This presentation introduces ESPARGOS, an innovative, low-cost system designed to visualize Wi-Fi signal propagation in real-time through an augmented reality interface. Developed by Florian Euchner (DO7JE), ESPARGOS utilizes arrays of phase-synchronized ESP32 Wi-Fi chips to form a coherent antenna array, overcoming the high cost typically associated with MIMO and phased array systems. The talk details the system's architecture, including its hardware components, the critical phase synchronization method (using a shared 40MHz clock and a reference signal for calibration), and the software processing pipeline that transforms raw Channel State Information (CSI) into visual representations of signal direction, reflections, and delays. Through live demonstrations, the system effectively illustrates complex RF phenomena such as signal reflections off metallic surfaces, penetration through building materials, and the tracking of devices within attenuating enclosures, offering a unique educational and practical tool for exploring the "quasi-optic" properties of radio waves.
Making WiFi Visible with ESPARGOS and ESP32s: A Detailed Summary
0:00 Project Introduction: Florian Euchner from the University of Stuttgart introduces ESPARGOS, a project that aims to make Wi-Fi signals visible using augmented reality with phase-synchronized, low-cost ESP32 chips.
2:01 ESPARGOS System Overview: The system comprises an antenna array (e.g., 8x4 antennas, each backed by an ESP32 chip), a webcam mounted on top, and a computer for GPU/CPU processing to generate real-time video overlays.
3:15 Augmented Reality Demos: Florian demonstrates how ESPARGOS visualizes Wi-Fi traffic by overlaying real-time signal propagation onto a live camera feed:
3:15 Signal Tracking: Visualizing the Wi-Fi signal emitted from a smartphone.
3:51 Reflections & Transparency: Showing how metallic walls act as RF mirrors, and surprisingly, how gypsum walls are more transparent to microwaves than glass.
4:47 Directional Antennas: Illustrating the focused beam of a Yagi antenna and its reflections off surfaces.
5:19 Signal Attenuation: Tracking a device inside a microwave oven, which provides about 30 dB of attenuation but doesn't completely block the signal.
5:58 Signal Blocking: Demonstrating a specialized "Faraday bag" with carbon fibers that effectively blocks Wi-Fi signals.
21:54 Accidental Antenna Effect: A cable poking out of a sealed Faraday bag can inadvertently act as an antenna, allowing signal leakage.
6:52 Leveraging Wi-Fi Packet Structure: The system analyzes the "Legacy Long Training Field" (LLTF) within Wi-Fi packets. This pilot symbol, with its known content, is used by the ESP32 to estimate channel propagation effects, including time/phase offsets, path loss, and multipath propagation.
8:33 Channel State Information (CSI): Wi-Fi's OFDM (Orthogonal Frequency-Division Multiplexing) modulation allows the ESP32 to extract Channel State Information (CSI) for each subcarrier. This CSI, comprising amplitude and phase as complex numbers, is readily accessible via the ESP32's API.
12:29 The Synchronization Challenge: A critical hurdle is achieving precise phase and time synchronization between all ESP32s in the array to accurately measure tiny phase differences caused by signal arrival times.
13:36 ESPARGOS Hardware Architecture: ESPARGOS employs a unique, low-cost architecture where each ESP32 functions as a complete processing chain (analog to digital CSI extraction), with a two-stage digital signal processing approach (on-chip and then on a central computer).
Frequency Synchronization: All ESP32s share a common, distributed 40MHz clock.
Phase Synchronization: To correct for unknown phase offsets generated by the ESP32's internal PLL, a reference ESP32 transmits a Wi-Fi packet over precisely known microstrip lines on the PCB, allowing software to calibrate and subtract the phase error.
16:22 Software Processing for Visualization: The raw CSI data (a 3D array of antenna elements, subcarriers, and time) undergoes a 3D Fourier Transform to convert it into "beam space." This beam space is then translated into azimuth and elevation angles (Angle of Arrival - AoA) and overlaid onto the webcam image.
22:24 Visualizing Delays and Velocities: By including the frequency axis in the Fourier transform, the system can measure and visualize signal delays (e.g., shorter paths in blue, longer in red) and has the potential to measure relative velocities, similar to radar signal processing.
23:10 Localization Applications: ESPARGOS can also be used for advanced localization by combining data from multiple antenna arrays, utilizing Time Difference of Arrival (TDoA) and Angle of Arrival (AoA).
23:46 Project Status & Availability: Florian offers a live demo of ESPARGOS at the German Amateur Radio Club booth and provides more information, including CSI datasets, on the project website. He is exploring small-scale manufacturing of the system.
25:29 ESP32 Documentation & Limitations: While ESP32s offer well-documented access to CSI, the phase synchronization method was custom-developed. The ESP32's 8-bit CSI resolution limits the dynamic range, and the overall image resolution of ESPARGOS is ultimately constrained by physical diffraction limits.