모병문 Byeong-mun Mo , 김도현 Do-hyun Kim , 오창근 Chang-geun Oh

DOI:10.12673/jant.2026.30.4.244 JANT Vol.30(No.4) 244-249, 2026

---

This study compared radar vectoring and time-based flow management (TBFM) in terms of air traffic control efficiency and human factors. A simulator experiment with five participants measured radio communications, total landing completion time, NASA-TLX workload, and SART situation awareness, while a survey of 49 respondents, including active controllers and ATC course graduates, examined perceptions of both methods. Under TBFM, communications decreased by 39.2%, landing completion time decreased by 10.4% from 42.2 to 37.8 minutes, NASA-TLX workload was significantly lower, and SART scores were consistently higher. The survey showed that TBFM was favored for traffic-flow predictability and stability, whereas radar vectoring was favored for direct control and unexpected-situation response despite higher mental demand. These findings indicate complementary use: TBFM-centered operation under normal conditions and radar vectoring for off-nominal situations.

이소망 Somang Lee , 은연주 Yeonju Eun , 전대근 Daekeun Jeon

DOI:10.12673/jant.2026.30.4.250 JANT Vol.30(No.4) 250-259, 2026

---

Air traffic flow management (ATFM) is an essential service for balancing air traffic demand with available capacity. In Korea, ATFM requires a modeling approach that considers both requests from neighboring flight information regions and domestic air traffic control procedures. This paper presents a traffic management initiative (TMI) modeling algorithm implemented in the Collaborative Tactical Flow Management System (CTFMS). The proposed heuristic algorithm assigns calculated time-over (CTO) values using estimated time-over (ETO) information and capacity constraints while supporting functions including fixed CTO assignment, order preservation, flow-rate constraints, multiple TMIs, and runway separation. The algorithm was evaluated using a TMI request scenario from a neighboring flight information region and simplified scenarios for each function. The results demonstrate that the algorithm successfully assigns CTOs while satisfying capacity constraints and supports the proposed operational functions. The modeling approach supports practical implementation of ATFM in Korea and provides a foundation for future research considering operational uncertainty.

김진욱 Jin-uk Kim

DOI:10.12673/jant.2026.30.4.260 JANT Vol.30(No.4) 260-266, 2026

---

This study analyzes residual range errors after atmospheric propagation correction using flight-test data from an airborne maritime target detection radar. A representative atmospheric refractivity was derived from marine buoy observations and applied using a Doerry-based correction method. The correction reduced the range-dependent error, but a residual linear trend remained in both test datasets. Linear regression showed that this trend was more pronounced in Test #1 than in Test #2. After removing the linear range-dependent component, the residual errors were separated into low- and high-frequency components using a moving-average filter. The low-frequency standard deviation remained about 1.5 m in both tests, and its relative variance ratio remained above 60% for the 10-sample window. A sensitivity analysis using 5, 10, and 20 samples showed the same overall tendency. These results indicate that slowly varying residual components remain even after propagation correction and linear detrending, and may reflect residual propagation-model error, target scattering characteristics, and other measurement effects.

조환희 Hwan-heui Cho , 박준상 Jun-sang Park , 염우성 Woo-sung Yeom , 우창욱 Chang-wuk Woo , 정동우 Dong-woo Jung

DOI:10.12673/jant.2026.30.4.267 JANT Vol.30(No.4) 267-276, 2026

---

Recently, Manned-Unmanned Teaming has garnered significant attention as a core operational concept to enhance the survivability of manned systems and increase operational efficiency. In this study, a UAV flight control system designed for light armed helicopter(LAH) MUM-T operations was integrated into a comprehensive verification environment, and scenario-based validation was performed based on requirements derived from collaboration with the military. The established environment consists of battlefield equipment, ground and airborne control stations, and simulators for manned and multiple unmanned aircraft, linked in real-time via a STANAG 4586-based communication interface. The system's functionality was verified through operational scenarios involving with automatic takeoff, control transfer, and missions of target tracking, schedule re-planning and automatic landing. The results demonstrated that the UAVs performed the missions stably following control transfer, and that collision avoidance and target tracking were executed successfully through collaborative flights between helicopters and UAVs. This research validates the integrated operational feasibility of the proposed UAV flight control system within a simulated environment.

김진욱 Jin-uk Kim

DOI:10.12673/jant.2026.30.4.277 JANT Vol.30(No.4) 277-284, 2026

---

An active electronically scanned array (AESA) radar can operate despite failures of some transmit/receive modules (TRMs), but its residual detection performance depends on the beamforming weights, failure locations, and number of failed TRMs. The conventional count-based TRM failure rate assumes equal contribution from all TRMs and therefore cannot consistently represent signal-to-noise ratio (SNR) loss in arrays with nonuniform amplitude weighting. This paper proposes an SNR-equivalent weighted TRM failure rate that combines amplitude- and power-weighted failure rates to reflect the beamforming weights and failure distribution. Simulations of a 16×16 planar array with uniform, Taylor, Chebyshev, and Gaussian weighting show that the conventional metric cannot distinguish different SNR losses and detection-range degradations at the same failure rate. The proposed metric consistently represents these differences and enables quantitative evaluation of the residual probability of detection and detection range, providing a practical measure of AESA radar performance under TRM failures.

