pISSN : 1226 - 4717 / eISSN : 2287 - 3880
ISSUER : KICS
ISSUER : KICS
THE JOURNAL OF KOREAN INSTITUTE OF COMMUNICATIONS AND INFORMATION SCIENCES (J-KICS) (formerly, the Korea Information and Communications Society) launched THE JOURNAL OF KOREAN INSTITUTE OF COMMUNICATIONS AND INFORMATION SCIENCES (J-KICS) (formerly, the Journal of Korean Institute of Communications Sciences) in 1976. During the early six years, this publication was published annually and now it is published monthly since 1990. This publication has been printed in three parts since 2001: Part A (Communication Theory and Systems), Part B (Networks and Services), and Part C (Convergence Technologies). Currently, it categorizes papers into regular papers (A: AI for ICT Applications, B: Communications Systems, C: Networks and Computings, D: Services Applications and Emerging Topics, E: ICT Convergence) and special papers. During the last ten years, about 400 papers have been published yearly. The title of this publication has been changed from THE JOURNAL OF KOREAN INSTITUTE OF COMMUNICATIONS AND INFORMATION SCIENCES to THE JOURNAL OF KOREAN INSTITUTE OF COMMUNICATIONS AND INFORMATION SCIENCES in 2007, under the same ISSN of 1226-4717.
Indexed in
Publishing with
Best papers (Vol. 50, No. 5, May. 2025)
Analysis of Random Characteristics of Pseudo-Random Real Number and Binary Sequences Generated by Cascade Chaotic MapsHyojeong Choi Gangsan Kim Hong-Yeop Song Sangung Shin Chulho Lee Hongjun Noh |
|
| In this paper, we evaluate the dynamic properties, complexity, and randomness of cascaded Logistic, Sine, and Chebyshev maps using the Lyapunov Exponent (LE), Approximate Entropy (ApEn), and Permutation Entropy (PE). Additionally, we convert the real... | |
Implementation of Multiple Transmitter Tracking Mechanism Based on DeepSORT for 2D MIMO Optical Camera CommunicationsEun-bi Shin Yeong Min Jang |
|
| This paper presents an enhanced 2D MIMO Optical Camera Communication (OCC) system that integrates a multi-transmitter tracking mechanism enhanced by the DeepSORT algorithm to address real-time communication challenges in dynamic environments. The sys... | |
Latest Publication (Vol. 50, No. 5, May. 2025)
Lightweight LiDAR—Camera Online Extrinsic Calibration with Multi-Dilation Encoder Blocks
Sang-Chul Kim Yeong-min Jang
As the integration of multi-sensor systems, such as cameras and LiDAR, becomes increasingly common in various fields, the development of efficient and accurate extrinsic calibration techniques is emerging as a critical task. This article presents a novel lightweight deep-learning network for LiDAR―Camera targetless extrinsic calibration, which consists of only 4 million parameters. The proposed method utilized CNN-based multi-dilation encoder blocks which can extract multi-scale features, especially for sparse LiDAR depth image.
The proposed block allows the network to be lightweight and excel in calibration performance. The proposed method achieved translation errors of 1.08 cm, 0.18 cm, and 0.56 cm along the X, Y, and Z axes, respectively. Additionally, it achieved rotation errors of 0.182°, 0.139°, and 0.141° for roll, pitch, and yaw, respectively. The proposed method also performs calibration in a one-shot approach, which is suitable for real-time applications. These results highlight the capabilities of the proposed method in enabling reliable fusion of LiDAR and camera data, enhancing the perception capabilities of autonomous vehicles.
Deep Learning-Based Lightweight LiDAR and Fisheye Camera Online Extrinsic Calibration
Sang-Chul Kim Yeong-min Jang
Autonomous driving has been extensively researched in recent years. To improve the mobility and decision-making capabilities of autonomous vehicles, multiple sensors have been integrated to complement the limitations of individual sensors. For example, Light Detection and Ranging (LiDAR) is frequently combined with camera data to overcome the narrow field-of-view (FoV) of traditional pinhole cameras. The fisheye cameras of LiDAR expand the FoV from up to 80° in traditional cameras to 180°, which is advantageous for autonomous driving applications. This study introduces a lightweight, deep learning-based LiDAR-fisheye camera fusion model for real-world environments. The mean translation errors are 1.375, 0.753, and 1.208 cm along the X, Y, and Z axes, respectively, and the mean rotation errors are 0.171°, 0.150°, and 0.089° in the roll, pitch, and yaw directions, respectively. These results demonstrate the efficiency and proficiency of our sensor-fusion approach for autonomous driving.
Reinforcement Learning-Based Autonomous Underwater Vehicle Waypoint Generation Algorithm in Dynamic Environments
Emily Jimin Roh Hyunsoo Lee Ilseok Song Seunghwan Kim Youngdae Kim Soohyun Park Joongheon Kim
This paper proposes a method to optimize the autonomous torpedo maneuver path for reaching the target of torpedoes, which are explosive projectile weapons in naval operations. For flexible maneuvering of torpedoes, movement in various directions is considered. Also, the obstacles in the actual marine environment and the minimization of the waypoint that occurs when the angle of the torpedoes is changed considered to increase the efficiency of torpedo maneuvering. Consequently, this study presents the environment that reflects the action of the torpedo in various directions according to the maximum rotation angle. Torpedo maneuver strategy is formulated by applying a Markov Decision Process based reinforcement learning algorithm, Q-Learning. Compared to the general Q-Learning algorithm, the superiority of the proposed algorithm is assessed and its applicability in the actual marine environment, through the success rate of reaching the target point and the number of waypoints.
