Loading...
|
Please use this identifier to cite or link to this item:
https://nccur.lib.nccu.edu.tw/handle/140.119/158705
|
| Title: | 分群策略於網路切片設計階段之比較與應用分析
A Comparative and Applied Analysis of Clustering Strategies in the Design Phase of Network Slicing |
| Authors: | 廖韋雅
Liao, Wei-Ya |
| Contributors: | 張宏慶
Jang, Hung-Chin
廖韋雅
Liao, Wei-Ya |
| Keywords: | 第五代行動通訊
網路切片
分群分析
K-means
HDBSCAN
5G
Network slicing
Clustering analysis
K-means
HDBSCAN |
| Date: | 2025 |
| Issue Date: | 2025-08-04 15:09:29 (UTC+8) |
| Abstract: | 隨著第五代行動通訊技術的蓬勃發展，網路切片已經成為支援多樣化應用場景的重要技術之一。在網路切片的準備階段，如何有效規劃初始網路切片設計，是網路架構規劃人員所要面臨的首要挑戰，而其設計品質將直接影響後續系統資源分配與服務品質。
本研究以分群分析為核心，從分群架構、資料類型以及分群演算法三種策略面向出發，設計八種實驗組合進行系統性比較分析，以探討適用於網路切片初始設計階段的分群策略，作為未來網路切片配置時參考依據。研究中採用K-means 與 HDBSCAN兩種演算法，於單層式與兩層式分群架構下，以無線端特徵與服務端特徵資料作為輸入資料集。透過Silhouette Coefficient、Calinski–Harabasz Index以及Davies–Bouldin Index三種分群評估指標以及樣本分佈均衡性，以兼顧分群品質與實務適用性為考量，找出適用於初始網路切片設計階段的分群參考策略。經由研究結果得到以下結論：(1)單層式分群在效能指標上表現優異，而兩層式分群則可有效改善樣本分布不均問題，提升資源配置公平性。(2)混合無線端與服務端特徵資料有助於分群結果的均衡性，優於僅使用無線端特徵資料。(3)K-means 在分群結果具有高穩定性，但常有極端群體，而 HDBSCAN具備識別雜訊資料能力，適合結構複雜場景。整體而言，將兩種演算法混合應用於兩層式架構中，能有效兼顧穩定性與分群結構解釋力。
With the rapid advancement of fifth-generation (5G) communication technologies, network slicing has become essential for supporting diverse application scenarios. Effectively planning the initial slice design is a primary challenge for network architects, significantly impacting resource allocation and service quality.
This study uses clustering analysis to systematically evaluate strategies for initial network slice design, exploring clustering architectures, data types, and algorithms. Eight experimental combinations with K-means and HDBSCAN algorithms are assessed under single-layer and two-layer architectures, using wireless-side and service-side datasets. Evaluation involves the Silhouette Coefficient, Calinski–Harabasz Index, Davies–Bouldin Index, and sample distribution balance.
Results indicate that:
(1) Single-layer clustering excels in clustering quality, while two-layer clustering reduces sample imbalance, enhancing resource allocation fairness.(2) Combining wireless and service-side data yields more balanced outcomes than wireless-side data alone.(3) K-means provides stable clusters but can yield extreme distributions, whereas HDBSCAN effectively identifies noise, suitable for complex scenarios. Integrating both algorithms within a two-layer framework balances clustering stability and interpretability. |
| Reference: | [1]Aloupogianni, E., Karyotis, C., Maniak, T., Iqbal, R., Passas, N., Vujičić, Z., & Doctor, F. (2024). Network slicing for beyond 5G networks using machine learning. In Proceedings of the 2024 IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing Workshops (CCGridW) (pp. 197–200). IEEE. https://doi.org/10.1109/ccgridw63211.2024.00031
[2]Amur, S. H., Zia, K., Chiumento, A., & Havinga, P. (2023). Autonomous network slicing and resource management for diverse QoS in IoT networks. In 2023 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops) (pp. 160–165). IEEE.
[3]Banchs, A., de Veciana, G., Sciancalepore, V., & Costa-Pérez, X. (2020). Resource allocation for network slicing in mobile networks. IEEE Access, 8, 214696–214706.
[4]Berry, M. J. A., & Linoff, G. S. (1997). Data mining techniques: For marketing, sales, and customer support. Wiley.
[5]Calinski, T., & Harabasz, J. (1974). A dendrite method for cluster analysis. Communications in Statistics, 3(1), 1–27.
[6]Campello, R. J. G. B., Moulavi, D., & Sander, J. (2013). Density-based clustering based on hierarchical density estimates. In Lecture Notes in Computer Science (Vol. 7819, pp. 160–172). Springer.
[7]Contu, C. A., Ciobanu, A., Borcoci, E., Vochin, M., Balapuwaduge, I. A. M., Topoloi, S.-G., & Trifan, R.-F. (2022). Deploying use case specific network slices using an OSM automation platform. In 2022 International Symposium on Wireless Personal Multimedia Communications (WPMC) (pp. 387–391). IEEE. https://doi.org/10.1109/WPMC55625.2022.10014883
[8]Raca, D., Leahy, D., Sreenan, C. J., & Quinlan, J. J. (2020). Beyond throughput, the next generation: A 5G dataset with channel and context metrics. In ACM Multimedia Systems Conference (MMSys), Istanbul, Turkey.
[9]Davies, D. L., & Bouldin, D. W. (1979). A cluster separation measure. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1(2), 224–227.
