Earticle

최근 대규모 개인정보 유출 사고로 인해 Credential Stuffing 공격이 급증하고 있다. 기존의 TOTP(Time-based One-Time Password) 기반 이중 인증 기법은 비밀번호 재사용 공격을 완화하는 데 효과적이지만, HMAC-SHA 기반 구조는 장기적으로 양자 컴퓨팅 환경에서 보안 강도가 저하될 수 있다. 본 논문에서는 격자 기반 및 해시 기반 양자 내성 암호 구조를 적용한 PQ-TOTP(Post-Quantum TOTP) 인증 기법을 제안한다. 제안 기법은 디바이스 바인딩과 세션 난수 결합 메커니즘을 통해 Credential Stuffing 공격의 성공 확률을 이론적으로 감소시킨다. 보안성 분석 결과, 제안 기법은 기존 TOTP 대비 양자 공격 저항성과 재사용 공격 방어 측면에서 향상된 보안 수준을 제공함을 확인하였다.

Credential stuffing attacks have significantly increased due to large-scale credential leaks. Conventional Time-based One-Time Password (TOTP) mechanisms mitigate password reuse attacks; however, HMAC-SHA-based constructions may suffer reduced security strength in the presence of quantum adversaries. This paper proposes a Post-Quantum TOTP (PQ-TOTP) scheme leveraging lattice-based and hash-based post-quantum cryptography. The proposed mechanism incorporates device binding and session randomness to reduce the success probability of credential stuffing attacks. Security analysis demonstrates that the proposed scheme provides enhanced resistance against quantum attacks and replay-based credential reuse compared to conventional TOTP mechanisms.

Credential Stuffing 공격의 2단계 인증에 사용되는 TOTP(Time based One Time Password)는 일회성 비밀번호(OTP)를 시간 기반으로 생성하는 간편하고 널리 사용되는 인증 방식이지만, 유효 시간 내 동일 OTP를 재사용할 수 있어 재사용 공격에 취약하다는 한계가 있다. 따라서 본 논문에서는 이러한 보안 취약점을 보완하기 위해 경량 자료 구조인 Counting Bloom Filter(CBF)를 결합하여 OTP 사용 이력을 추적하여 재사용 여부를 효 율적으로 검출할 수 있는 방법을 제안하였다. 결론적으로 제안한 CBF+TOTP 방식은 기존 TOTP의 재사용 공격 과 같은 보안 취약점을 보완하면서도 경량성과 확장성을 유지할 수 있는 효과적인 인증 기법이다.

Credential stuffing attacks exploit reused or compromised credentials to gain unauthorized access to user accounts. As a countermeasure, Time-Based One-Time Passwords (TOTP) are widely used in two-factor authentication (2FA) systems. However, TOTP remains vulnerable to replay attacks, as the same OTP can be reused within its valid time window. To address this limitation, we propose an enhanced authentication method that combines TOTP with a Counting Bloom Filter (CBF), a lightweight and space-efficient data structure. By recording hashed OTPs in the CBF, the system can efficiently detect and prevent OTP reuse without the need for persistent storage. In conclusion, the proposed CBF+TOTP scheme effectively addresses security vulnerabilities of traditional TOTP, such as susceptibility to replay attacks, while maintaining lightweight operation and scalability.

Trash or garbage is one of the most dangerous health and environmental problems that affect pollution. Pollution affects nature, human life, and wildlife. In this paper, we propose modern solutions for cleaning the environment of trash pollution by enforcing strict action against people who dump trash inappropriately on streets, outside the home, and in unnecessary places. Artificial Intelligence (AI), especially Deep Learning (DL), has been used to automate and solve issues in the world. We availed this as an excellent opportunity to develop a system that identifies trash using a deep convolutional neural network (CNN). This paper proposes a real-time garbage identification system based on a deep CNN architecture with eight distinct classes for the training dataset. After identifying the garbage, the CCTV camera captures a video of the individual placing the trash in the incorrect location and sends an alert notice to the relevant authority.

