⚡ electrical engineering

Leveraging Space-Time Synchronization for Ultra-Spot Detection in mmWave/THz UAV-to-UAV Communications

This paper proposes a space-time synchronization technique using multiple Wireless Two-way Interferometry (multi-Wi-Wi) devices with spatially diverse antennas to accurately detect mmWave/THz ultra-spot locations and optimize UAV flight paths, achieving an experimental localization error of 37.16 cm and a 186 ms prediction latency.

Original authors: Phuc Duc Nguyen, Ryosuke Isogai, Keitarou Kondou, Satoshi Yasuda, Nobuyasu Shiga, Yozo Shoji

Published 2026-05-26

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Phuc Duc Nguyen, Ryosuke Isogai, Keitarou Kondou, Satoshi Yasuda, Nobuyasu Shiga, Yozo Shoji

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine two drones flying past each other at high speed. They want to share a massive amount of data (like a whole movie) in just a split second. To do this, they use a super-fast "laser beam" of radio waves (called mmWave or Terahertz). However, this beam is incredibly narrow—think of it not as a flashlight, but as a laser pointer. If the drones miss each other by even a few inches, or if they are slightly off-angle, the connection breaks instantly, and the data transfer fails.

The challenge is that these drones are moving, shaking in the wind, and the "laser beam" (the ultra-spot) is tiny. They need to know exactly where the beam is and exactly when to turn on their high-speed radios to catch it.

Here is how the researchers solved this problem, broken down into simple concepts:

1. The Problem: The "Blind Flyer"

Standard GPS is like a map that tells you you are in a city, but it's not precise enough to tell you which specific window of a building you are looking at. Since the "laser beam" is so small, standard GPS is too blurry to help the drone find it. The drone needs to know its position within centimeters, not meters.

2. The Solution: The "Wi-Fi Walkie-Talkie" (Wi-Wi)

The team used a special technology called Wi-Wi (Wireless Two-way Interferometry). Think of this as a super-precise walkie-talkie that doesn't just talk, but listens to the echo of the signal to measure distance.

How it works: One drone (Drone A) sends a signal. The other drone (Drone B) listens. By measuring tiny changes in the signal's "phase" (like the timing of a wave), they can calculate the distance between them with millimeter-level accuracy.
The Catch: The drone's own body (its arms, motors, and frame) acts like a wall. As the drone spins or moves, its own body blocks the signal, causing the "walkie-talkie" to lose its voice or hear static. This is called "shadowing."

3. The Innovation: The "Four-Eyed" Drone

To fix the problem of the drone blocking its own signal, the researchers didn't just put one sensor on the drone; they put four of them, spaced out on different arms of the drone.

The Analogy: Imagine you are trying to hear a friend in a noisy room, but your own shoulder keeps blocking the sound. If you have four ears placed on different sides of your head, it's very unlikely that your shoulder will block all four ears at the exact same time.
The Result: If one sensor gets blocked or loses the signal, the others keep working. The system picks the best data from the group, ensuring the drone never loses track of the other one. This is called Spatial Diversity.

4. The Strategy: The "Dance Partner" Algorithm

The drone uses a two-step dance to find the beam:

The "Smell" Test (RSSI): From far away, the drone checks the "strength" of the signal coming from the four sensors. It compares this pattern to a pre-memorized map. If the signal is strongest on the left sensor, it knows it needs to turn left to align with the beam.
The "Ruler" Test (Phase): As it gets closer, it switches to the ultra-precise distance measurement. It calculates exactly how many seconds it has left before it flies past the beam.

5. The Results: Hitting the Bullseye

The researchers tested this in real-world flights.

Accuracy: They could predict exactly where the tiny "laser beam" was, with an error of only about 15 inches (37 cm).
Timing: They could predict when to enter the beam with an error of less than two-tenths of a second (186 ms).
Reliability: By using four sensors instead of one, they drastically reduced the times the system got confused or lost the signal due to the drone's own body blocking the view.

Summary

In short, this paper describes a way for flying drones to find a tiny, invisible "high-speed data tunnel" in the sky. They do this by using a team of four precise sensors that help each other see through the drone's own body, allowing them to fly through the tunnel and download huge files in a fraction of a second without crashing or missing the connection.

Technical Summary: Leveraging Space-Time Synchronization for Ultra-Spot Detection in mmWave/THz UAV-to-UAV Communications

Problem Statement
In UAV-to-UAV communications utilizing millimeter-wave (mmWave) and Terahertz (THz) bands, data transmission relies on "ultra-spots"—spatially confined, ultra-narrow beams ranging from centimeters to a few meters in width. Effective communication requires precise beam alignment between transmitter and receiver. However, the inherent mobility of UAVs, combined with environmental factors like wind and the narrow beamwidth of mmWave/THz signals, makes maintaining this alignment highly challenging. Traditional positioning systems like GNSS lack the necessary precision (meter-level error) for ultra-spot detection, while Real-Time Kinematic (RTK) GNSS, though more accurate, incurs additional costs and relies on base stations. Furthermore, existing positioning technologies struggle in indoor environments or when UAVs are in close proximity where body shadowing obstructs signals. The core problem addressed is the need for a high-precision, robust method to detect the location and timing of dynamic ultra-spots during flight to enable optimal path planning and beam steering without relying solely on expensive or obstructed positioning systems.

