본문

서브메뉴

Machine Learning for Traffic Prediction and Communication Efficient Data Analytic in Wireless Networks
Machine Learning for Traffic Prediction and Communication Efficient Data Analytic in Wirel...
Machine Learning for Traffic Prediction and Communication Efficient Data Analytic in Wireless Networks

Detailed Information

자료유형  
 학위논문 서양
최종처리일시  
20260202105544
ISBN  
9798263393373
DDC  
300
저자명  
Alamoudi, Abdulrahman Mohammed A.
서명/저자  
Machine Learning for Traffic Prediction and Communication Efficient Data Analytic in Wireless Networks
발행사항  
[Sl] : Georgia Institute of Technology, 2024
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2024
형태사항  
120 p
주기사항  
Source: Dissertations Abstracts International, Volume: 87-05, Section: A.
주기사항  
Advisor: Fekri, Faramarz.
학위논문주기  
Thesis (Ph.D.)--Georgia Institute of Technology, 2024.
초록/해제  
요약With the exponential growth of available data, deep learning has emerged as a fundamental tool for interpreting data abstractions and constructing computational models. It has revolutionized our understanding of information processing, facilitating exploration across diverse domains such as text and signal analysis, image and audio recognition, social network analysis, and bioinformatics. The overarching goal of this research is to minimize wireless network traffic and operational costs for mobile users and network operators, respectively. The integrated framework endeavors to develop predictive models by analyzing the behavior of mobile users within wireless networks and designing efficient, task-oriented models in wireless networks. Particularly, our research taps into machine learning to learn and forecast mobile user behaviors, design semantic communication systems over noisy channels, and implement unsupervised distributed functional compression over wireless channels.Machine Learning for Predicting Mobile Users Web Applications Traffic and Access-time Behavior in Wireless Networks: We first highlight the significant limitations of traditional browsing prediction models, primarily due to their lack of access to mobile user behavioral data. The browsing behavior of mobile users exhibits complex spatial and temporal dependencies with the accessed content. Further, mobile users are often unwilling to share their information with a third party for privacy concerns. Our research aims to address this gap by leveraging machine learning techniques to learn and predict mobile user web content and time access, thereby effectively capturing statistical dependencies and sequential behavior patterns. This approach seeks to alleviate traffic congestion on cellular networks during peak periods and enhance network connectivity by predicting users' browsing patterns. Particularly, we develop a Recurrent Neural Network (RNN) with Long Short-Term Memory (LSTM), leveraging engineered features to effectively learn and predict users' web browsing behaviors. Additionally, we propose a Self-Exciting Memory Neural Network (SMNN) to forecast the future access times of mobile users, utilizing self-exciting point processes and intensity modeling for prediction. We further expand our framework to predict group users behaviors within cell towers, where we resort on clustering methods to group the users based on their similar behaviors. Then, we develop a separate LSTM model for each cluster to predict web domain traffic activities for the cell tower.Since mobile user browsing content is dynamic, we focus next on mining non-overlapping clickstream patterns with noisy interleaving clicks within the mobile user clickstream history. Furthermore, we aim to label these identified patterns in the user browsing history. To accomplish these objectives, we employ a modified suffix tree to extract patterns accurately from the user browsing database. Subsequently, we propose a probabilistic graphical approach to model and capture dependencies between these extracted patterns and predict future clickstream pattern to be accessed.Machine Learning for Designing Semantic Communication Systems over Wireless Channels: Semantic communication systems play a crucial role in modern wireless networks by aligning communication processes with specific tasks or objectives, thereby optimizing resource utilization, enhancing efficiency, and improving overall system performance. In this research, we focus