請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/15281
標題: | 蜂巢式網路天線傾斜角的分散式優化法 A Distributed Optimization Method of Antenna Downtilt in Cellular Networks |
作者: | Ming-Yu Chiang 江明昱 |
指導教授: | 蔡志宏(Zse-Hong Tsai) |
關鍵字: | 指向性天線,傾斜角,分散式演算法,基因演算法,自我組織網路, Directional antenna,Downtilt angle,Distributed algorithm,Genetic algorithm,Self-organizing network, |
出版年 : | 2020 |
學位: | 碩士 |
摘要: | 由於近年來無線裝置大幅增加和連線速率的要求大幅成長,使得無線網路中的物理層管理就變得相當重要。指向性天線可以讓電磁波在特定方向產生高的增益,使得接收區域的信號強度增加。而調整天線的傾斜角可以讓主波束更集中在使用者裝置的位置。然而,當多個基地台同時在調整傾斜角時,基地台間所造成的干擾會嚴重影響訊號的品質,因此有一套完整調整傾斜角的方法變顯得相當重要。 本論文因而提出一套分散式的調整演算法,基於斯塔克伯格競爭(Stackelberg competition)的基礎上,我們提出了兩階段的演算法。在第一階段中,指向性天線根據選擇自己的使用者的位置進行調整,讓其使用者接收到最強加總訊號。在第二階段,我們提出了干擾懲罰權重和覆蓋面積兩參數,使不同天線的主波束觸及的位置保持一定的距離。最後,我們在部分範圍內利用啟發式演算法找出最佳解。 在模擬測試中,我們在高密度的狀態下約有基因演算法的百分之九十的速率。在低密度的狀態下,速率和運算效能皆超越基因演算法。然而相較於集中式演算法如基因演算法,運算時間和傳輸時間上大幅的降低,在有新的裝置連線和部分裝置離線的情況下,本演算法可快速地找出新的角度,使優化效果高於基因演算法。 Due to the variety of wireless devices and the requirement for throughput have increased dramatically in recent years, the management of physical layer in wireless networks becomes more important than before. The directional antenna has high gain and increase the signal strength in specified direction. Adjusting downtilt angle of directional antenna can let the main beam focus on the position of the user devices. However, when multiple base transceiver stations parallelly adjust downtilt angle, the interference caused between them will seriously affect the signal quality. Therefore, it is important to find a complete method to adjust the downtilt angle for base station antenna, especially when the network is large. This thesis proposes a distributed algorithm based on Stackelberg competition. We propose a two-stage algorithm. In the first stage, the directional antenna adjusts tilt and let main beam toward the user devices which select this antenna, the total received power would be maximum. In the second stage, we proposed two parameters, the interference penalty weight and the coverage area size, to keep the position of main beam of different antenna at appropriate distance. Finally, we use heuristic algorithm to find the best solution. In our simulation, we achieve about 90% of the total throughput of the genetic algorithm in the environment with a high population density. In a low population density, the throughput improvement is more than 100%. Compared with the genetic algorithm which is a centralized algorithm, computing time and transmission time in our algorithm are small. While wireless devices start to create a connection and some devices leave, this algorithm can quickly find new angles and make the higher optimization than the genetic algorithm. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/15281 |
DOI: | 10.6342/NTU202000288 |
全文授權: | 未授權 |
顯示於系所單位: | 電信工程學研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-109-1.pdf 目前未授權公開取用 | 4.63 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。