefinewidebandnoisesignalassignalwith
It is important to collect general network performance bandwidth greater than twenty times the occupied bandwidth
measurements to characterize the non-DSA specific metrics of the DSA signal.
of the PBR network in a lab environment. Network metrics 4) FrequencySweptSignal: Frequencysweptsignalisused
are network and application specific, and as such, cannot be to characterize the performance of the DSA network in the
exhaustivelylistedintheframework.Theappropriatenetwork presenceofasweptsignal.Asweptsignalisasignalgenerated
metrics must be determined by the user, based on the specific to sweep the operating band of the PBR. This may also affect
application, type of DSA PBR, network routing protocol, the general networking measurements.
and other parameters. The following items represent a small 5) FrequencyHoppingSignal: Frequencyhoppingsignalis
sampling of the network metrics to be performed as part of
used to characterize the performance of the DSA network in
the DSA test:
the presence of a frequency hopping signal. A frequency hop-
• Payload Throughput pingsignalisgeneratedtotransmitintheoperatingfrequency
• Bit Error Rate band of the PBR. The frequency hop rate should be set to
• Packet Error Rate characterizetheradioindifferentsimulatedEMenvironments.
• Latency Thismayalsobeperformedinconjunctionwiththewideband
• Jitter or narrowband signals.
186
6) High Power Signal: This is used to characterize the distancesseparatingtheDSAnodes.Thischaracterizationwill
performance of the DSA network in the presence of a high also help to establish anomalies associated with the DSA
powersignal.Ahighpowersignalisdefinedasasignalwhose implementation
powerisapre-determinedamounthigherrelativetotheDSA- 3) Unlevel Terrain Performance Characterization: This al-
enabled PBR signal. lows for characterization of the DSA network performance in
various terrains, such as hilly, mountainous, valley, as well
IV. SUGGESTEDFIELDTESTING
as various salinity and sea state conditions. The battlefield
A. Rationale today is occurring on increasing varied terrain, with many
In this paper we are mostly concerned with the DSA test hills, mountains, and valleys serving as great obstacles. The
frameworkforlaboratorytesting.Laboratorytestingisaprac- marine environment has also always posed a challenge due
tical and efficient method to thoroughly exercise all DSA fea- to the variety of ever-changing sea conditions present over
tures. The laboratory allows us to recreate a variety of EMEs, substantial distances. By characterizing the DSA network
interfering signals and network configurations. Furthermore, performance while dealing with varied locations, terrains, and
laboratorytestingfacilitatesmeasurementsofvariousnetwork seaconditions,arealisticcharacterizationofDSAperformance
and DSA metrics. However, a complete characterization of a will be obtained.
DSAPBRrequiresanunderstandingofhowtheradiobehaves
V. CONCLUSION
under realistic field conditions.
The proposed framework provides a baseline for charac-
B. Suggested Field Test Proposals terization and testing of the emerging DSA technology. Key
To gain a full understanding of DSA behavior in the DSA-specific network metrics are identified and quantified
field, the framework suggests repeating the tests provided for characterization of a network of DSA-enabled PBRs. A
in the DSA Network Performance and Policy-based Control primary aspect of this framework is that it is extensible to
sections of the framework, but without the ideally simulated a variety of network configurations. The framework provides
or controlled conditions seen in the lab. The flexibility of methods of identifying the primary performance aspects of
the framework allows for the suggested field tests to be the DSA, devoid of the random effects that are encountered
performed in conjunction with a combination of any of the inafieldenvironment.Furthermoretheframeworkestablishes
characterizations and measurements previously described. By extensionsthatenablefieldexperimentationperformanceeval-
comparing characterization results from Test X performed uations and clearly identifies component properties of the
under ideal conditions, Test X performed in the lab with DSA susceptible to interactions with defined environmental
specificEMEsignalsintroduced,TestXperformedinthefield elements. As DSA technology matures, and a varied number
under ideal conditions, and Text X performed in the field in of DSA-enabled systems become available, the framework
anenvironmentassuggestedbelow,acompleteunderstanding will evolve to incorporate comprehensive and standardized
of the capability of the PBR and the DSA network can be methods for performance comparisons and characterization.
obtained.
REFERENCES
The suggested tests are not an exhaustive list of strictly
[1] K. Zhang, D. Swain, and M. Lin, “Dynamic spectrum access enabled
required tests. These are only suggestions to enable a more
DoD net-centric spectrum management,” in Military Communications
complete characterization of DSA performance. It is up to the Conference,2007.MILCOM2007.IEEE,oct.2007,pp.1–7.
user to decide to perform field tests. However, the utilization [2] S. Chan, “Shared spectrum access for the DoD,” in New Frontiers in
Dynamic Spectrum Access Networks, 2007. DySPAN 2007. 2nd IEEE
oftheproposedframeworkenablestheusertocharacterizethe
InternationalSymposiumon,april2007,pp.524–534.
behavior of any additional tests in a standardized format. [3] “IEEEstandarddefinitionsandconceptsfordynamicspectrumaccess:
1) UrbanEnvironmentPerformanceCharacterization: This Terminologyrelatingtoemergingwirelessnetworks,systemfunctional-
ity, and spectrum management,” IEEE Std 1900.1-2008, pp. c1–48, 26
characterizes the DSA network performance in an urban
2008.
environment. The battlefield today is increasingly urban. By [4] E. Hossain, D. Niyato, and Z. Han, Dynamic Spectrum Access and
characterizing the DSA network performance in an urban ManagementinCognitiveRadioNetworks. Cambridge,UK:Cambridge
UniversityPress,2009.
environment,withbuildingobstructions,indoorsandoutdoors
[5] E. Hossain and V. K. Bhargava, Cognitive Wireless Communication
communications, and very dense RF spectrum utilization (due Networks. NewYorkNY:Springer,2007.
to RF interference or assigned spectrum users, among others), [6] K.ChowdhuryandI.Akyildiz,“Cognitivewirelessmeshnetworkswith
dynamic spectrum access,” Selected Areas in Communications, IEEE
a better understanding of DSA capabilities can be obtained.
Journalon,vol.26,no.1,pp.168–181,jan.2008.
2) ClearLineofSightPerformanceCharacterization: This [7] Q.ZhaoandB.Sadler,“Asurveyofdynamicspectrumaccess,”Signal
characterizes the DSA network’s line of sight performance ProcessingMagazine,IEEE,vol.24,no.3,pp.79–89,May2007.
[8] National Telecommunications and Information Administration. (2009)
on unobstructed, level terrain (or naval or aerial environment)
Final Phase I Test Plan. [Online]. Available: http://www.ntia.doc.gov/
where line of sight is always maintained. This may also be frnotices/2006/spectrumshare/comments.htm/
a good test to confirm the accuracy of the DSA network [9] RadioFrequencySpectrumCharacteristicsMeasurementof,Department
ofDefenseMilitaryStandardMIL-STD-449D,1973.
performance characterization done in the lab, with simu-
[10] Requirements for the Control of Electromagnetic Interference Charac-
lated distance between the DSA nodes by using attenuators, teristicsofSubsystemsandEquipment,DepartmentofDefenseMilitary
compared to the same test done in the field with physical StandardMIL-STD-461F,2007.
187