Wi-Fi 6E無人機群對地面控制站之雙向高頻寬低延遲韌性傳輸設計

Abstract

隨著無人機應用於各領域已帶來大量商機，無人機群的無線通訊是近期熱門的議題。無人機群作為空間偵蒐影像傳輸常有帶寬不足並導致高延遲的問題，三機對單站之節點傳輸需滿足規格:寬頻(>63Mbps)、低延遲(飛控訊號≤ 50ms及影像傳輸≤140 ms)、畫面無遺失(0%)，以及單機對地面控制站鏈結不上時，可經由其他無人機轉發圖像給地面控制站，因此有對雙向、高頻寬、低延遲、韌性傳輸的需求。 本研究旨在設計一套基於 Wi-Fi 6E的星狀通訊架構，作為小型無人機群空間偵蒐影像對地面控制站傳輸的基礎無線寬頻Wi-Fi 6E雙向傳輸系統，同時擬定符合3GPP效能規範的無線寬頻即時傳輸需求規範，以支援三機對單一地面控制站之攝像傳輸，於6GHz頻段以TCP可靠傳輸作Wi-Fi 6E雙向傳輸系統，提供三機對單站之地面控制站達280Mbps的高寬頻傳輸，端到端距離20公尺處傳輸延遲由飛控訊號的420毫秒降低至30毫秒內。為提升小機群通訊網路韌性，本論文並考慮其中一機與地面站失去鏈結的狀況，設計和實作星狀拓樸之自我修復機制，以及基於週期性鄰機探測、鏈結建立、影像傳輸紀錄來自動化驗證通訊狀況的轉發演算法，在單一無人機和地面控制站間通道不良時可彈性建立轉傳鏈結和完成傳輸，實驗顯示影像經轉發後影像遺失率為0%，地面控制站可區隔影像來源無人機，且在延遲3.59秒完成轉發並完整取得三機攝像資訊。

With the growing commercial opportunities brought by the application of drones across various fields, wireless communication for drone swarms has become a popular topic in recent years. Drone swarms used for aerial surveillance often face issues such as insufficient bandwidth, resulting in high latency. The transmission requirements for a three-UAV-to-one-ground-station setup are as follows: high bandwidth (>63 Mbps), low latency (flight control ≤50 ms, image transmission ≤140 ms), zero image loss (0%), and in the event that a single UAV cannot establish a direct link with the ground control station, image data must be relayed to the ground station via other UAVs. Therefore, there is a growing demand for bidirectional, high-bandwidth, low-latency, and resilient transmission. This study aims to design a Wi-Fi 6E-based star topology communication architecture between a small drone swarm and a ground control station (GCS), serving as the foundational wireless broadband communication platform for aerial surveillance image transmission. Additionally, it tailors a set of wireless broadband real-time transmission requirements that meet 3GPP performance specifications. This study specifically proposed and implemented a Wi-Fi 6E communication network architecture operating in the 6 GHz band via TCP. It enables bidirectional transmission of 280 Mbps, end-to-end transmission latency reduced from 420 to within 30 msec at 20 m. Furthermore, this study designs and implements a self-healing mechanism for the star topology, as well as an automated forwarding algorithm for communication status verification based on periodic neighbor detection, link establishment, and image transmission logging. This algorithm enhances the resilience of small UAV communication networks by flexibly establishing relay links and completing transmission when the communication channel between a single UAV and the ground control station is poor. Experiments show that after relay transmission, the image loss rate is 0%, the ground control station can differentiate the image sources from different UAVs, and the image is successfully forwarded and fully received with a delay of 3.59 seconds, collecting information from three UAV cameras.