AI gym for Networks
Autor(a) principal: | |
---|---|
Data de Publicação: | 2022 |
Tipo de documento: | Dissertação |
Idioma: | eng |
Título da fonte: | Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
Texto Completo: | http://hdl.handle.net/10362/152022 |
Resumo: | 5G Networks are delivering better services and connecting more devices, but at the same time are becoming more complex. Problems like resource management and control optimization are increasingly dynamic and difficult to model making it very hard to use traditional model-based optimization techniques. Artificial Intelligence (AI) explores techniques such as Deep Reinforcement Learning (DRL), which uses the interaction between the agent and the environment to learn what action to take to obtain the best possible result. Researchers usually need to create and develop a simulation environment for their scenario of interest to be able to experiment with DRL algorithms. This takes a large amount of time from the research process, while the lack of a common environment makes it difficult to compare algorithms. The proposed solution aims to fill this gap by creating a tool that facilitates the setting up of DRL training environments for network scenarios. The developed tool uses three open source software, the Containernet to simulate the connections between devices, the Ryu Controller as the Software Defined Network Controller, and OpenAI Gym which is responsible for setting up the communication between the environment and the DRL agent. With the project developed during the thesis, the users will be capable of creating more scenarios in a short period, opening space to set up different environments, solving various problems as well as providing a common environment where other Agents can be compared. The developed software is used to compare the performance of several DRL agents in two different network control problems: routing and network slice admission control. A novel DRL based solution is used in the case of network slice admission control that jointly optimizes the admission and the placement of traffic of a network slice in the physical resources. |
id |
RCAP_f550ff13ed3898f9e3020d8a194907ae |
---|---|
oai_identifier_str |
oai:run.unl.pt:10362/152022 |
network_acronym_str |
RCAP |
network_name_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository_id_str |
7160 |
spelling |
AI gym for Networks5G NetworksDeep Reinforcement LearningContainernet,Ryu ControllerDomínio/Área Científica::Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática5G Networks are delivering better services and connecting more devices, but at the same time are becoming more complex. Problems like resource management and control optimization are increasingly dynamic and difficult to model making it very hard to use traditional model-based optimization techniques. Artificial Intelligence (AI) explores techniques such as Deep Reinforcement Learning (DRL), which uses the interaction between the agent and the environment to learn what action to take to obtain the best possible result. Researchers usually need to create and develop a simulation environment for their scenario of interest to be able to experiment with DRL algorithms. This takes a large amount of time from the research process, while the lack of a common environment makes it difficult to compare algorithms. The proposed solution aims to fill this gap by creating a tool that facilitates the setting up of DRL training environments for network scenarios. The developed tool uses three open source software, the Containernet to simulate the connections between devices, the Ryu Controller as the Software Defined Network Controller, and OpenAI Gym which is responsible for setting up the communication between the environment and the DRL agent. With the project developed during the thesis, the users will be capable of creating more scenarios in a short period, opening space to set up different environments, solving various problems as well as providing a common environment where other Agents can be compared. The developed software is used to compare the performance of several DRL agents in two different network control problems: routing and network slice admission control. A novel DRL based solution is used in the case of network slice admission control that jointly optimizes the admission and the placement of traffic of a network slice in the physical resources.As redes 5G oferecem melhores serviços e conectam mais dispositivos, fazendo com que se tornem mais complexas e difíceis de gerir. Problemas como a gestão de recursos e a otimização de controlo são cada vez mais dinâmicos e difíceis de modelar, o que torna difícil usar soluções de optimização basea- das em modelos tradicionais. A Inteligência Artificial (IA) explora técnicas como Deep Reinforcement Learning que utiliza a interação entre o agente e o ambiente para aprender qual a ação a ter para obter o melhor resultado possível. Normalmente, os investigadores precisam de criar e desenvolver um ambiente de simulação para poder estudar os algoritmos DRL e a sua interação com o cenário de interesse. A criação de ambientes a partir do zero retira tempo indispensável para a pesquisa em si, e a falta de ambientes de treino comuns torna difícil a comparação dos algoritmos. A solução proposta foca-se em preencher esta lacuna criando uma ferramenta que