PromptSan-Suffix: ISAC in Low-Altitude Security
- PromptSan-Suffix is an integrated sensing and communication framework that transforms cellular networks into distributed platforms for continuous low-altitude security.
- It leverages advanced ISAC techniques such as MIMO beamforming, spatiotemporal fusion, and edge-cloud processing to enhance target detection and identity verification.
- The framework enables end-to-end situational awareness, collaborative deconfliction, and dynamic trust authentication for secure, urban low-altitude operations.
Integrated sensing and communication for low-altitude security denotes the use of dual-functional wireless infrastructure to provide continuous wide-area sensing, intelligent cognition, collaborative decision support, and dynamic trust authentication for low-altitude, slow-speed, and small-size aircraft operating in dense, heterogeneous, and obstructed low-altitude airspace. In this formulation, cellular networks are upgraded from pure communication infrastructure into distributed sensing-and-communication systems, with base stations acting as sensing nodes while preserving communication services. The framework has been presented as the first review dedicated to ISAC for low-altitude security, and it sits within a broader 5G-Advanced and 6G trajectory in which shared spectrum, hardware, and signal processing are used to couple connectivity with environmental perception (Ren, 20 Jan 2026, Zhang et al., 9 Apr 2025).
1. Low-altitude security as an ISAC problem
The motivating setting is the rapidly growing low-altitude economy, in which low-altitude, slow-speed, small-size aircraft operate in cluttered urban and suburban geometries. The resulting low-altitude security problem is not limited to target detection. It also includes persistent sensing blind spots, ambiguous target identity and intent, weak conflict resolution capability, and vulnerable wireless links and authentication. The review on low-altitude security isolates four core bottlenecks: full-domain continuous sensing failure, lack of target cognition and intention understanding, poor high-dynamic conflict resolution and avoidance, and vulnerable wireless links and identity authentication (Ren, 20 Jan 2026).
These bottlenecks arise because terrain and building occlusions, adverse weather, and dense heterogeneous traffic degrade radar and optical coverage, while position and velocity alone are insufficient for discriminating UAVs from birds or balloons or for inferring behaviors such as loitering, reconnaissance, and formation flight. In addition, dense static obstacles and mixed dynamic traffic in urban canyons exceed the range and robustness of onboard sensors, and conventional broadcast identifiers remain weakly bound to physical entities. The result is a governance gap: existing radar, electro-optical, and regulatory systems do not provide a unified, real-time, wide-area picture of low-altitude airspace or a reliable mechanism for linking digital identity, physical behavior, and operational intent (Ren, 20 Jan 2026).
Within the broader ISAC literature, this application fits a larger shift from isolated link design toward system- and network-level sensing. Papers about ISAC describe the field as moving from coexistence and resource sharing toward true dual-functional operation, multi-cell collaboration, and cross-layer integration, which is precisely the regime required by low-altitude security rather than by isolated point sensing (Zhang et al., 9 Apr 2025, Wymeersch et al., 16 May 2025).
2. Cellular-ISAC architecture for low-altitude security
The low-altitude security architecture is organized as a cellular-ISAC technical system. At the infrastructure layer, macro base stations, micro base stations, and future millimeter-wave base stations form a dense, planned network at urban commanding heights. With software-defined radio upgrades and enhanced antenna arrays, each base station can emit optimized ISAC waveforms, or reuse communication signals, while receiving target echoes within licensed mobile spectrum and power limits. At the signal and sensing layer, ISAC waveforms with strong auto/cross-correlation, together with MIMO beamforming and robust signal processing, extract range, velocity, and angle in clutter and interference. At the fusion and computation layer, low-latency mobile backhaul carries local detections to edge or cloud nodes for spatiotemporal registration, data association, and trajectory fusion, producing a unified 4D situational map containing position, velocity, heading, and timestamp, with centimeter-level accuracy for cooperative and non-cooperative targets. Communication and distribution then deliver this situational map through 5G-A or 6G capabilities such as low latency, high reliability, and network slicing to aircraft, edge nodes, and central platforms. Decision and control close the loop through autonomous conflict prediction, route replanning, distributed deconfliction, and macro airspace adjustment. A security and authentication layer continuously compares claimed digital identities with ISAC-derived physical behavioral fingerprints such as precise position, 3D trajectory, velocity profile, and angle of arrival (Ren, 20 Jan 2026).
A notable feature of this architecture is that sensing, communication, and control are coupled in three loops that close a perception-cognition-action cycle. This is more than a colocated radar-and-network deployment. It is a networked sensing fabric in which the base station becomes both a radio access point and a distributed sensor.
