第五代行動通訊網路 (5G) 藉由軟體定義網路、網路功能虛擬化,以及網路切片等技術,使得實體網路設施得以虛擬化,讓網路之管理與擴充更為彈性,且能劃分成不同服務,將有助於滿足5G多樣化之需求與垂直應用。 本論文聚焦於組織內之5G網路,且資料經過加密之即時通訊垂直網路切片。然而,通訊的可靠度極為重要,為確保通訊過程不會因故中斷之服務品質,基於此需求,我們提出了一個數學規劃模型,在滿足高可靠度、高安全性、低延遲的情況下,以最小成本為目標建置出網路切片,希冀能提供決策建議給廠商或承租人參考,以作為較佳之資源規劃與佈建之依據。 我們將此複雜之問題以數學模型表示,目標為最小化建置成本,並且使用拉格朗日鬆弛法來解此問題,並開發了一個演算法來求得可行解,並藉由一系列的實驗證實此演算法具有最佳化網路資源規劃與建置的效能。
With the help of SDN, NFV, and network slicing, the 5G network can virtualize physical networks into isolated logical networks, which makes the management and extension of network more agile and effective. Therefore, it is more likely to satisfy the goals of 5G, which is diverse applications and vertical industries. Focusing on a network slicing of instant messaging in an organization which requires data encryption and high reliability, the fluent transmission is vital for ensuring uninterrupted transportation. Based on these considerations, we proposed an optimization deployment model, which deploys the network slicing with minimum cost and meanwhile ensures high security, high reliability, and low delay. We hope to provide the industry segments and tenants a reference for better service planning and purchase decisions. The complex problem is formulated to a mathematical model and solved optimally by using the Lagrangian relaxation method. Besides, we develop a Lagrangian relaxation based approach for getting feasible solutions. In the end, the proposed approach is validated by computational experiments to be effectively to optimally deploy a network slicing that satisfies economic efficiency, high reliability, high security, and quality of service (QoS).