[10]Ericsson. (2024, November). Ericsson Mobility Report. https://www.ericsson.com/zh-tw/reports-and-papers/mobility-report/reports/november-2024
[11]Ester, M., Kriegel, H. P., Sander, J., & Xu, X. (1996). A density-based algorithm for discovering clusters in large spatial databases with noise. In Proceedings of the Second International Conference on Knowledge Discovery and Data Mining (pp. 226–231). AAAI Press.
[12]3GPP. (2020). System architecture for the 5G system (Release 16). 3GPP TS 23.501.
[13]Habibi, M. A., Han, B., Yousaf, F. Z., & Schotten, H. D. (2021). How should network slice instances be provided to multiple use cases of a single vertical industry. arXiv preprint arXiv:2101.00001.
[14]Intelligent-Slicing. (2023). An AI-assisted network slicing framework for 5G-and-beyond networks. IEEE Transactions on Network and Service Management, 20(2), 1024–1039. https://doi.org/10.1109/tnsm.2023.3274236
[15]ITU-R. (2015). IMT vision – Framework and overall objectives of the future development of IMT for 2020 and beyond. Recommendation ITU-R M.2083-0.
[16]Khan, S., Kiani, S. H., & Iqbal, M. Z. (2019). Network slicing and clustering in vehicular communications: Design and implementation. Vehicular Communications, 19, 100180.
[17]Kumar, D., Iyer, S., & López, O. A. (2024). Evolution of mobile networks. In S. Iyer, A. Kalla, O. A. López, & C. De Alwis (Eds.), Intelligent Spectrum Management: Towards 6G (pp. 1–25). Wiley.
[18]MacQueen, J. (1967). Some methods for classification and analysis of multivariate observations. In Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability (Vol. 1, pp. 281–297). University of California Press.
[19]Priyanka, R., & Chandrasekar, C. (2024). Efficient slice creation in network slicing using K-prototype clustering and context-aware slice selection for service provisioning. Retrieved from https://www.researchgate.net/publication/377807469
[20]Rappaport, T. S., Sun, S., Mayzus, R., Zhao, H., Azar, Y., Wang, K., ... & Gutierrez, F. (2017). Overview of millimeter wave communications for fifth-generation (5G) wireless networks—With a focus on propagation models. IEEE Transactions on Antennas and Propagation, 65(12), 6213–6230.
[21]Rousseeuw, P. J. (1987). Silhouettes: A graphical aid to the interpretation and validation of cluster analysis. Journal of Computational and Applied Mathematics, 20, 53–65.
[22]Singh, R., Soni, N., & Kumar, N. (2020). Machine learning-based sub-slicing in 5G networks: A step towards efficient resource management. Computer Communications, 154, 456–464. https://doi.org/10.1016/j.comcom.2020.02.013
[23]Singhal, G., & Roy, S. (2017). A survey on data clustering. International Journal of Advanced Engineering and Management, 2(8), 183. https://doi.org/10.24999/IJOAEM/02080042
[24]Sonawane, R. C., & Patil, H. D. (2020). Clustering techniques and research challenges in machine learning. In 2020 International Conference on Computing Methodologies and Communication (ICCMC) (pp. 227–233). IEEE. https://doi.org/10.1109/ICCMC48092.2020.ICCMC-00054
[25]Steinhaus, H. (1956). Sur la division des corps matériels en parties. Bulletin de l’Académie Polonaise des Sciences, 1(12), 801–804.
[26]Taleb, T., Bensalem, D. E., & Laghrissi, A. (2019). Smart service-oriented clustering for dynamic slice configuration. In 2019 IEEE Global Communications Conference (GLOBECOM) (pp. 1–6). IEEE. https://doi.org/10.1109/GLOBECOM38437.2019.9013564
[27]Tang, D., Rui, D., Tang, L., Sijia, Z., & Jianping, M. (2018). Low-rate DoS attack detection based on two-step cluster analysis. In Artificial Intelligence and Security (pp. 92–104). Springer. https://doi.org/10.1007/978-3-030-01950-1_6
[28]Wang, W., Wang, H., Fan, W., & Fu, J. (2024). Task clustering and resource allocation based on network slicing in IoV scenario. In Proceedings of the 7th International Conference on Information Technology and Network Engineering (pp. 123–127). IEEE. https://doi.org/10.1109/itnec60942.2024.10733163
[29]Yin, H., Aryani, A., Petrie, S., Nambissan, A., Astudillo, A., & Cao, S. (2024). A rapid review of clustering algorithms. arXiv preprint arXiv:2401.07389. https://arxiv.org/abs/2401.07389
[30]Zhang, H., Liu, N., Chu, X., Long, K., Aghvami, A. H., & Leung, V. C. M. (2017). Network slicing based 5G and future mobile networks: Mobility, resource management, and challenges. IEEE Communications Magazine, 55(8), 138–145. |
| Description: | 碩士
國立政治大學
資訊科學系碩士在職專班
109971019 |
| Source URI: | http://thesis.lib.nccu.edu.tw/record/#G0109971019 |
| Data Type: | thesis |
| Appears in Collections: | [資訊科學系碩士在職專班] 學位論文
|
Files in This Item:
| File |
Description |
Size | Format | |
|---|
| 101901.pdf | | 9846Kb | Adobe PDF | 0 | View/Open |
|
All items in 政大典藏 are protected by copyright, with all rights reserved.
|
