| 摘要: | 在科技日新月異的現在，網路成為我們密不可分的一部份，而隨著人類使用到網路的比例越大，網路性能的要求也越來越高。目前行動通訊已經發展到5G，為了網路對生活的不同應用，ITU-R(ITU Radiocommunication Sector)制定了三大場景，分別為URLLC(超可靠低延遲通訊)、eMBB(增強型行動頻寬)、mMTC(大規模機器型通訊)。三種場景所要求的性能各有不同，基地台中資源利用的設計也有不同考量的點，在現今許多環境中可能會同時需要多種場景使用，因此多個場景的共存非常考驗系統架構的設計。
網路環境中的使用者，在需要上傳資料至網路時，必須先與基地台進行連線，這個連線的過程稱之為隨機存取(Random Access)，而由於基地台資源是有限的，因此對於所有使用者必須採取競爭的方法來進行連線，只有競爭成功才能開始上傳資料，而對於URLLC來說，更發展了另一種上傳方法Grant-free(無允諾上行)來協助其低延遲的特性，本論文主要探討URLLC與eMBB兩種場景應用共存，在兩者使用不同的隨機存取方法時，必須要讓兩者皆達到設想的性能，但不同的資源分配方法各有好壞，因此本論文除了探討分組的動態調整方法之外，同時也設計出用來切換不同分組模式的判斷方法，讓基地台可以根據目前環境中UE的成功率或者數量來決定當前所選擇的模式，以求讓URLLC的效能增加。
本論文通過討論兩種分組方法、三種切換方法和兩種調整參數方式來進行模擬並分析，從中一一進行分析並找出每種方法對應的適合場景、優缺點等，同時做出統整並挑選出最理想的切換方法，並對兩種調整參數可以如何優化做進一步說明。
;In the rapidly evolving world of technology, the internet has become an integral part of our lives. As the percentage of people using the internet continues to grow, the demand for network performance is also increasing. Currently, mobile communication has advanced to 5G, and in order to cater to different applications of the network in daily life, the ITU-R (ITU Radiocommunication Sector) has defined three major scenarios: URLLC (Ultra-Reliable Low-Latency Communication), eMBB (Enhanced Mobile Broadband), and mMTC (Massive Machine Type Communication). The performance requirements for three different scenarios vary, and the design considerations for resource utilization in base stations also differ. In many present-day environments, there may be a simultaneous need for multiple scenario usage, making the coexistence of multiple scenarios a significant challenge for system architecture design.
In a network environment, when users need to upload data to the network, they must first establish a connection with the base station. This connection process is known as random access. However, since base station resources are limited, a competitive method must be employed to establish connections for all users. Only those who successfully compete can begin uploading data. For URLLC, another upload method called Grant-free has been developed to assist its low-latency characteristics. This paper primarily explores the coexistence of URLLC and eMBB scenarios. When different random access methods are used for each scenario, it is crucial to ensure that both scenarios achieve the desired performance. However, different resource allocation methods have their pros and cons. Therefore, in addition to investigating dynamic adjustment methods for grouping, this paper also designs a decision-making method to switch between different grouping modes. This allows the base station to select the current mode based on factors such as the success rate or quantity of User Equipment (UE) in the current environment, aiming to improve the performance of URLLC.
This paper simulates and analyzes two grouping methods, three switching methods, and two parameter adjustment methods. Through this analysis, each method is examined to determine its suitable scenarios, advantages, disadvantages, and overall performance. The paper also aims to identify the most ideal switching method and provide further explanations on how to optimize the two parameter adjustment methods. |
