DTIC ADA283243: Performance Modeling and Analysis of Parallel Processing and Low Earth Orbit Satellite Communication Systems

264 Pages·10.4 MB·English

by  Defense Technical Information Center

Form Approved 10
OATION PAGE
A D- A 283 243 0MB No. 0704-01498
... ge I h~our per re~por~e, including the tine forir~e"," nsrctions. searching exisn dt~oc1
1q the colletinon of information, Send commentsradi n thi br ensetimate or any Othe:r ap fti
I *.IIEE U I to Washington Headcluaneri Services. Directorate for information Operations and Reports. 121 5 1 !ferson
IEEI E I! ,fManagemernt and Budget. Paperworks Reduction Project (0704-0188). Washington. Dc- 20503.
ATE 3R.E PORT TYPE AND DATES COVERED
TIL N SUBTITLE S . FUNDING NUMBERS
6. SPTON(SORN/OIOIGAEC)AES N DRS(S 0 PNOIGMNTRN
DEPARTMENT OF THE AIR FORCE AGENCY REPORT NUMBER
AFIT/CI
2950 P STREET
WRIGHT-PATTERSON AFB OR 45433-7765 X
11. SUPPLEMENTARY NOTES
Approved for Public Release IAW 190-1 2.OSTBUINCD
Distribution Unlimited
MICHEAL M. BRICK.ER, SMSgt, USAF
Cheif Administration
13. ABSTRACT (Maximum 200 words)
S94-25343
14. SUBJECT TERMS 1.NMSOF PAGES
16. PRICE CODE
17. SECURITY CLASSIFICATION I18. SECURITY CLASSIFICATION I19. SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT
OF REPORT I FTHIS PAGE :11 OF ABSTRACT
QuAJ M LV5?.SCMV I
NSN 7540-01-280-5500 Me~
GENERAL INSTRUCTIONS FOR COMPLETING SF 298
1,q The Report Documentation Page (RDP) is used in announcing and cataloging reports. It is important
that this information be consistent with the rest of the report, particularly the cover and title page.
Instructions for filling in each block of the form follow. It is important to stay within the lines to meet
opticals canning requirements.
Block 1. Agency Use Only (Leave blank). Block 12a. Distribution/Availability Statement.
Denotes public availability or limitations. Cite any
Block 2. Report Date. Full publication date availability to the public. Enter additional
including day, month, and year, if available (e.g. 1 limitations or special markings in all capitals (e.g.
Jan 88). Must cite at least the year. NOFORN, REL, ITAR).
Block 3. Type of Report and Dates Covered. DOD See DoDD 5230.24, "Distribution
State whether report is interim, final, etc. If Statements on Technical
applicable, enter inclusive report dates (e.g. 10 ts
Jun 87 -30 Jun 88). DOE See authorities.
Block 4. Title and Subtitle. A title is taken from NASA - See Handbook NHB 2200.2.
the part of the report that provides the most NTIS - Leave blank.
meaningful and complete information. When a
report is prepared in more than one volume, Block 12b. Distribution Code.
repeat the primary title, add volume number, and
include subtitle for the specific volume. On
classified documin entsa reenttheere tsh.De OtitEle classification DOD - Eave DbOlaEn kd.istribution categories
-Enter
in parentheses.
from the Standard Distribution for
Block S. Funding Numbers. To include contract Unclassified Scientific and Technical
and grant numbers; may include program Reports.
element number(s), project number(s), task NASA - Leave blank.
number(s), and work unit number(s). Use the NTIS - Leave blank.
following labels:
C - Contract PR - Project Block 13. Abstract. Include a brief (Maximum
G - Grant TA - Task 200 words) factual summary of the most
PE - Program WU - Work Unit significant information contained in the report.
Element Accession No.
Block 6. Author(s). Name(s) of person(s) Block 14. Subject Terms. Keywords or phrases
responsible for writing the report, performing identifying major subjects in the report.
the research, or credited with the content of the
report. If editor or compiler, this should follow
the name(s). Block 15. Number of Pages. Enter the total
number of pages.
Block 7. Performing Oroanization Name(s) and
Address(es). Self-explanatory. Plock 16. Price Code. Enter appropriate price
Block 8. Performing Oroanization Report code (NTIS only).
