By 2030, projected UAV operations may exceed one million concurrent flights in urban airspace, yet traditional fixed-distance separation methods fail to accommodate heterogeneous platforms. This paper introduces a three-tier hierarchical dynamic separation framework adapting minima across strategic (30–80 m category-specific baselines), pre-tactical (0.7–1.8× encounter-dependent scaling), and tactical (real-time 3D decomposition) timescales. Monte Carlo simulation across 100,000+ flight hours demonstrates 47% collision rate reduction versus fixed 30 m separation (0.008 vs. 0.015 per 1000 h, p < 0.001), 50% airspace utilization increase (18.4 vs. 12.3 UAVs/km³), 44% flight time penalty decrease (8.5% vs. 15.2%), and 99.97% ICAO-compliant TLS achievement (≤10⁻⁷ per flight hour) with real-time performance (78.5 ms for 20 UAVs). The framework provides an immediately deployable foundation for heterogeneous UAV traffic management.

Cooperative localization for multi-Unmanned Aerial Vehicle (UAV) systems in GPS-degraded environments is often compromised by ideal-communication or uniform-quality assumptions. This paper proposes Quality-Aware Distributed State Estimation (QA-DSE), which combines three operational quality factors—freshness (Age of Information), accuracy (covariance trace), and link reliability (packet loss and channel noise)—into a single multiplicative score $q_{ij}$, modulated by a bounded history-consistency factor based on velocity-propagated self-trajectory continuity. A dual-constraint AND-gate on AoI and covariance trace excludes jointly degraded neighbors, while admitted neighbors are fused through a quality-squared information-matrix update under a stated bounded residual cross-correlation assumption, with an adaptive Covariance-Intersection fallback when the assumption is stressed. Under explicit observability, bounded-noise, bounded-quality, joint-connectivity, and bounded residual cross-correlation assumptions, we establish mean-square bounded error, exponential convergence at a rate inherited from the Kalman update operator, $O\left(n^3+nm\right)$ per-step complexity, Bounded-Input Bounded-Output (BIBO) stability, soft attenuation of single-axis faults (Theorem 4), and hard exclusion under joint AoI–covariance violation (Theorem 5). Under a Ultra-Wideband (UWB)-style cooperative-observation model, Monte Carlo experiments across five scenarios show 74.08–74.24% position- Root Mean Square Error (RMSE) reductions over Covariance Intersection, with the relative advantage held within 73.04–74.24% as the fleet scales from 3 to 50 UAVs; QA-DSE remains within 8.1% of an idealized no-cooperation single-vehicle Kalman filter, demonstrating graceful degradation rather than improvement above that floor. Per-step Central Processing Unit (CPU) time scales from 0.09 ms (5 UAVs) to 0.31 ms (50 UAVs); embedded validation is left to future work.

The increasing complexity of general aviation flight training poses significant challenges to tower controllers, who must reconcile safety separation standards with training efficiency under diverse and often unexpected conflict scenarios. This study proposes a BERT-BiLSTM-XGBoost multimodal feature fusion ensemble learning model that automatically matches optimal allocation schemes based on predicted conflict characteristics, thereby providing real-time decision support for tower controllers. The framework operates in three stages. First, a BERT model extracts high-level semantic features from textual data—including aircraft status descriptions, controller characteristics, pilot profiles, and airspace constraints—by leveraging bidirectional contextual encoding. Second, a Bidirectional Long Short-Term Memory (BiLSTM) network captures temporal dynamic features from sequential aircraft operation data. Third, the extracted semantic and temporal feature vectors are fused and fed into an XGBoost classifier, which performs the final allocation scheme matching. The XGBoost algorithm additionally provides feature importance scores, supporting the traceability requirements of tower control decision-making. Experimental results demonstrate that the proposed model achieves a matching accuracy of 97.2%, substantially outperforming baseline methods including Naive Bayes, linear SVM, and LightGBM. Moreover, the model reduces controller decision-making time by 75–80% in simulated tower control tests, validating its practical utility as an effective decision-support tool for general aviation tower operations.