In this research, a remote control system has been developed and implemented, which combines autonomous obstacle avoidance in real-time with force-reflective tele-operation. A tele-operated mobile robot is controlled by a local two-degrees-of-freedom force-reflective joystick that a human operator holds while he is monitoring the screen. In the system, the force-reflective joystick transforms the relation between a mobile robot and the environment to the operator as a virtual force which is generated in the form of a new collision vector and reflected to the operator. This reflected force makes the tele-operation of a mobile robot safe from collision in an uncertain and obstacle-cluttered remote environment. A mobile robot controlled by a local operator usually takes pictures of remote environments and sends the images back to the operator over the Internet. Because of limitations of communication bandwidth and the narrow view-angles of the camera, the operator cannot observe shadow regions and curved spaces frequently. To overcome this problem, a new form of virtual force is generated along the collision vector according to both distance and approaching velocity between an obstacle and the mobile robot, which is obtained from ultrasonic sensors. This virtual force is transferred back to the two-degrees-of-freedom master joystick over the Internet to enable a human operator to feel the geometrical relation between the mobile robot and the obstacle. It is demonstrated by experiments that this haptic reflection improves the performance of a tele-operated mobile robot significantly.

This paper describes a new digital signal processor (DSP) imbalance measurement system dedicated to real-time vibration analysis on rotating machine. To accomplish real-time analysis, the vibration signals are on-line acquired and processed to analyze the mass imbalance and phase position. This is achieved through the use of FFT and Lissajous diagram. The method followed to analyze the mass imbalance with the chosen hardware and software solutions are described in detail in this paper. Several experimental tests demonstrate the efficiency and accuracy in imbalance analysis performance of the DSP system.

Fall accidents are increasing, and monitoring them using real-time CCTV systems remains challenging. This paper compares the performance of YOLOv11 and RT-DETRv2 models for real-time fall detection. Experimental results show that YOLOv11 outperforms RT-DETRv2 in terms of inference speed, making it more suitable for real-time applications. Unlike earlier studies, we propose feature map-based knowledge distillation during the model training process to improve model performance. The proposed YOLO-based fall detection system transfers intermediate representations from a teacher to a student network and optimises two complementary objectives: spatial alignment via Mean-Squared-Error (MSE) loss and channel-wise distribution alignment via Kullback?Leibler (KL) divergence. Experiments improved the mean Average Precision (mAP) and reduced processing time by 0.8ms. Evaluation on AI-hub abnormal behavior datasets confirmed a 0.02 increase in accuracy and F1-score, demonstrating the effectiveness of the proposed distillation method in real-time environments.

This paper proposes a novel Korean phoneme boundary detection method that can be applied to phoneme-based Korean speech recognition systems. The proposed method employs two time-domain metrics-volatility and bulk metrics-as the foundation for phoneme boundary detection. The input speech signal is divided into blocks of 300 integer samples. For each block, the volatility metric is computed that adds up all the changes between neighboring samples within the block. A bulk is a grouping of consecutive samples with the same sign. For each bulk, two bulk metrics are calculated: bulk size and bulk length. Three dedicated algorithms that utilize both types of metrics are used to detect phoneme boundaries by recognizing vowels, voiced consonants, and voiceless consonants in turn. The experimental results show that the proposed method can significantly reduce the error rate compared to an existing boundary detection method.

The new realtime applications like multimedia and realtime services in a wireless network will be dramatically increased. However, many realtime services of mobile hosts in a cell cannot be continued because of insufficiency of useful channels. Conventional channel assignment approaches didn't properly consider the problem to serve realtime applications in a cell. This paper proposes a new realtime channel assignment algorithm based on time constraint analysis of channel requests. The proposed algorithm dynamically borrows available channels from neighboring cells. It also supports a smooth handoff which continuously serves realtime applications of the mobile hosts.

As application-specific requirements for wireless sensor networks emerge, both real-time and reliable communications become major research challenges in wireless sensor networks due to the many constraints on nodes and wireless links. To support these services, several protocols have been proposed. However, since most of them were designed as well as developed for general purpose applications, it is not recommended that they be directly adapted to applications with special requirements. In this paper, we propose a way to extend the current reliable transport protocol to cover a special real-time service, the (m,k)-firm deadline stream, in wireless sensor networks. While the proposed scheme is basically built on the PSFQ protocol for reliability, some features have been newly developed to support the (m,k)-firm stream efficiently. Finally, simulation results are given to demonstrate the feasibility of the proposed scheme in high traffic and with failed links.