Number. Enter the unique appr'umeric re~ort
number(s) assigned by the organization Blocks 17.- 19. Security Classifications. Self-
explanatory. Enter U.S. Security Classification in
Block 9. Soonsorinq/Monitorn,- AceEc.i2,:m, e(s) accordance with U.S. Security Regulations (i.e.,
and Address(es). Self-explanatory. UNCLASSIFIED). if form contains classified
Information, stamp classification on the top and
Block 10. Soonsorin_!Monitorinq Agency bottom of the page.
Report Number. (if known)
Block 11. Supplementary Notes. Enter Block 20. Limitation of Abstract. This block must
information not included elsewhere such as: be completed to assign a limitation to the
Prepared in cooperation with...; Trans. of...; To be abstract. Enter either UL (unlimited) or SAR (same
published in.... When a report is revised, include as report). An entry in this block is necessary if
a statement whether the new report supersedes the abstract is to be limited. If blank, the abstract
or supplements the older report. is assumed to be unlimited.
S.. -A-4 C'.-_(cid:127) Q clop (cid:127) (cid:127)R~i v 2-89)
PERFORMANCE MODELING AND ANALYSIS OF PARALLEL PROCESSING
AND LOW EARTH ORBIT SATELLITE COMMUNICATION SYSTEMS o
by DTIC TAB
Unannounced []
Richard A. Raines Justification
By ... ..
BDist, ibution I
Dr. Nathaniel J. Davis, IV, Chairman Dist,
________
Electrical Engineering Availability Codes
Avail and/ or
Dist Special
(ABSTRACT)
This dissertation presents unique and valuable insight into the analysis of packet-switched
data communication systems. The research described in this dissertation examines performance
characteristics of two types of packet-switcbed data communication systems. The first system to
be analyzed operates in a parallel processing environment where cooperating processors
independently perform assigned tasks. In this environment, the packet delay performance is
dominated by queuing delays. The second type of system examined operates in a low earth orbit
(LEO) satellite communications network environment. In this type of network, delay performance
is affected by both queuing and propagation effects.
The objectives of this research are to study the effects of queuing and propagation on the
average packet delay, the number of buffers required to implement the networks that interconnect
the parallel processors, and the satellite resource utilization rates. For both types of
communication systems, mathematical metamodels [Agr85] are developed to capture the effects on
packet delay caused by incremental changes in network dependent parameters.
Part I of this research performs average packet delay and buffer cost comparisons of the
augmented shuffle exchange network (ASEN) and the multistage cube (MSC) network. It is
shown that the packet delay associated with the ASEN is between 20 and 25 percent lower than
that of a similar sized MSC network. In addition to the delay benefits of the ASEN, network
implementation cost savings for the ASEN are shown to be 9 to 16 percent lower than the MSC.
94 8 11 096
Innovative mathematical design tools are developed and applied to the parallel processing
interconnection network environment. These tools are used for predictive modeling of packet delay
given network dependent parameters. The simple and concise models are shown to have predictive
accuracy within I percent of the observed simulation delay results.
Part II of this research focuses on LEO satellite system communications. Six different
constellations, providing whole-earth coverage are modeled and analyzed. The number of satellites
within the constellations examined range from 36 to 77. The analysis of these LEO satellite
systems consists of examining the packet delay characteristics of these dynamic systems as well
how resource requests to the satellites are distributed. It is shown that when packet delay is the
only design criteria, the differences in delay between the 36-satellite system and the 77-satellite are
minimal and do not warrant the use the 77-satellite system over the 36-satellite system. The
satellite resource utilization analysis captures the resource request load balancing characteristics of
the systems. From a load balancing perspective, the 54-satellite system yields the best performance
while the 36-satellite system the worst. A third unique aspect of the research presented in Part H is
the application of metamodeling to the LEO satellite system environment. The metamodels
developed reduce a complex 8-factor packet delay representation to simple, yet accurate, 2 and 3-
factor relationships. These metamodel delay relationships are shown to have a predicted versus
observed packet delay "best case" accuracy of 8 and 4 percent for the 3 and 2-factor models,
respectively. Predicted versus observed packet delays are typically within 20 percent of agreement.
This research makes two contributions to the state-of-the-art knowledge in packet-switched
communications system analysis. First, the metamodeling of interconnection networks and LEO
systems are first of their kind. These models can allow for reduced simulation trials and more
expeditious design decision making. The second contribution is the development of an integrated
LEO satellite system model not seen in previously published research. This model can be used to
further advance research in the LEO satellite system environment.
