Machine learning based heuristic BBU-RRH switching scheme for C-RAN in 5G
| dc.contributor.advisor | Falowo, Olabisi | |
| dc.contributor.author | Liu, Jiamo | |
| dc.date.accessioned | 2020-03-02T09:27:13Z | |
| dc.date.available | 2020-03-02T09:27:13Z | |
| dc.date.issued | 2019 | |
| dc.date.updated | 2020-03-02T08:41:05Z | |
| dc.description.abstract | The immense increase in bandwidth demand by various services such as high definition video streaming, online gaming, and virtual reality has made it increasingly challenging for operators to provide satisfactory services to the end users while making a profit. Cloud Radio Access Network (C-RAN) is a new architecture that has been proposed to facilitate the mobile networks' ability to meet the increase in bandwidth demand. C-RAN consists of three parts, namely Remote Radio Head (RRH), the front haul link, and Baseband Processing Units (BBU) pool. Many RRHs are associated with one BBU pool, and all RRHs within the pool are logically connected to every BBU in the pool. Thus, a BBU-RRH switching algorithm needs to be developed as it is able to enhance the performance of such architecture while managing the resource efficiently. This work mainly focuses on developing a traffic profile prediction-based BBU-RRH switching algorithm using a real life dataset. In the literature, there are related works that have proposed algorithms to achieve this purpose. However some of the existing algorithms suffer from high switching complexity while others fall short in QoS provision. Therefore, this work develops a BBU-RRH algorithm that to enhance the QoS while reducing the switching complexity, with the aid of machine learning techniques. The algorithm developed consists of three parts. The first part consists of an efficient RRH clustering mechanism that determines which RRHs are associated with a specific BBU pool. The second part utilizesrecurrent neural networks (RNN) to predict the daily traffic profile of RRHs, so that a relatively accurate traffic profile prediction can be obtained to facilitate the switching algorithm. Finally, the third part comprises the BBU-RRH switching scheme that works in conjunction with the predicted traffic profile to make an informed decision about the associations between RRHs and BBUs within the BBU pool. The performance of the proposed algorithm has been evaluated through simulations. The simulation results show that the proposed algorithm reduces the number of BBUs used and therefore save on energy. In addition, the algorithm reduces the occurrence of congestion and failure states, and thus improve the quality of the service of the network. Finally, the developed switching algorithm also reduces the switching complexity when compared with existing algorithms. | |
| dc.identifier.apacitation | Liu, J. (2019). <i>Machine learning based heuristic BBU-RRH switching scheme for C-RAN in 5G</i>. (). ,Engineering and the Built Environment ,Department of Electrical Engineering. Retrieved from http://hdl.handle.net/11427/31427 | en_ZA |
| dc.identifier.chicagocitation | Liu, Jiamo. <i>"Machine learning based heuristic BBU-RRH switching scheme for C-RAN in 5G."</i> ., ,Engineering and the Built Environment ,Department of Electrical Engineering, 2019. http://hdl.handle.net/11427/31427 | en_ZA |
| dc.identifier.citation | Liu, J. 2019. Machine learning based heuristic BBU-RRH switching scheme for C-RAN in 5G. . ,Engineering and the Built Environment ,Department of Electrical Engineering. http://hdl.handle.net/11427/31427 | en_ZA |
| dc.identifier.ris | TY - Thesis / Dissertation AU - Liu, Jiamo AB - The immense increase in bandwidth demand by various services such as high definition video streaming, online gaming, and virtual reality has made it increasingly challenging for operators to provide satisfactory services to the end users while making a profit. Cloud Radio Access Network (C-RAN) is a new architecture that has been proposed to facilitate the mobile networks' ability to meet the increase in bandwidth demand. C-RAN consists of three parts, namely Remote Radio Head (RRH), the front haul link, and Baseband Processing Units (BBU) pool. Many RRHs are associated with one BBU pool, and all RRHs within the pool are logically connected to every BBU in the pool. Thus, a BBU-RRH switching algorithm needs to be developed as it is able to enhance the performance of such architecture while managing the resource efficiently. This work mainly focuses on developing a traffic profile prediction-based BBU-RRH switching algorithm using a real life dataset. In the literature, there are related works that have proposed algorithms to achieve this purpose. However some of the existing algorithms suffer from high switching complexity while others fall short in QoS provision. Therefore, this work develops a BBU-RRH algorithm that to enhance the QoS while reducing the switching complexity, with the aid of machine learning techniques. The algorithm developed consists of three parts. The first part consists of an efficient RRH clustering mechanism that determines which RRHs are associated with a specific BBU pool. The second part utilizesrecurrent neural networks (RNN) to predict the daily traffic profile of RRHs, so that a relatively accurate traffic profile prediction can be obtained to facilitate the switching algorithm. Finally, the third part comprises the BBU-RRH switching scheme that works in conjunction with the predicted traffic profile to make an informed decision about the associations between RRHs and BBUs within the BBU pool. The performance of the proposed algorithm has been evaluated through simulations. The simulation results show that the proposed algorithm reduces the number of BBUs used and therefore save on energy. In addition, the algorithm reduces the occurrence of congestion and failure states, and thus improve the quality of the service of the network. Finally, the developed switching algorithm also reduces the switching complexity when compared with existing algorithms. DA - 2019 DB - OpenUCT DP - University of Cape Town KW - Electrical Engineering LK - https://open.uct.ac.za PY - 2019 T1 - Machine learning based heuristic BBU-RRH switching scheme for C-RAN in 5G TI - Machine learning based heuristic BBU-RRH switching scheme for C-RAN in 5G UR - http://hdl.handle.net/11427/31427 ER - | en_ZA |
| dc.identifier.uri | http://hdl.handle.net/11427/31427 | |
| dc.identifier.vancouvercitation | Liu J. Machine learning based heuristic BBU-RRH switching scheme for C-RAN in 5G. []. ,Engineering and the Built Environment ,Department of Electrical Engineering, 2019 [cited yyyy month dd]. Available from: http://hdl.handle.net/11427/31427 | en_ZA |
| dc.language.rfc3066 | eng | |
| dc.publisher.department | Department of Electrical Engineering | |
| dc.publisher.faculty | Faculty of Engineering and the Built Environment | |
| dc.subject | Electrical Engineering | |
| dc.title | Machine learning based heuristic BBU-RRH switching scheme for C-RAN in 5G | |
| dc.type | Master Thesis | |
| dc.type.qualificationlevel | Masters | |
| dc.type.qualificationname | MSc |