In this research, we study the outage probability for distributed space-time coding-based cooperative (DSTC) systems with amplify-and-forward relaying over Rayleigh fading channels with a high temporal correlation where the direct link between the source and the destination is available. In particular, we derive the upper and lower bounds of the outage probability as well as their corresponding asymptotic expressions. In addition, using only the average channel powers for the source-to-relay and relay-to-destination links, we propose an efficient power allocation scheme between the source and the relay to minimize the asymptotic upper bound of the outage probability. Through a numerical investigation, we verify the analytical expressions as well as the effectiveness of the proposed efficient power allocation. The numerical results show that the lower and upper bounds tightly correspond to the exact outage probability, and the proposed efficient power allocation scheme provides an outage probability similar to that of the optimal power allocation scheme that minimizes the exact outage probability.

Finding a valid timing constraint is one of the most important issues in the real-time monitoring area. To get the valid timing constraint, a developer executes a real-time process and changes the constraint on a regular basis. This is an exhaustive and time-consuming process. To improve this approach, we propose a timing constraint search technique. This technique uses two load models and one condition proposed in this paper to support the developer in determining the valid timing constraint range in an easy and systematic manner.

The increasing reliance on smartphones necessitates secure, accurate user identification methods. This study introduces “digital pheromones”, a novel digital identity creation method based on smartphone user behaviors. Represented by a 56-bit binary number, this identity is derived from quantization and concatenation of 14 time-varying smartphone variables. Motivated by the need for enhanced security and privacy, the study employs statistical analysis to measure user behavior dispersion, showing that wider dispersion variables facilitate faster user uniqueness at lower quantization levels. The results demonstrate 100% uniqueness, ensuring no two users share the same 56-bit identity. This research offers a robust framework for future advancements in smartphone user authentication and identification systems.

New High Throughput Satellite (HTS) systems allow high throughput IP uplinks/contribution at Ka-band frequencies for relatively lower costs when compared to broadcasting satellite uplinks at Ku band. This technology offers an advantage for live video contribution from remote areas, where the terrestrial infrastructure may not be adequate. On the other hand, the Ka-band is more subject to impairments due to rain or bad weather. This paper addresses the target system specification and provides an optimized approach for the transmission of IP-based video flows through HTS commercial services operating at Ka-band frequencies. In particular, the focus of this study is on the service requirements and the propagation analysis that provide a reference architecture to improve the overall link availability. The approach proposed herein leads to the introduction of a new concept of live service contribution using pairs of small satellite antennas and cheap satellite terminals.

Objective To estimate the effects of a relatively protruded head and neck posture on postural balance, in computer based worker. Method Th irty participants, who work with computers for over 6 hrs per day (Group I), and thirty participants,who rarely work with computers (Group II), were enrolled. Th e head and neck posture was measured by estimating angles A and B. A being the angle between the tragus of the ear, the lateral canthus of the eye, and horizontal line and B the angle between the C7 spinous process, the tragus of the ear, and the horizontal line. Th e severity of head protrusion with neck extension was assessed by the subtraction of angle A from angle B. We also measured the center of gravity (COG) and postural balance by using computerized dynamic posturography to determine the eff ect of computer-based work on postural balance. Results Results indicated that group I had a relatively more protruded head with extensive neck posture (angle B-A of group I and group II, 28.2±8.3, 32.9±6.0; p<.05). Th e COG of group I tended more toward the anterior than that of group II. Postural imbalance and impaired ability to regulate movement in forward and backward direction were also found. Conclusion The results of this study suggest that forward head postures during computer-based work may contribute to some disturbance in the balance of healthy adults. These results could be applied to education programs regarding correct postures when working at a computer for extended periods of time.

Affected by external factors, errors in time series data collected by sensors are common. Using the traditional method of constraining the speed change rate to clean the errors can get good performance. However, they are only limited to the data of stable changing speed because of fixed constraint rules. Actually, data with uneven changing speed is common in practice. To solve this problem, an online cleaning algorithm for time series data based on dynamic speed change rate constraints is proposed in this paper. Since time series data usually changes periodically, we use the extreme learning machine to learn the law of speed changes from past data and predict the speed ranges that change over time to detect the data. In order to realize online data repair, a dual-window mechanism is proposed to transform the global optimal into the local optimal, and the traditional minimum change principle and median theorem are applied in the selection of the repair strategy. Aiming at the problem that the repair method based on the minimum change principle cannot correct consecutive abnormal points, through quantitative analysis, it is believed that the repair strategy should be the boundary of the repair candidate set. The experimental results obtained on the dataset show that the method proposed in this paper can get a better repair effect.