Mask-Based Selective Downsampling in Convolutional Neural Networks
Chulyoung Kwak Saewoong Bahk
This letter presents a study on selective down-sampling techniques designed to enhance the efficiency of convolutional neural networks (CNNs).
By selectively adjusting the resolution of feature layers based on the down-sampling mask, this method aims to improve efficiency. We develop a network that is 80% lighter than the baseline scheme, and introduce several methods to mitigate the performance degradation of light weight networks. The performance of the proposed methods is evaluated through extensive experiments.
Resizing Method for Applying RF-based Data to ViT in Human Activity Recognition
Jeongjun Park Saewoong Bahk
This paper applies RF-based data, obtained through the commonly used Radio Frequency (RF) approach in human activity recognition (HAR), to the Vision Transformer (ViT), a state-of-the-art machine learning method for image classification.
Through this process, we analyze the challenges arising from applying RF-based data, which have different sizes compared to standard image dimensions, to ViT. To address these challenges, we propose various input resizing methods. Furthermore, through a comparison of these resizing methods, we identify the most effective resizing approach for RF-based data, achieving an average accuracy improvement of 9.57%.
Terminal Mobility Prediction for Deep Reinforcement Learning-Based Handover Optimization in Non-Terrestrial Networks
Junyoung Kim Huiyeon Jang In-Sop Cho Minsu Shin Soyi Jung
Low Earth orbit (LEO) satellites are crucial for global coverage and real-time communication services.
However, their rapid mobility and unique channel characteristics pose challenges for conventional handover techniques, leading to frequent disruptions and limited seamless connectivity. Optimized methods are needed to address the satellites' movement and the stochastic mobility of user terminals. This paper proposes a novel approach combining deep learning and reinforcement learning to optimize handovers. Time-series data of satellite and terminal movements are analyzed to predict the received signal strength (RSSI) using deep learning. Based on the predicted RSSI, a reinforcement learning-based framework determines the optimal handover timing. This integration achieves faster convergence and precise handover decisions, enhancing RSSI and overall system performance.
Best Papers Analysis of Random Characteristics of Pseudo-Random Real Number and Binary Sequences Generated by Cascade Chaotic Maps
Hyojeong Choi Gangsan Kim Hong-Yeop Song Sangung Shin Chulho Lee Hongjun Noh
In this paper, we evaluate the dynamic properties, complexity, and randomness of cascaded Logistic, Sine, and Chebyshev maps using the Lyapunov Exponent (LE), Approximate Entropy (ApEn), and Permutation Entropy (PE). Additionally, we convert the real-number sequences generated by these maps into binary sequences using two different binary mapping methods, and we compare the autocorrelation and cross-correlation properties of the binary sequences with those of m-sequences. The experimental results show that the binary sequences generated from cascaded chaotic maps can produce a large set of codes with excellent correlation properties and randomness, depending on various initial conditions and control parameters.
Indoor Visible Light Positioning Technique Based on RIS Position and AOA Estimation
Yong Up Lee
In the RIS (reconfigurable intelligent surface) assisted visible light (VL) positioning based on angle-of-arrival (AOA) estimation for 6G communication service, VL positioning channel is largely dependent of RIS position and it is difficult to obtain the exact AOA parameter estimate by conventional technique. In this paper, the new AOA estimation technique based on weighted mean steering vector concept is proposed to estimate the angle-of-arrival (AOA) parameters of multiple clusters under different visible light positioning channel due to RIS position. It is seen that the proposed AOA estimation technique gives an optimal AOA estimate, however the previously published method meets a failure because the widely distributed multipath signals by RIS reflection arrive at the receiver under the positioning environment based on RIS around receiver. The proposed method has also the advantages of the simpler system architecture, the better positioning accuracy, and the lower computation, compared with the conventional one.
Efficient WL Beamforming with Projection onto Interference Null Space
Yang-Ho Choi
In case noncircular signals are incident into a sensor array the performance of a beamformer can be improved by introducing the widely linear (WL) scheme. When the incident signals are strictly noncircular, in which the magnitudes of the noncircular coefficients are one, an efficient WL beamforming method is proposed through the interference null space projection. In the proposed method, the arrival angle and the initial phase of the desired signal are estimated via the search over its direction sector by NC-MUSIC (noncircular multiple signal classification) to attain an extended steering vector. Utilizing it, the interference subspace is effectively estimated by removing the desired signal component from the eigendecomposition of the sample matrix. The weight vector is computed such that it is orthogonal to the estimated subspace. The computational burden is light since most computations are carried out with real numbers by using the conjugate symmetry of the extended received vector. Simulation results show that it has robustness against pointing errors, outperforming existing WL beamformers including the one dealing with a very complex nonlinear problem to overcome steering vector errors.
K-means Assisted Simulated Annealing Algorithm for Controller Placement
Haeun Kim Dongkyun Ryoo Hongrok Choi Sanghoon Lee Junhyeong Kim Jinho Park Hyun Park Kihun Kim Sungjoon Ahn Sangheon Pack
The multi-controller placement problem (MCPP) in software-defined networking is a complex optimization problem. This paper proposes the K-means Assisted Simulated Annealing Controller Placement (KASA-CP) algorithm, which combines k-means and simulated annealing to address MCPP efficiently. KASA-CP improves the computational efficiency of simulated annealing by using k-means for initial placement, ensuring effective operation in large-scale networks. Experimental results show that KASA-CP outperforms the basic simulated annealing algorithm in terms of execution time and average latency.
pISSN : 1226 - 4717
eISSN : 2287 - 3880