in text semantic communication systems which have largely been driven by the integration of machine learning techniques. The primary motivation behind adopting a text semantic approach, which focuses on conveying meaning rather than exact reconstruction, is to conserve resources such as bandwidth and improve the efficiency of information transmission to receiver. In particular, we propose a deep learning communication framework specifically designed for learning semantic text over orthogonal Additive White Gaussian Noise (AWGN) channels. This framework utilizes on the Information Bottleneck (IB) principle to reduce data size in compliance with the required Communication rate while preserving the semantic meaning of the text. Moreover, we incorporate sentence similarity into the learning objective function to ensure consistency between model training and performance evaluation. We leverage the capabilities of Large Language Models (LLMs) to guide the learning process, employing a knowledge distillation approach for online training. Specifically, we train a neural semantic encoder at the transmitter and a neural semantic decoder at the receiver using the proposed learning objective function in an online training paradigm.
일반주제명  
Privacy
일반주제명  
Traffic congestion
일반주제명  
Neural networks
일반주제명  
Transportation
기타저자  
Georgia Institute of Technology.
기본자료저록  
Dissertations Abstracts International. 87-05A.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008260126s2024        us                              c    eng  d
■001000017360541
■00520260202105544
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798263393373
■035    ▼a(MiAaPQ)AAI32315479
■035    ▼a(MiAaPQ)GeorgiaTech75326
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a300
■1001  ▼aAlamoudi,  Abdulrahman  Mohammed  A.
■24510▼aMachine  Learning  for  Traffic  Prediction  and  Communication  Efficient  Data  Analytic  in  Wireless  Networks
■260    ▼a[Sl]▼bGeorgia  Institute  of  Technology▼c2024
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2024
■300    ▼a120  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  87-05,  Section:  A.
■500    ▼aAdvisor:  Fekri,  Faramarz.
■5021  ▼aThesis  (Ph.D.)--Georgia  Institute  of  Technology,  2024.
■520    ▼aWith  the  exponential  growth  of  available  data,  deep  learning  has  emerged  as  a  fundamental  tool  for  interpreting  data  abstractions  and  constructing  computational  models.  It  has  revolutionized  our  understanding  of  information  processing,  facilitating  exploration  across  diverse  domains  such  as  text  and  signal  analysis,  image  and  audio  recognition,  social  network  analysis,  and  bioinformatics.  The  overarching  goal  of  this  research  is  to  minimize  wireless  network  traffic  and  operational  costs  for  mobile  users  and  network  operators,  respectively.  The  integrated  framework  endeavors  to  develop  predictive  models  by  analyzing  the  behavior  of  mobile  users  within  wireless  networks  and  designing  efficient,  task-oriented  models  in  wireless  networks.  Particularly,  our  research  taps  into  machine  learning  to  learn  and  forecast  mobile  user  behaviors,  design  semantic  communication  systems  over  noisy  channels,  and  implement  unsupervised  distributed  functional  compression  over  wireless  channels.Machine  Learning  for  Predicting  Mobile  Users  Web  Applications  Traffic  and  Access-time  Behavior  in  Wireless  Networks:  We  first  highlight  the  significant  limitations  of  traditional  browsing  prediction  models,  primarily  due  to  their  lack  of  access  to  mobile  user  behavioral  data.  The  browsing  behavior  of  mobile  users  exhibits  complex  spatial  and  temporal  dependencies  with  the  accessed  content.  Further,  mobile  users  are  often  unwilling  to  share  their  information  with  a  third  party  for  privacy  concerns.  Our  research  aims  to  address  this  gap  by  leveraging  machine  learning  techniques  to  learn  and  predict  mobile  user  web  content  and  time  access,  thereby  effectively  capturing  statistical  dependencies  and  sequential  behavior  patterns.  This  approach  seeks  to  alleviate  traffic  congestion  on  cellular  networks  during  peak  periods  and  enhance  network  connectivity  by  predicting  users'  browsing  patterns.  