PERFORMANCE MODELING AND ANALYSIS OF PARALLEL PROCESSING
AND LOW EARTH ORBIT SATELLITE COMMUNICATION SYSTEMS
by
Richard A. Raines
Dr. Nathaniel J. Davis, IV, Chairman
Electrical Engineering
(ABSTRACT)
This dissertation presents unique and valuable insight into the analysis of packet-switched
data communication systems. The research described in this dissertation examines performance
characteristics of two types of packet-switched data communication systems. The first system to
be analyzed operates in a parallel processing environment where cooperating processors
independently perform assigned tasks. In this environment, the packet delay performance is
dominated by queuing delays. The second type of system examined operates in a low earth orbit
(LEO) satellite communications network environment. In this type of network, delay performance
is affected by both queuing and propagation effects.
The objectives of this research are to study the effects of queuing and propagation on the
average packet delay, the number of buffers required to implement the networks that interconnect
the parallel processors, and the satellite resource utilization rates. For both types of
communication systems, mathematical metamodels [Agr85] are developed to capture the effects on
packet delay caused by incremental changes in network dependent parameters.
Part I of this research performs average packet delay and buffer cost comparisons of the
augmented shuffle exchange network (ASEN) and the multistage cube (MSC) network. It is
shown that the packet delay associated with the ASEN is between 20 and 25 percent lower than
that of a similar sized MSC network. In addition to the delay benefits of the ASEN, network
implementation cost savings for the ASEN are shown to be 9 to 16 percent lower than the MSC.
Innovative mathematical design tools are developed and applied to the parallel processing
interconnection network environment. These tools are used for predictive modeling of packet delay
given network dependent parameters. The simple and concise models are shown to have predictive
accuracy within I percent of the observed simulation delay results.
Part H of this research focuss on LEO satellite system communications. Six different
constellations, providing whole-earth coverage are modeled and analyzed. The number of satellites
within the constellations examined range from 36 to 77. The analysis of these LEO satellite
systems consists of examining the packet delay characteristics of these dynamic systems as well
how resource requests to the satellites are distributed. It is shown that when packet delay is the
only design criteria, the differences in delay between the 36-satellite system and the 77-satellite are
minimal and do not warrant the use the 77-satellite system over the 36-satellite system. The
satellite resource utilization analysis captures the resource request load balancing characteristics of
the systems. From a load balancing perspective, the 54-satellite system yields the best performance
while the 36-satellite system the worst. A third unique aspect of the research presented in Part II is
the application of metamodeling to the LEO satellite system environment. The metamodels
developed reduce a complex 8-factor packet delay representation to simple, yet accurate, 2 and 3-
factor relationships. These metamodel delay relationships are shown to have a predicted versus
observed packet delay "best case" accuracy of 8 and 4 percent for the 3 and 2-factor models,
respectively. Predicted versus observed packet delays are typically within 20 percent of agreement.
This research makes two contributions to the state-of-the-art knowledge in packet-switched
communications system analysis. First, the metamodeling of interconnection networks and LEO
systems are first of their kind. These models can allow for reduced simulation trials and more
expeditious design decision making. The second contribution is the development of an integrated
LEO satellite system model not seen in previously published research. This model can be used to
further advance research in the LEO satellite system environment.
PERFORMANCE MODELING AND ANALYSIS OF PARALLEL PROCESSING
AND LOW EARTH ORBIT SATELLITE COMMUNICATION SYSTEMS
by
Richard A. Raines
Dissertation submitted to the Faculty of the
Virginia Polytechnic Institute and State University
in partial fulfillment of the requirements for the degree of
DOCTOR OF PHILOSOPHY
in
Electrical Engineering
APPROVED:
Dr.1q avsTCann
/Dr. C. W. Bostian Dr. T. Pratt
Dr. S. (cid:127)M. ddif Dr. M. Abr (cid:127)'-s
June 1994
Blacksburg, Virginia
PERFORMANCE MODELING AND ANALYSIS OF PARALLEL PROCESSING
AND LOW EARTH ORBIT SATELLITE COMMUNICATION SYSTEMS
by
Richard A. Raines
Dr. Nathaniel J. Davis, IV, Chairman
Electrical Engineering
(ABSTRACT)
This dissertation presents unique and valuable insight into the analysis of packet-switched
data communication systems. The research described in this dissertation examines performance
characteristics of two types of packet-switched data communication systems. The first system to
be analyzed operates in a parallel processing environment where cooperating processors
independently perform assigned tasks. In this environment, the packet delay performance is
dominated by queuing delays. The second type of system examined operates in a low earth orbit
(LEO) satellite communications network environment. In this type of network, delay performance
is affected by both queuing and propagation effects.
