Loading...

Evaluation of Intelligent Automation Solutions in 5G Cloud-Native Mobile Network

Mamaghani, Ali | 2024

0 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 57252 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Hossein Khalaj, Babak; Shah Mansouri, Hamed
  7. Abstract:
  8. In today's world, mobile network technology and artificial intelligence models have made significant advancements. In this project, we intend to combine the two elements of mobile networks and artificial intelligence to reveal new aspects of these two fields, which have been less practically addressed in the past. In implementing this mobile network, we use the cloud-native software concept, which existed in the software world before mobile networks and became widespread around 2014 with Google’s introduction of Kubernetes. Another notable point in the world of computer science that helps us is the recent advancements in the field of artificial intelligence. In fact, with increased processing power and a vast amount of training data for machine learning models, these models have made remarkable progress in recent years. Therefore, by using them, one can achieve intelligent automation. One category of these models that play an effective role in this project is large language models, which act as assistants for network engineers. In this project, we have three layers: infrastructure, network components, and higher-level software. In the infrastructure and higher-level software layers, we incorporate large language models. In the infrastructure layer, we create an assistant to check for failures, whereby this assistant identifies infrastructure problems and suggests solutions to the network engineer using large language models. Thus, the network engineer can save time finding solutions. In the higher-level software layer, we develop a chatbot that utilizes large language models to generate commands and settings for the fifth-generation network. This chatbot translates English into a comprehensible language for fifth-generation network tasks. The chatbot's functionality is based on text classification and text generation using artificial intelligence. Given that the chatbot’s output is not one hundred percent reliable, a feedback loop is activated in case of an error in the chatbot’s output, making improvements based on the error message. This way, even individuals who are not fully familiar with network programming can create their private network by describing their desired network to this system
  9. Keywords:
  10. Fifth Generation of Mobile Networks ; Machine Learning ; Large Language Model ; Cloud-Native Software ; Intelligent Automation

 Digital Object List

 Bookmark

No TOC