/Low-Power-Long-Range-UAV-Detection

Enemy UAV Detection with TensorFlow Lite Micro in LoRa mesh network

Primary LanguageC++

BTT - Be The Top

🧑‍💻👩‍💻 Member

Soonchan Kwon Gihwan Kim Nahyeong Kim Nawon Kim Karteikay Dhuper Prakshi Chander
Chonnam Univ Chungnam Univ Chungnam Univ Chonnam Univ Purdue Univ Purdue Univ

🏛️ Project title

Airspace Counter Drone System (ACDS) using observer drones to detect unidentified aerial object in LoRa mesh network.

🔍 Research problem statements

With the increasing popularity of UAVs, they have become more easily accessible to the public, companies, and even terrorists. This has raised the need for having a Counter Drone System (CDS). There have already been several attempts of implementing a CDS, but Ground Counter Drone System (GCDS) has a geological restriction and Airspace Counter Drone System (ACDS) in NASA research uses a WiFi mesh network which is fast but consumes high-power. The system is ACDS uses observer drones to detect unidentified aerial object in LoRa mesh network.

🎯 Research novelty

to the ground for detection, so we can use low-data-rate and long-range communication technology. network coverage to be expanded just by adding a new drone.

🤖 Overview of the system

The system is classified into three:

  • Ground LoRa mesh network - Base station where communicate with UAVs by LoRa mesh network
  • Node LoRa mesh network - UAV which communicate with ground and UAV by LoRa mesh network
  • Detection - Small and simple UAV detection model

🛠️ Environment setting

The system consist of LoRa mesh network and detection is developed on Arduino and run on ESP32-WROVER.

The Structure of hardware

Software development platform


Arduino




arduino-esp32




RadioHead




TensorFlow Lite for Microcontrollers