Purpose: This study is to investigate the muscle activity based on real-time visual feedback training methods by rehabilitative ultrasound image in elderly and correlation between Heckmatt scale grade, muscle tone and thickness. Design: Cross-sectional study: Pilot study Methods: 6 elderly participated in the study with 2 conditions. Under the condition of rehabilitation ultrasound imaging equipment, all subjects performed voluntary maximal muscle contraction of the quadriceps 3 times using visual feedback based on Rehabilitative Ultrasound Imaging 1.0 (RUSI 1.0). Under the condition of only ultrasound images, all subjects performed voluntary maximal muscle contraction of the quadriceps 3 times using ultrasound image-based visual feedback. The muscle thickness and tone of the quadriceps were measured and the grades were classified by Heckmatt scale and all variables were comparative analyzed. Results: Heckmatt scale grade showed a negative correlation with muscle thickness at relaxation (p<0.05), and a negative correlation with the difference value obtained by subtracting muscle thickness at relaxation from muscle thickness at contraction in ultrasound image condition (p<0.05). The muscle tone during relaxation showed a negative correlation with the muscle thickness during relaxation (p<0.05). Conclusion: In the case of voluntary maximum muscle contraction of the quadriceps muscle in the elderly, it can be seen that the muscle thickness is getting larger when the RUSI 1.0-based visual feedback is provided than with only ultrasound image provided. And the lower Heckmatt scale grade is, the thicker the muscle is, and the lower the muscle tone is.

The deployment of embedded time-series sensing devices enabled better understanding of user environments and contexts. However, classifying them solely on extremely limited devices under data-scarce conditions is still a remaining challenge. We introduce GMT, a memory-efficient parameter-free classifier that uses gzip compressor and k-nearest neighbors (kNN) for classifying multi-channel time-series data. GMT tackles issues due to high data fidelity, multi-channel characteristics, and numerical properties of sensor data using techniques such as floating point quantization, channel-wise compression, and hybrid distance. Experiments show that GMT provides superior accuracy and memory efficiency compared to other classifiers across various tasks and applications.

Nowadays most vehicles are equipped with a variety of electronic devices to improve user convenience as well as its performance itself. In order to efficiently interconnect these devices with each other, Controller Area Network (CAN) is commonly used. However, the CAN requires reconfiguration of the entire network when a new device, which is capable of supporting both of transmission and reception of data, is added to the existing network. In addition, since CAN is based on the collision avoidance using address priority, it is difficult that a new node is assigned high priority and eventually it results in transmission delay of the entire network. Therefore, in this paper we propose a new system component, called CAN coordinator, and design a new CAN framework capable of supporting plug and play functionality. Through experiments, we also prove that the proposed framework can improve real-time ability based on plug and play functionality.

The real-time detection of malware remains an open issue, since most of the existing approaches for malware categorization focus on improving the accuracy rather than the detection time. Therefore, finding a proper balance between these two characteristics is very important, especially for such sensitive systems. In this paper, we present a fast portable executable (PE) malware detection system, which is based on the analysis of the set of Application Programming Interfaces (APIs) called by a program and some technical PE features (TPFs). We used an efficient feature selection method, which first selects the most relevant APIs and TPFs using the chi-square ($KHI^2$) measure, and then the Phi (${\varphi}$) coefficient was used to classify the features in different subsets, based on their relevance. We evaluated our method using different classifiers trained on different combinations of feature subsets. We obtained very satisfying results with more than 98% accuracy. Our system is adequate for real-time detection since it is able to categorize a file (Malware or Benign) in 0.09 seconds.

Multiple expression levels of genes obtained using time series microarray experiments have been exploited effectively to enhance understanding of a wide range of biological phenomena. However, the unique nature of microarray data is usually in the form of large matrices of expression genes with high dimensions. Among the huge number of genes presented in microarrays, only a small number of genes are expected to be effective for performing a certain task. Hence, discounting the majority of unaffected genes is the crucial goal of gene selection to improve accuracy for disease diagnosis. In this paper, a non-Gaussian weight matrix obtained from an incremental model is proposed to extract useful features of multivariate time series microarrays. The proposed method can automatically identify a small number of significant features via discovering hidden variables from a huge number of features. An unsupervised hierarchical clustering representative is then taken to evaluate the effectiveness of the proposed methodology. The proposed method achieves promising results based on predictive accuracy of clustering compared to existing methods of analysis. Furthermore, the proposed method offers a robust approach with low memory and computation costs.