본문

서브메뉴

Computer Network Optimization Using the Power Metric
Computer Network Optimization Using the Power Metric
Computer Network Optimization Using the Power Metric

상세정보

자료유형  
 학위논문 서양
최종처리일시  
20250211152836
ISBN  
9798384091707
DDC  
004
저자명  
Tsai, Meng-Jung.
서명/저자  
Computer Network Optimization Using the Power Metric
발행사항  
[Sl] : University of California, Los Angeles, 2024
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2024
형태사항  
241 p
주기사항  
Source: Dissertations Abstracts International, Volume: 86-03, Section: B.
주기사항  
Advisor: Kleinrock, Leonard.
학위논문주기  
Thesis (Ph.D.)--University of California, Los Angeles, 2024.
초록/해제  
요약Modern network research focuses on optimizing performance through congestion control, quality of service, and fairness. With the rapid expansion of networks and increasing traffic, balancing throughput and response time has become critical. This thesis explores this tradeoff and introduces the Power metric as a tool for optimizing network performance and expands its investigation to achieving optimized performance with optimum fairness. The Power metric, defined as the ratio of normalized throughput to normalized mean response time, serves as our performance optimization goal. Previous research primarily focused on single-flow systems, but contemporary networks involve multiple flows with more complex scenarios. This work extends Power analysis in performance optimization to modern network environments, developing a model that also accommodates multiple flows. We further examine different queueing disciplines that implement various levels of flow discrimination. In addition, we examine fairness metrics coupled with performance optimization.Our research focuses on three aspects: performance, flow priority discrimination, and fairness. We introduce performance metrics, including individual power, sum of power, and average power, and optimize these metrics using an M/M/1 system model with multiple flows under different queueing disciplines. We also explore fairness metrics such as throughput, delay, and power, and investigate scenarios where optimum performance and equal fairness can be achieved simultaneously.Additionally, we study generalized power, which allows specifying the relative preference for throughput versus delay, providing a flexible approach to optimizing network performance based on specific requirements.In summary, this research represents a first step in incorporating performance, fairness, and priority flow discrimination into the Power metric analysis for modern multi-flow network environments. Our goal is to provide insights, guidance, and "rules of thumb" for system designers to create more efficient and equitable network systems.
일반주제명  
Computer science
일반주제명  
Computer engineering
일반주제명  
Information technology
일반주제명  
Electrical engineering
키워드  
Congestion control
키워드  
Network optimization
키워드  
Power metric
키워드  
Queueing
키워드  
Scheduling
키워드  
Performance optimization
기타저자  
University of California, Los Angeles Computer Science 0201
기본자료저록  
Dissertations Abstracts International. 86-03B.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008250123s2024        us                              c    eng  d
■001000017164139
■00520250211152836
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798384091707
■035    ▼a(MiAaPQ)AAI31561597
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a004
■1001  ▼aTsai,  Meng-Jung.
■24510▼aComputer  Network  Optimization  Using  the  Power  Metric
■260    ▼a[Sl]▼bUniversity  of  California,  Los  Angeles▼c2024
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2024
■300    ▼a241  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  86-03,  Section:  B.
■500    ▼aAdvisor:  Kleinrock,  Leonard.
■5021  ▼aThesis  (Ph.D.)--University  of  California,  Los  Angeles,  2024.
■520    ▼aModern  network  research  focuses  on  optimizing  performance  through  congestion  control,  quality  of  service,  and  fairness.  With  the  rapid  expansion  of  networks  and  increasing  traffic,  balancing  throughput  and  response  time  has  become  critical.  This  thesis  explores  this  tradeoff  and  introduces  the  Power  metric  as  a  tool  for  optimizing  network  performance  and  expands  its  investigation  to  achieving  optimized  performance  with  optimum  fairness. The  Power  metric,  defined  as  the  ratio  of  normalized  throughput  to  normalized  mean  response  time,  serves  as  our  performance  optimization  goal.  Previous  research  primarily  focused  on  single-flow  systems,  but  contemporary  networks  involve  multiple  flows  with  more  complex  scenarios.  This  work  extends  Power  analysis  in  performance  optimization  to  modern  network  environments,  developing  a  model  that  also  accommodates  multiple  flows.  We  further  examine  different  queueing  disciplines  that  implement  various  levels  of  flow  discrimination.  In  addition,  we  examine  fairness  metrics  coupled  with  performance  optimization.Our  research  focuses  on  three  aspects:  performance,  flow  priority  discrimination,  and  fairness.  We  introduce  performance  metrics,  including  individual  power,  sum  of  power,  and  average  power,  and  optimize  these  metrics  using  an  M/M/1  system  model  with  multiple  flows  under  different  queueing  disciplines.  We  also  explore  fairness  metrics  such  as  throughput,  delay,  and  power,  and  investigate  scenarios  where  optimum  performance  and  equal  fairness  can  be  achieved  simultaneously.Additionally,  we  study  generalized  power,  which  allows  specifying  the  relative  preference  for  throughput  versus  delay,  providing  a  flexible  approach  to  optimizing  network  performance  based  on  specific  requirements.In  summary,  this  research  represents  a  first  step  in  incorporating  performance,  fairness,  and  priority  flow  discrimination  into  the  Power  metric  analysis  for  modern  multi-flow  network  environments.  Our  goal  is  to  provide  insights,  guidance,  and  "rules  of  thumb"  for  system  designers  to  create  more  efficient  and  equitable  network  systems.
■590    ▼aSchool  code:  0031.
■650  4▼aComputer  science
■650  4▼aComputer  engineering
■650  4▼aInformation  technology
■650  4▼aElectrical  engineering
■653    ▼aCongestion  control
■653    ▼aNetwork  optimization
■653    ▼aPower  metric
■653    ▼aQueueing
■653    ▼aScheduling
■653    ▼aPerformance  optimization
■690    ▼a0984
■690    ▼a0544
■690    ▼a0489
■690    ▼a0464
■71020▼aUniversity  of  California,  Los  Angeles▼bComputer  Science  0201.
■7730  ▼tDissertations  Abstracts  International▼g86-03B.
■790    ▼a0031
■791    ▼aPh.D.
■792    ▼a2024
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17164139▼nKERIS▼z이  자료의  원문은  한국교육학술정보원에서  제공합니다.

미리보기

내보내기

chatGPT토론

Ai 추천 관련 도서


    신착도서 더보기
    최근 3년간 통계입니다.

    소장정보

    • 예약
    • 소재불명신고
    • 나의폴더
    • 우선정리요청
    • 비도서대출신청
    • 야간 도서대출신청
    소장자료
    등록번호 청구기호 소장처 대출가능여부 대출정보
    TF10712 전자도서 대출가능 마이폴더 부재도서신고 비도서대출신청 야간 도서대출신청

    * 대출중인 자료에 한하여 예약이 가능합니다. 예약을 원하시면 예약버튼을 클릭하십시오.

    해당 도서를 다른 이용자가 함께 대출한 도서

    관련 인기도서

    로그인 후 이용 가능합니다.