Specific Beamforming for Multi-UAV Networks: A Dual Identity-based ISAC Approach

TL;DR

This paper introduces a dual identity-based ISAC approach for multi-UAV networks that enables fast, accurate beamforming by dynamically associating physical and digital identities, improving alignment and communication performance.

Contribution

It proposes the first dual identity association method for beamforming in UAV networks, combining feature similarity and Kalman filtering for dynamic, precise beam alignment.

Findings

Significantly improves association accuracy over conventional methods

Enhances beamforming speed and precision in dynamic UAV networks

Achieves better communication performance through novel identity tracking

Abstract

Beam alignment is essential to compensate for the high path loss in the millimeter-wave (mmWave) Unmanned Aerial Vehicle (UAV) network. The integrated sensing and communication (ISAC) technology has been envisioned as a promising solution to enable efficient beam alignment in the dynamic UAV network. However, since the digital identity (D-ID) is not contained in the reflected echoes, the conventional ISAC solution has to either periodically feed back the D-ID to distinguish beams for multi-UAVs or suffer the beam errors induced by the separation of D-ID and physical identity (P-ID). This paper presents a novel dual identity association (DIA)-based ISAC approach, the first solution that enables specific, fast, and accurate beamforming towards multiple UAVs. In particular, the P-IDs extracted from echo signals are distinguished dynamically by calculating the feature similarity according to their prevalence, and thus the DIA is accurately achieved. We also present the extended Kalman filtering scheme to track and predict P-IDs, and the specific beam is thereby effectively aligned toward the intended UAVs in dynamic networks. Numerical results show that the proposed DIA-based ISAC solution significantly outperforms the conventional methods in association accuracy and communication performance.