facilite a configuração de ambientes de treino DRL para cenários de rede. A ferramenta desenvolvida utiliza três softwares open source, o Containernet para simular as conexões entre os dispositivos, o Ryu Controller como Software Defined Network Controller e o OpenAI Gym que é responsável por configurar a comunicação entre o ambiente e o agente DRL. Através do projeto desenvolvido, os utilizadores serão capazes de criar mais cenários em um curto período, abrindo espaço para configurar diferentes ambientes e resolver diferentes problemas, bem como fornecer um ambiente comum onde diferentes Agentes podem ser comparados. O software desenvolvido foi usado para comparar o desempenho de vários agentes DRL em dois problemas diferentes de controlo de rede, nomeadamente, roteamento e controlo de admissão de slices na rede. Uma solução baseada em DRL é usada no caso do controlo de admissão de slices na rede que otimiza conjuntamente a admissão e a colocação de tráfego de uma slice na rede nos recursos físicos da mesma.Amaral, PedroRUNCapelo, Pedro Alexandre Francisco2023-04-21T18:38:18Z2022-122022-12-01T00:00:00Zinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/masterThesisapplication/pdfhttp://hdl.handle.net/10362/152022enginfo:eu-repo/semantics/openAccessreponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãoinstacron:RCAAP2024-03-11T05:34:09Zoai:run.unl.pt:10362/152022Portal AgregadorONGhttps://www.rcaap.pt/oai/openaireopendoar:71602024-03-20T03:54:41.675440Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãofalse |
dc.title.none.fl_str_mv |
AI gym for Networks |
title |
AI gym for Networks |
spellingShingle |
AI gym for Networks Capelo, Pedro Alexandre Francisco 5G Networks Deep Reinforcement Learning Containernet, Ryu Controller Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática |
title_short |
AI gym for Networks |
title_full |
AI gym for Networks |
title_fullStr |
AI gym for Networks |
title_full_unstemmed |
AI gym for Networks |
title_sort |
AI gym for Networks |
author |
Capelo, Pedro Alexandre Francisco |
author_facet |
Capelo, Pedro Alexandre Francisco |
author_role |
author |
dc.contributor.none.fl_str_mv |
Amaral, Pedro RUN |
dc.contributor.author.fl_str_mv |
Capelo, Pedro Alexandre Francisco |
dc.subject.por.fl_str_mv |
5G Networks Deep Reinforcement Learning Containernet, Ryu Controller Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática |
topic |
5G Networks Deep Reinforcement Learning Containernet, Ryu Controller Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Eletrotécnica, Eletrónica e Informática |
description |
5G Networks are delivering better services and connecting more devices, but at the same time are becoming more complex. Problems like resource management and control optimization are increasingly dynamic and difficult to model making it very hard to use traditional model-based optimization techniques. Artificial Intelligence (AI) explores techniques such as Deep Reinforcement Learning (DRL), which uses the interaction between the agent and the environment to learn what action to take to obtain the best possible result. Researchers usually need to create and develop a simulation environment for their scenario of interest to be able to experiment with DRL algorithms. This takes a large amount of time from the research process, while the lack of a common environment makes it difficult to compare algorithms. The proposed solution aims to fill this gap by creating a tool that facilitates the setting up of DRL training environments for network scenarios. The developed tool uses three open source software, the Containernet to simulate the connections between devices, the Ryu Controller as the Software Defined Network Controller, and OpenAI Gym which is responsible for setting up the communication between the environment and the DRL agent. With the project developed during the thesis, the users will be capable of creating more scenarios in a short period, opening space to set up different environments, solving various problems as well as providing a common environment where other Agents can be compared. The developed software is used to compare the performance of several DRL agents in two different network control problems: routing and network slice admission control. A novel DRL based solution is used in the case of network slice admission control that jointly optimizes the admission and the placement of traffic of a network slice in the physical resources. |
publishDate |
2022 |
dc.date.none.fl_str_mv |
2022-12 2022-12-01T00:00:00Z 2023-04-21T18:38:18Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/masterThesis |
format |
masterThesis |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
http://hdl.handle.net/10362/152022 |
url |
http://hdl.handle.net/10362/152022 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.format.none.fl_str_mv |
application/pdf |
dc.source.none.fl_str_mv |
reponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação instacron:RCAAP |
instname_str |
Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação |
instacron_str |
RCAAP |
institution |
RCAAP |
reponame_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
collection |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository.name.fl_str_mv |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação |
repository.mail.fl_str_mv |
|
_version_ |
1799138135351754752 |