This architectural direction is consistent with several adjacent ISAC developments. Coordinated cellular multistatic radar has been proposed precisely to avoid the severe self-interference that arises when a monostatic base station simultaneously transmits the ISAC waveform and receives the echo; the multistatic formulation spatially separates transmission and reception and jointly optimizes transmit and receive beamforming under communication and sensing QoS constraints (Xu et al., 2023). At the network level, ISAC has also been formalized as a routing-and-sensing resource allocation problem with a convex sensing-throughput region and a piecewise linear Pareto boundary in general topologies, which suggests that low-altitude security deployments will inherit explicit coverage-versus-capacity trade-offs once they scale beyond local clusters (Andrews et al., 15 Jan 2026).
The low-altitude security review is conceptual and architectural rather than formula-driven. It does not present explicit LaTeX formulas or closed-form mathematical models for sensing, inference, collaboration, authentication, or performance; instead, it specifies the information elements and algorithmic families that instantiate the architecture (Ren, 20 Jan 2026).
3. Workflow, information fusion, and intent inference
The end-to-end workflow couples sensing and communication at every stage. First, base stations collect radar echoes and channel state information. Feature extraction includes micro-Doppler and high-resolution range profiles to capture rotor count, rotor speed, size, and materials, enabling non-cooperative physical fingerprinting of UAVs versus birds or balloons. Second, edge or cloud nodes perform spatiotemporal registration, association, and trajectory fusion across base stations, generating a unified 4D map with single-source consistency rather than fragmented local views. Third, intelligent cognition and intent inference fuse physical fingerprints with cyberspace signaling, such as temporary device IDs and signal-strength history, and with planned information, such as declared flight plans, electronic fences, and approved routes. Temporal deep learning or knowledge graph engines then analyze trajectories, speed changes, dwell patterns, and GIS context to detect abnormal behaviors and classify intent as straying, snooping, or malicious. Fourth, network slicing and low-latency delivery disseminate the situational product to aircraft, edge nodes, and central platforms according to mission latency and reliability requirements. Fifth, collaborative decision-making executes onboard conflict prediction and route replanning, edge-coordinated distributed deconfliction, and central traffic instruction. Sixth, dynamic trust authentication parses remote IDs and access identifiers, compares them with sensed position, trajectory, velocity variation, and angle of arrival, applies consistency rules, decays trust scores on deviations, raises alarms for no-fly zone incursions or identity-model mismatch, and preserves audit trails (Ren, 20 Jan 2026).
This workflow is notable because it treats low-altitude security as a multimodal inference pipeline rather than as a radar tracking task. Physical sensing, network signaling, planning data, and policy constraints are fused in a single operational graph.
Related UAV-oriented ISAC work points in the same direction at a smaller scale. One design uses a single ISAC waveform to parallelize radar sensing and communication interrogation for identification friend or foe, reducing the communication delay metric by up to 50%, and applies EKF fusion of communication-derived location information with radar measurements to improve target sensing accuracy by 24.2% when communication location information and radar sensing information have the same sensing accuracy (Jiang et al., 2022). That result does not solve the wide-area low-altitude governance problem, but it supports the broader claim that tightly time-aligned sensing and communication can improve both target classification latency and state estimation.
4. Core security bottlenecks and corresponding ISAC mechanisms
| Bottleneck | Failure mode | ISAC mechanism |
|---|---|---|
| Full-domain continuous sensing failure | Blind spots, intermittent tracks, fragmented views | Dense cellular sensing nodes, MIMO processing, edge/cloud fusion |
| Lack of target cognition and intention understanding | Identity ambiguity, poor intent inference | Physical fingerprints, signaling fusion, temporal models, knowledge graphs |
| Poor high-dynamic conflict resolution and avoidance | No wide-area low-latency collaborative decision infrastructure | Shared unified map, network slicing, distributed deconfliction |
| Vulnerable wireless links and identity authentication | Spoofing, jamming, hijacking, impersonation | Cross-layer identity-behavior binding, trust scoring, audit trails |
The first bottleneck is coverage continuity. Dense cellular base stations provide a planned sensing geometry that can bridge occlusions and weather-induced discontinuities more effectively than isolated radar or optical systems. The second is cognition. CSI and echo-derived signatures provide non-cooperative fingerprints, but their operational value emerges only after fusion with cyberspace identifiers and planning data. The third is deconfliction. A shared, low-latency situational map allows aircraft, edge nodes, and central platforms to act on a consistent world model rather than on locally inconsistent tracks. The fourth is trust. Authentication is no longer a one-shot credential check; it becomes a continuous comparison between claimed mission, model, and route on one side and sensed physical behavior on the other (Ren, 20 Jan 2026).
This cross-layer binding of protocol identities to physical trajectories aligns with broader 6G cross-layer ISAC thinking, in which sensing, protocol control, resource management, and privacy or security functions are designed together rather than as separate modules (Wymeersch et al., 16 May 2025). In the low-altitude security setting, the practical consequence is that spoofing and hijacking become detectable as behavioral inconsistency rather than only as protocol anomalies.