박상현 Sanghyeon Park , 손표웅 Pyo-woong Son

DOI:10.12673/jant.2026.30.4.285 JANT Vol.30(No.4) 285-290, 2026

---

Global navigation satellite system (GNSS) radio frequency interference (RFI) can affect ship and aircraft navigation and surveillance. This paper presents a real-time GNSS RFI candidate monitoring system using automatic identification system (AIS) and automatic dependent surveillance-broadcast (ADS-B) position reports. Each new report is screened with an interacting multiple model Kalman filter for AIS and a constant-acceleration Kalman filter for ADS-B. Accumulated candidates are grouped through daily spatiotemporal clustering, and cluster locations and pre- and post-transition tracks are provided through a unified replay interface. Compared with existing services, the system additionally provides AIS/ADS-B integration and multi-object clustering/replay. From 1 to 7 July 2026, it processed 8.13 million AIS and 27.62 million ADS-B reports, identifying 3,608 and 3,314 candidates and 75 and 209 clusters, respectively. Amman and St. Petersburg cases involving 69 vessels and 12 aircraft demonstrate real-time screening, daily clustering, and case replay.

추헌민 Hunmin Choo , 박성현 Seonghyeon Park , 정윤호 Yunho Jung

DOI:10.12673/jant.2026.30.4.291 JANT Vol.30(No.4) 291-298, 2026

---

This paper proposes an object classification system for autonomous delivery robots based on a 60 GHz frequency modulated continuous wave (FMCW) radar sensor. The proposed system extracts the micro-Doppler features of detected objects through radar sensor signal processing and performs deep learning-based inference to classify the objects into pedestrians, cyclists, static obstacles, and delivery robots. To minimize the memory requirements and computational complexity of the deep learning model, a binarized neural network (BNN) was introduced, and the evaluation results confirmed an excellent classification accuracy of 95.23%. The system was implemented on a Raspberry-Pi 5 and a field programmable gate array (FPGA) board, and a BNN accelerator was implemented on the FPGA for real-time inference. The implementation results achieved a short inference time of 6.5 ms, confirming that efficient operation is possible in an embedded environment.

임성준 Seoung-jun Lim , 최원혁 Won-hyuck Choi

DOI:10.12673/jant.2026.30.4.299 JANT Vol.30(No.4) 299-306, 2026

---

Status-message delay in low-altitude unmanned aircraft system traffic management (UTM) is safety-critical for unmanned aerial vehicles (UAVs), yet handovers, traffic, and channel variations obscure its causes. Using ns-3 Long-Term Evolution/Evolved Packet Core (LTE/EPC) simulations without competing traffic, we combined 1 to 5 base stations with 10, 20, 30, and 40 UAVs over 10 seeds (200 runs), with 120 B, 5 Hz uplink, path-loss exponent 2.6, hysteresis 3 dB, and time-to-trigger (TTT) 256 ms. Handovers per UAV rose from 0 to 3.33; mean delay grew from 29.4 to 137.6 ms (about 368%), p95 delay from 40.0 to 514.1 ms (about 1,185%), and packet delivery ratio (PDR) fell from 1.000 to 0.981. Reference signal received power (RSRP) and signal-to-interference-plus-noise ratio (SINR) held near -104 dBm and 18.7 dB; regression indicated about 27 ms per handover, implicating handover over signal quality. The trend persisted under path-loss exponent 3.0, and lowering hysteresis from 6 to 1 dB cut p95 delay by up to 45%. UTM design should treat handover-induced tail delay alongside mean delay.

임성준 Seoung-jun Lim , 최원혁 Won-hyuck Choi

DOI:10.12673/jant.2026.30.4.307 JANT Vol.30(No.4) 307-314, 2026

---

Low-altitude unmanned aircraft system traffic management (UTM) requires timely delivery of status information from unmanned aerial vehicles (UAVs). This study conducts controlled ns-3 LTE/EPC simulations with a fixed cell structure and combinations of three background-load levels and two burst-traffic conditions. Under heavy background and burst traffic with 40 UAVs, the 95th-percentile delay (p95) reached approximately 1.13 s, while the packet delivery ratio (PDR) remained 0.956. The burst-induced increase in p95 delay also tended to grow with background load. A delay-based information freshness metric derived from the message generation interval and received-packet delay distribution reached approximately 1.33 s under the worst condition. Delay-threshold exceedance probabilities were 12.7% at 500 ms and 7.5% at 1 s. Although these metrics are proxies rather than direct measurements of time-average or peak age of information, they show that a high PDR can conceal degraded information timeliness. Low-altitude UTM communication should therefore be evaluated using tail delay and delay-based freshness metrics together with conventional delivery metrics.

이다인 Da-in Lee , 노윤희 Yun-hee Ro , 김동영 Dong-young Kim , 황치호 Chi-ho Hwang , 권영갑 Younggap Kwon

DOI:10.12673/jant.2026.30.4.315 JANT Vol.30(No.4) 315-321, 2026

---

This paper proposes methods to optimize the performance of frame synchronization pattern detection and cyclic redundancy check (CRC) verification, which are performed repeatedly in telemetry data processing program. To optimize frame synchronization pattern detection, a preceding 2-byte priority validation algorithm is applied to create a 2-byte candidate shift table and prioritizes validating the first 2 bytes of the data stream buffer is applied to reduce the repetitive comparisons. To optimize CRC verification, the process was optimized by applying a direct-16-bit operation structure that eliminates the process of rearranging and initializing data into byte buffers and then directly refer to words to be checked when using a 16 bits word-based frame structure. It was confirmed through a benchmark program that the proposed algorithm reduced the data processing time. This reduced the elapsed time for data processing and improved the efficiency of the data processing.