Abstract
Integrated sensing and communications (ISAC) enables the sharing of hardware and spectrum resources, facilitating efficient data transmission and environmental sensing. This fusion is critical for unmanned aerial vehicle (UAV) swarms to enhance performance and flexibility. This paper proposes an ISAC solution based on pseudo-random time-division multiplexing (TDM)-multiple input multiple output (MIMO) millimeter-wave (mmWave) frequency modulated continuous wave (FMCW). A novel ISAC chirp waveform modulates data in both the delay domain and complex amplitude while maintaining high-precision sensing. To overcome TDM-MIMO velocity ambiguity, we employ pseudo-random antenna selection and compressed sensing. Furthermore, a chirp-division multiple access (Chirp-DMA) scheme is proposed for interference-free multi-user transmission. Simulation results demonstrate that the proposed scheme achieves robust ISAC performance in dynamic UAV flight scenarios, outperforming existing baselines like mmWave-LoRadar.
Introduction: The Velocity Wall in UAV ISAC
Unmanned Aerial Vehicle (UAV) swarms require simultaneous sensing for navigation and communication for coordination. Traditionally, these functions use separate hardware, increasing weight and power consumption—a significant constraint for agile platforms.
While ISAC aims to merge these functions, current solutions face a "velocity wall." Most mmWave radars use TDM-MIMO, where antennas transmit sequentially. This creates a "waiting period" that reduces the maximum unambiguous velocity. For fast-moving UAVs, this limitation is a dealbreaker. Additionally, managing interference between multiple UAVs in a swarm requires sophisticated time-frequency resource coordination.
Methodology
1. Dual-Domain Modulation
The core innovation is a modified FMCW waveform that embeds data into the radar chirp using two methods:
- Delay Domain: Shifting the start time of the chirp, which the receiver interprets as an "equivalent distance."
- Complex Amplitude: Utilizing Differential Phase Shift Keying (DPSK) on the signal's complex amplitude.
2. Pseudo-Random TDM-MIMO and Compressed Sensing
To solve velocity ambiguity, antennas fire in a pseudo-random order guided by a Latin Matrix rather than a fixed sequence. This randomizes phase errors caused by motion. The receiver then applies the Orthogonal Matching Pursuit (OMP) algorithm for compressed sensing. Because the sampling is randomized, velocity estimation becomes a sparse recovery problem, allowing the system to track speeds significantly higher than traditional TDM-MIMO limits.
3. Chirp-Division Multiple Access (Chirp-DMA)
The system utilizes time-staggered chirps to create orthogonal "Resource Blocks" (RBs) in the time-frequency plane. This allows multiple antennas or multiple UAVs to operate simultaneously without mutual interference, enabling dynamic resource allocation.
4. J-SPEDD Algorithm
The Joint Sensing Parameter Estimation and Data Demodulation (J-SPEDD) scheme employs an Extended Kalman Filter (EKF) to track physical movement. By predicting the "true" distance and velocity, the receiver can subtract radar components from the signal to cleanly extract communication bits.
Results and Performance
The proposed framework was evaluated using high-fidelity numerical simulations of a 3-UAV swarm in an urban environment with Rayleigh-distributed clutter and ground reflections.
- Communication Performance: The system achieved bit rates between 500 kbps and 700 kbps, significantly outperforming the mmWave-LoRadar baseline in Symbol Error Rate (SER), particularly at low Signal-to-Noise Ratios (SNR).
- Sensing Accuracy: The sensing "hit rate" reached nearly 100% at SNRs above -20 dB. Crucially, the maximum unambiguous velocity was expanded by a factor equal to the number of transmit antennas ($L_{tx}$).
- Urban Robustness: Despite performance degradation from urban clutter, the EKF-based tracking remained stable, proving the system's viability in complex real-world environments.
Conclusion
This research presents a hardware-friendly path for integrating sensing and communication on UAV platforms. By treating data as a time delay and utilizing pseudo-randomization to overcome physical sensing limits, the framework provides a robust solution for swarm coordination. Future work will address the computational complexity of running compressed sensing in real-time for larger swarms and developing self-compensation algorithms for hardware phase noise.
Summary updated