| 摘要: | 歐洲電信標準協會(ETSI)倡議在第五代行動網路架構中利用多重接取邊緣計算 (Multi-Access Edge Computing, MEC)、軟體定義網路(Software-Defined Network, SDN)和網路功能虛擬化(Network Function Virtualization, NFV)等技術之整合以支援增強型行動頻寬(eMBB)、超可靠低延遲通訊(URLLC)和大規模物聯網通訊(mMTC)等服務。然而，就網路傳輸和資源使用而言，MEC雖具備低傳輸延遲之優勢，但是若僅單純依賴邊緣區域的計算資源，將因龐大資料流匯聚在邊緣區域而增加邊緣伺服器之計算和維運負擔，另一方面，就資料內容傳遞而言，5G核心網路即使導入SDN和NFV，若無法解決多路徑傳輸效率過低、路由鏈結容錯與恢復、以及網路資源分配不均等問題，將無法提升整體的網路服務效能。因此，在5G行動網路環境下，如何整合ETSI MEC和SDN，進而發展適切的資料傳輸與網路資源管理技術，為極具貢獻與價值之研究議題。 本計畫將針對5G行動網路環境的資料傳輸、網路資源配置與管理等議題，逐步進行一系列之子議題研究與技術開發。本計畫首先將由無線接取網路(Radio Access Network, RAN)、接取邊緣區域、延伸至核心網路，分析在5G環境下結合MEC、NFV與SDN的特性、限制及面臨的挑戰，並致力於提出有效的解決方法。本計畫之內容分為五個部分，依序進行；第一部分，研發合適的多接取邊緣計算機制，導入賽局理論方法進行邊緣區域之動態資源分配；第二部分，研發可適用於串流服務的多路徑資源分配機制，並導入SDN以有效管控網路資源之使用；第三部分，研發具服務品質保證(QoS)優先差異之路徑恢復機制，以保證不同串流服務之差異化資料傳輸，且維持整體網路傳輸效能；第四部分，透過網路切片技術(Network Slicing)，研發可適用於5G eMBB、URLLC和mMTC等不同資料服務之資料傳輸與路徑規劃機制；第五部份，研發可適用於5G多重接取邊緣計算之合作型導向的快取機制，透過對裝置移動的預測與網路編碼技術的結合，期能提升資料服務的效能。因此，本計畫執行至最後，我們將展現在5G行動網路環境下結合MEC、NFV與SDN之資料傳輸與網路資源管理之研發成果。 ;The European Telecommunications Standards Institute (ETSI) proposes to use Multi-Access Edge Computing (MEC), Software-Defined Network (SDN), and Network Function Virtualization (NFV) technologies in the fifth generation (5G) of mobile networks to support three major data services, such as Enhanced Mobile Broadband (eMBB), Ultra Reliable Low Latency Communications (URLLC) and Massive Machine Type Communications (mMTC). From the viewpoint of data transmission and network resource usage, MEC has the advantage of low transmission delay. However, if MEC only depends on the computing resources in the edge area, it will impose serious computation load and maintenance cost on the edge server because substantial data streams converge in the edge area. On the other hand, from the viewpoint of data content transfer, even if SDN and NFV technologies are introduced into the 5G core network, the network system cannot sustain the overall service performance if the issues of inefficient multi-path transmission, unreliable routing recovery, fault tolerance, and uneven distribution of network resources cannot be resolved yet. Therefore, it is crucial to study how to integrate ETSI MEC and SDN, and then develop appropriate data dissemination and network resource management technologies in the 5G mobile network environments. Regarding the theme of data dissemination, network resource allocation and management in the 5G mobile network environments, this research project will carry out a series of sub-topic study. Our study will analyze the characteristics, limitations and challenges of combining MEC, NFV and SDN under the 5G environment that comprise the radio access network (RAN), edge area, and core network. The content of this project includes five parts. First, we plan to employ the game theory in the multi-access edge computing environment to perform dynamic resource allocation in the edge area. Second, we plan to develop a multi-path resource allocation mechanism for high-throughput streaming services, and adopt the SDN technology to effectively control the use of network resources. Third, to ensure streaming services with differentiated service guarantee, we will design a QoS-oriented and priority-based routing path recovery mechanism. Fourth, through the network slicing technology, we will design respective data transfer and path planning mechanisms for eMBB, URLLC and mMTC. Fifth, we will propose a cooperative-oriented caching mechanism that can be applied to multiple MEC servers in 5G network. Using both device movement prediction and network coding techniques, we expect to enhance the performance of data caching services. Therefore, in accordance with all above, we will exhibit the research and development efforts of data dissemination and network resource management by advantages of MEC, SDN and NVF in the 5G mobile network environments. |