Methodology
The paper proposes a space-time synchronization technique utilizing multiple Wireless Two-way Interferometry (Wi-Wi) devices, termed "multi-Wi-Wi," mounted on UAVs. The system operates on a 920 MHz carrier frequency with a 200 kHz bandwidth, providing nanosecond-level time synchronization and millimeter-level distance measurement capabilities.

The methodology consists of three primary components:

Spatial Diversity via Multi-Wi-Wi: To mitigate packet loss and signal degradation caused by the UAV's own body (shadowing), the system deploys four Wi-Wi slave devices at distinct positions on the receiving UAV (UAV-B), while the transmitting UAV (UAV-A) acts as a master. This configuration leverages spatial diversity; if one device is obstructed, others may maintain a clear line-of-sight, ensuring continuous data flow.
RSSI-Based Pose and Proximity Detection: At longer distances (e.g., >25 m), the system analyzes the Received Signal Strength Indicator (RSSI) "footprint" from the four slave devices. By comparing the real-time RSSI ranking and magnitude against pre-calibrated reference patterns, UAV-B can detect deviations in its orientation and adjust its flight path to align with the ultra-spot.
Phase-Variation-Based Distance and Timing Estimation: As UAV-B approaches the ultra-spot (within ~25 m), the system switches to using carrier phase variation data from the Wi-Wi devices to calculate inter-UAV distance with sub-centimeter accuracy. This data, combined with the initial distance (established via RTK-GNSS before flight), allows the algorithm to estimate the exact time of entry ( $t_{ent}$ ) and exit ( $t_{exit}$ ) from the ultra-spot. The paper presents algorithms for two scenarios: (1) UAV-A hovering while UAV-B flies past, and (2) both UAVs flying in opposite directions.

Key Contributions

Multi-Wi-Wi Architecture: The paper introduces the novel concept of deploying multiple Wi-Wi devices on a single UAV platform to achieve high-precision space-time synchronization, a configuration not previously explored for ultra-spot detection.
Hybrid Detection Algorithm: It develops an algorithm that fuses RSSI magnitude analysis for long-range pose correction with phase-variation analysis for short-range distance and timing estimation.
Spatial Diversity Proof: The study provides theoretical analysis and experimental validation demonstrating that multi-Wi-Wi configurations significantly reduce packet loss and improve reliability by mitigating the shadowing effects of the UAV body.
Flight Route Optimization: For the first time, the work addresses the optimization of UAV flight routes to maximize ultra-spot utilization, enabling UAVs to predict and align with ultra-narrow communication zones.

Experimental Results
The proposed method was validated through real-world UAV-to-UAV communication experiments and anechoic chamber tests.

Spatial Diversity: Anechoic chamber tests confirmed that body obstructions cause significant RSSI and phase instability in single-device setups. The multi-Wi-Wi approach successfully maintained stable data by selecting the most reliable device at any given moment, effectively doubling detection stability.
Distance and Timing Accuracy: In flight experiments, the system achieved an ultra-spot location estimation error of 37.16 cm. This corresponds to a temporal prediction error of 186 ms for ultra-spot entry (assuming a UAV speed of 2 m/s).
Comparison with Other Methods: The Wi-Wi-based approach demonstrated an entry-time error of 54 ms in a static-hover scenario, outperforming traditional GNSS and showing comparable performance to RTK-GNSS (78 ms error) in that specific scenario. However, in dynamic counter-directional flight scenarios, the error increased to 518 ms, highlighting current limitations in complex relative motion.
Likelihood Estimation: The joint use of RSSI and phase data from multiple devices improved the detection likelihood of the ultra-spot position by 3.5 times compared to single-device RSSI thresholding, reducing the deviation from the ground truth from ~7.5 m to within the width of the ultra-spot.

Significance and Claims
The paper claims that this work establishes a comprehensive and fundamentally new approach to high-accuracy sensing in UAV-borne ultra-spot detection systems. It demonstrates the feasibility of predicting the spatial position of mmWave beams in advance from a flying UAV without relying exclusively on expensive base stations or obstructed sensors.

The authors position this research as a critical enabler for future mmWave and THz UAV-to-UAV communications, particularly for scenarios requiring the rapid transfer of large data volumes (gigabytes) in brief timeframes (seconds or less). By allowing UAVs to adaptively adjust their flight direction, altitude, and velocity to align with ultra-narrow communication zones, the proposed method enhances communication efficiency and energy conservation. The authors modestly note that while the method is effective for static or simple dynamic scenarios, further algorithmic refinement is needed to address the higher error rates observed in complex, counter-directional flight paths.