Particularly,  we  develop  a  Recurrent  Neural  Network  (RNN)  with  Long  Short-Term  Memory  (LSTM),  leveraging  engineered  features  to  effectively  learn  and  predict  users'  web  browsing  behaviors.  Additionally,  we  propose  a  Self-Exciting  Memory  Neural  Network  (SMNN)  to  forecast  the  future  access  times  of  mobile  users,  utilizing  self-exciting  point  processes  and  intensity  modeling  for  prediction.  We  further  expand  our  framework  to  predict  group  users  behaviors  within  cell  towers,  where  we  resort  on  clustering  methods  to  group  the  users  based  on  their  similar  behaviors.  Then,  we  develop  a  separate  LSTM  model  for  each  cluster  to  predict  web  domain  traffic  activities  for  the  cell  tower.Since  mobile  user  browsing  content  is  dynamic,  we  focus  next  on  mining  non-overlapping  clickstream  patterns  with  noisy  interleaving  clicks  within  the  mobile  user  clickstream  history.  Furthermore,  we  aim  to  label  these  identified  patterns  in  the  user  browsing  history.  To  accomplish  these  objectives,  we  employ  a  modified  suffix  tree  to  extract  patterns  accurately  from  the  user  browsing  database.  Subsequently,  we  propose  a  probabilistic  graphical  approach  to  model  and  capture  dependencies  between  these  extracted  patterns  and  predict  future  clickstream  pattern  to  be  accessed.Machine  Learning  for  Designing  Semantic  Communication  Systems  over  Wireless  Channels:  Semantic  communication  systems  play  a  crucial  role  in  modern  wireless  networks  by  aligning  communication  processes  with  specific  tasks  or  objectives,  thereby  optimizing  resource  utilization,  enhancing  efficiency,  and  improving  overall  system  performance.  In  this  research,  we  focus  in  text  semantic  communication  systems  which  have  largely  been  driven  by  the  integration  of  machine  learning  techniques.  The  primary  motivation  behind  adopting  a  text  semantic  approach,  which  focuses  on  conveying  meaning  rather  than  exact  reconstruction,  is  to  conserve  resources  such  as  bandwidth  and  improve  the  efficiency  of  information  transmission  to  receiver.  In  particular,  we  propose  a  deep  learning  communication  framework  specifically  designed  for  learning  semantic  text  over  orthogonal  Additive  White  Gaussian  Noise  (AWGN)  channels.  This  framework  utilizes  on  the  Information  Bottleneck  (IB)  principle  to  reduce  data  size  in  compliance  with  the  required  Communication  rate  while  preserving  the  semantic  meaning  of  the  text.  Moreover,  we  incorporate  sentence  similarity  into  the  learning  objective  function  to  ensure  consistency  between  model  training  and  performance  evaluation.  We  leverage  the  capabilities  of  Large  Language  Models  (LLMs)  to  guide  the  learning  process,  employing  a  knowledge  distillation  approach  for  online  training.  Specifically,  we  train  a  neural  semantic  encoder  at  the  transmitter  and  a  neural  semantic  decoder  at  the  receiver  using  the  proposed  learning  objective  function  in  an  online  training  paradigm.
■590    ▼aSchool  code:  0078.
■650  4▼aPrivacy
■650  4▼aTraffic  congestion
■650  4▼aNeural  networks
■650  4▼aTransportation
■690    ▼a0800
■690    ▼a0709
■71020▼aGeorgia  Institute  of  Technology.
■7730  ▼tDissertations  Abstracts  International▼g87-05A.
■790    ▼a0078
■791    ▼aPh.D.
■792    ▼a2024
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17360541▼nKERIS▼z이  자료의  원문은  한국교육학술정보원에서  제공합니다.

Preview

Export

ChatGPT Discussion

AI Recommended Related Books


    New Books MORE
    Statistics for the past 3 years. Go to brief

    Подробнее информация.

    • Бронирование
    • не существует
    • моя папка
    • Первый запрос зрения
    • Non-Book Loan Application
    • Nighttime Book Loan Application
    материал
    Reg No. Количество платежных Местоположение статус Ленд информации
    TF14796 전자도서 대출가능 My Folder 부재도서신고 비도서대출신청 야간 도서대출신청

    * Бронирование доступны в заимствований книги. Чтобы сделать предварительный заказ, пожалуйста, нажмите кнопку бронирование

    Books borrowed together with this book

    Related Popular Books

    Available after logging in.