5. Relation to multistatic, networked, RIS-assisted, and multi-beam ISAC
Low-altitude security draws on several technical strands of the wider ISAC literature. Cooperative passive sensing in mobile systems addresses asynchronous multi-base-station reception with a non-line-of-sight and line-of-sight signal cross-correlation method for mitigating carrier frequency offset and time offset, and symbol-level multi-base-station fusion is reported to achieve an order of magnitude higher sensing accuracy than single-base-station passive sensing (Wei et al., 2024). For low-altitude surveillance, this is directly relevant to blind-spot reduction and multi-base-station fusion quality.
Networked distributed sensing offers another relevant pattern. A networked ISAC framework with multiple users acting as sensors replaces centralized raw-data fusion with a two-step distributed cooperative sensing algorithm based on low-dimensional intermediate estimate exchange, block-wise sparsity, and interference cancellation (Li et al., 2024). This suggests a plausible distributed implementation path for low-altitude edge fusion in which robustness is increased by reducing reliance on a single leader node.
RIS-assisted ISAC contributes yet another set of mechanisms. RISs provide passive beamforming gain, extra spatial degrees of freedom, virtual line-of-sight links, interference suppression, and improved detection probability, localization, mapping accuracy, and coverage (Liu et al., 2022). A related tutorial argues that joint sensing-and-communications designs are most beneficial when the sensing and communication channels are coupled, and that RISs provide a means to control this coupling through subspace expansion and subspace rotation (Chepuri et al., 2022). Hybrid STAR-RIS extends this to full-space coverage with passive reflective elements for local users and targets and low-power active transmissive elements for distant users and targets, but it also introduces explicit energy-versus-noise and cross-side interference trade-offs (Yigit et al., 2024). In dense urban low-altitude geometries, these results make RIS-aided blockage bypassing and virtual LoS creation technically attractive, although the low-altitude security review treats them as indicative rather than foundational (Ren, 20 Jan 2026).
Multi-beam ISAC is also germane. In mmWave and higher-frequency systems, multi-beam operation separates communication and sensing in beamspace by maintaining time-invariant communication beams and time-varying sensing beams, while joint beamforming stabilizes gain and suppresses sidelobes (Zhuo et al., 2024). This beamspace separation is closely aligned with the need to maintain communication services while continuously scanning or tracking in low-altitude corridors.
6. Open problems, research directions, and significance
The low-altitude security review identifies four major open problem classes. The first is sensing performance in complex environments: extremely low signal-to-clutter ratios for low-altitude, slow-speed, small-size targets under occlusions, terrain clutter, interference, and severe weather; cumulative registration errors in multi-base-station fusion; heterogeneous data consistency validation; and passive detection of non-cooperative targets with stable feature and motion fingerprint extraction. The second is the generalization of intelligent cognition and intent understanding: high heterogeneity across consumer UAVs, industrial aircraft, manned aircraft, and birds; adversarial camouflage; narrow-domain AI models that fail in dynamic low-altitude contexts; and the tension between data privacy and cross-domain sharing. The third is real-time guarantees for collaborative decision-making: millisecond-level latency requirements for detection, replanning, and command delivery; balancing global optimality and local autonomy in dense swarms; and maintaining consistency and safety under limited communication resources. The fourth is robust authentication under adversarial environments: spoofed physical fingerprints, tampered GNSS data, cross-operator and cross-airspace interoperability, anti-counterfeiting physical-layer features, non-repudiation, and lightweight distributed cross-domain trust networks (Ren, 20 Jan 2026).
These problems indicate that low-altitude security is not merely an application vertical of existing ISAC. It is a stress test for networked sensing, multimodal inference, URLLC-grade coordination, and physical-layer security. A plausible implication is that future progress will depend on moving from isolated PHY optimization toward task-level and cross-layer formulations. That direction is visible elsewhere in the literature: planning-oriented ISAC explicitly links transmit power, sensing uncertainty, and safe navigable space for autonomous driving (Jin et al., 27 Oct 2025), while a cross-layer industrial-academic perspective frames ISAC as a programmable and context-aware 6G platform spanning physical layer, hardware, protocol functions, APIs, and privacy or security controls (Wymeersch et al., 16 May 2025).
For industry, the low-altitude security architecture promises scalable, compliant deployment by reusing cellular infrastructure and licensed spectrum, offering BVLOS awareness, reduced infrastructure cost versus standalone radar, and end-to-end governance from detection through intent inference and trusted control. For academia, it crystallizes a research agenda around ISAC waveform and beamforming design, multi-base-station fusion, multimodal intent inference under privacy constraints, low-latency collaborative control, and verifiable physical-layer authentication for future perceptive 6G networks (Ren, 20 Jan 2026).