The objectives of this research are to study the effects of queuing and propagation on the
average packet delay, the number of buffers required to implement the networks that interconnect
the parallel processors, and the satellite resource utilization rates. For both types of
communication systems, mathematical metamodels [Agr85] are developed to capture the effects on
packet delay caused by incremental changes in network dependent parameters.
Part I of this research performs average packet delay and buffer cost comparisons of the
augmented shuffle exchange network (ASEN) and the multistage cube (MSC) network. It is
shown that the packet delay associated with the ASEN is between 20 and 25 percent lower than
that of a similar sized MSC network. In addition to the delay bedefits of the ASEN, network
implementation cost savings for the ASEN are shown to be 9 to 16 percent lower than the MSC.
Innovative mathematical design tools are developed and applied to the parallel processing
interconnection network environment. These tools are used for predictive modeling of packet delay
given network dependent parameters. The simple and concise models are shown to have predictive
accuracy within I percent of the observed simulation delay results.
Part II of this research focuses on LEO satellite system communications. Six different
constellations, providing whole-earth coverage are modeled and analyzed. The number of satellites
within the constellations examined range from 36 to 77. The analysis of these LEO satellite
systems consists of examining the packet delay characteristics of these dynamic systems as well
how resource requests to the satellites are distributed. It is shown that when packet delay is the
only design criteria, the differences in delay between the 36-satellite system and the 77-satellite are
minimal and do not warrant the use the 77-satellite system over the 36-satellite system. The
satellite resource utilization analysis captures the resource request load balancing characteristics of
the systems. From a load balancing perspective, the 54-satellite system yields the best performance
while the 36-satellite system the worst. A third unique aspect of the research presented in Part II is
the application of metamodeling to the LEO satellite system environment. The metamodels
developed reduce a complex 8-factor packt delay representation to simple, yet accurate, 2 and 3-
factor relationships. These metamodel delay relationships are shown to have a predicted versus
observed packet delay "best case" accuracy of 8 and 4 percent for the 3 and 2-factor models,
respectively. Predicted versus observed packet delays are typically within 20 percent of agreement.
This research makes two contributions to the state-of-the-art knowledge in packet-switched
communications system analysis. First, the metamodeling of interconnection networks and LEO
systems are first of their kind. These models can allow for reduced simulation trials and more
expeditious design decision making. The second contribution is the development of an integrated
LEO satellite system model not seen in previously published research. This model can be used to
further advance research in the LEO satellite system environment.
ACKNOWLEDGMENTS
This dissertation culminates a three year effort and completion of yet another chapter in my
life. These three years in Virginia have seen a period of physical recovery, spiritual growth, and
academic learning. During this time, many individuals have made direct impacts on my life and
have assisted me in getting to where I am now.
First, I thank Dr. Nathaniel J. Davis, IV for accepting me as his Ph.D. student. His efforts
in guiding this research, enduring the many questions I had, as well as the numerous revisions of
technical papers and this dissertation are greatly appreciated. I also thank you Nat for being more
than an advisor, for being a friend. I value this friendship as well as the working relationship we
have established over the years.
I also thank my committee members, Drs. C. W. Bostian, T. Pratt, S. F. Midkiff, and M.
Abrams for their knowledge and guidance during this effort. Each was always available for my
questions and concerns. Special thanks go to Dr. Tim Pratt for teaching me satellite
communications, setting me straight when needed, and always having the time to answer my many
questions.
I cannot begin to thank my family enough for their love and support during these past three
years. I promised Helen in 1987 that a Masters degree was my last, but broke that promise when
accepting this assignment. Without her love and support, I know I could not have made it. Thank
you Garrett and Kelley for understanding when I could not spend as much time with you as I would
have liked to.
Lastly, and by no means least, I thank the good Lord for the many blessings he has bestowed
upon me and my family. Through Jesus Christ, all things are possible.
iv