Jung-Hyok Kwon

Neighbor stability-based VANET clustering for urban vehicular environments

In this paper, we propose a neighbor stability-based VANET clustering (NSVC) that can efficiently deliver data in urban vehicular environments. The salient features of urban vehicles are their high mobility and unpredictable direction of movement, so vehicle-to-vehicle and vehicle-to-infrastructure (V2X) communication should take into consideration the frequent changes in the topology of vehicular ad hoc networks (VANETs). These technical challenges are addressed with NSVC by including a neighbor stability-based VANET clustering scheme and the corresponding supplementary transmission scheduling method. Thereby, NSVC supports fast cluster formation, minimizes the number of cluster head elections, and moreover guarantees the reliable delivery of data for emergency messages. The results of the simulation indicate that NSVC achieves better network performance when compared to existing approaches.

Adaptive multi-channel allocation for vehicular infrastructure mesh systems

This paper focuses on a wireless solution for vehicular infrastructure systems. In order to achieve both low cost and high efficiency, infrastructures can be connected to each other in vehicular networks by a wireless link similar to a mesh router in wireless mesh networks (WMNs). However, the existing WMN solutions cannot appropriately support various vehicular applications that require high rate and low latency communications. Therefore, in this paper, we present the design and performance evaluation of an adaptive multi-channel allocation for vehicular infrastructure mesh systems (abbreviated AMCA). In order to meet both high rate and low latency communications, AMCA is designed to provide optimal channel assignment duration for each flow to efficiently utilize multiple non-overlapping channels. The performance evaluation of AMCA is conducted by the QualNet 5.0 simulator under various network scenarios to consider diverse network conditions. Simulation results show that AMCA can achieve higher network throughput and lower average packet delay than other well known wireless solutions.

Low-Concentration Indium Doping in Solution-Processed Zinc Oxide Films for Thin-Film Transistors

We investigated the influence of low-concentration indium (In) doping on the chemical and structural properties of solution-processed zinc oxide (ZnO) films and the electrical characteristics of bottom-gate/top-contact In-doped ZnO thin-film transistors (TFTs). The thermogravimetry and differential scanning calorimetry analysis results showed that thermal annealing at 400 °C for 40 min produces In-doped ZnO films. As the In content of ZnO films was increased from 1% to 9%, the metal-oxygen bonding increased from 5.56% to 71.33%, while the metal-hydroxyl bonding decreased from 72.03% to 9.63%. The X-ray diffraction peaks and field-emission scanning microscope images of the ZnO films with different In concentrations revealed a better crystalline quality and reduced grain size of the solution-processed ZnO thin films. The thickness of the In-doped ZnO films also increased when the In content was increased up to 5%; however, the thickness decreased on further increasing the In content. The field-effect mobility and on/off current ratio of In-doped ZnO TFTs were notably affected by any change in the In concentration. Considering the overall TFT performance, the optimal In doping concentration in the solution-processed ZnO semiconductor was determined to be 5% in this study. These results suggest that low-concentration In incorporation is crucial for modulating the morphological characteristics of solution-processed ZnO thin films and the TFT performance.

Asymmetric Directional Multicast for Capillary Machine-to-Machine Using mmWave Communications

The huge demand for high data rate machine-to-machine (M2M) services has led to the use of millimeter Wave (mmWave) band communications with support for a multi-Gbps data rate through the use of directional antennas. However, unnecessary sector switching in multicast transmissions with directional antennas results in a long delay, and consequently a low throughput. We propose asymmetric directional multicast (ADM) for capillary M2M to address this problem in mmWave communications. ADM provides asymmetric sectorization that is optimized for the irregular deployment pattern of mulicast group members. In ADM, an M2M gateway builds up asymmetric sectors with a beamwidth of a different size to cover all multicast group members with the minimum number of directional transmissions. The performance of ADM under various simulation environments is evaluated through a comparison with legacy mmWave multicast. The results of the simulation indicate that ADM achieves a better performance in terms of the transmission sectors, the transmission time, and the aggregate throughput when compared with the legacy multicast method.

Unified Medium Access Control Architecture for Resource-Constrained Machine-to-Machine Devices

In capillary machine-to-machine (M2M) communications, which is being considered as a feasible network solution for M2M applications, because of physical resource constraints and deployment conditions, an energy-efficient and scalable medium access control (MAC) protocol is crucial for numerous M2M devices to concurrently access wireless channels. Therefore, this paper presents a unified MAC layer architecture for resource-constrained M2M devices in capillary M2M networks [named as resource-constrained MAC architecture (RCMA)], which has a unified (monolithic) framework consisting of essential functional components to support MAC-related operations of M2M devices: multi-channel hybrid MAC (McHM), logical link control (LLC), time synchronizer (TS), and device on--off scheduler (DO2S). McHM provides a baseline MAC protocol for an entire capillary M2M system that combines the benefit of both contention-based carrier sense multiple access and schedule-based time division multiple access schemes, whereas the other three components help in the McHM operations. To demonstrate the effectiveness of the RCMA, we implement the whole stack using the QualNet simulator. Experimental results show that the RCMA outperforms the conventional ZigBee stack in terms of energy efficiency and scalability, even under heavy traffic conditions.

Data Aggregation Gateway Framework for CoAP Group Communications

In this paper, a data aggregation gateway framework (DA-GW) for constrained application protocol (CoAP) group communications is proposed. The DA-GW framework is designed to improve the throughput performance and energy efficiency of group communication to monitor and control multiple sensor devices collectively with a single user terminal. The DA-GW consists of four function blocks—the message analyzer, group manager, message scheduler and data handler—and three informative databases—the client database, resource database and information database. The DA-GW performs group management and group communication through each functional block and stores resources in the informative databases. The DA-GW employs international standard-based data structures and provides the interoperability of heterogeneous devices used in various applications. The DA-GW is implemented using a Java-based open source framework called jCoAP to evaluate the functions and performance of the DA-GW. The experiment results showed that the DA-GW framework revealed better performance than existing group communication methods in terms of throughput and energy consumption.

FS-IIoTSim: a flexible and scalable simulation framework for performance evaluation of industrial Internet of things systems

This paper presents a flexible and scalable network simulation framework (FS-IIoTSim) for performance evaluation of the Industrial Internet of Things (IIoT) system. FS-IIoTSim provides flexibility and scalability for performance analysis of the IIoT system by supporting not only existing legacy industrial sensor network communication models, but also user protocol modification, in order to accommodate the complexity of systems such as communication standards, smart machine traffic, and applications over the complex IIoT environment. The FS-IIoTSim consists of three parts: (1) scenario modeling, (2) performance evaluation, and (3) user interface. Each of these parts interacts and cooperates with each other to simulate a target IIoT system. Scenario modeling utilizes user simulation scenarios to create an industrial sensor network (ISN) system model and perform simulations to generate a trace file recorded with log information about the device’s operations. Performance evaluation is an analytical process of a trace file, generated in the scenario modeling part, which is used to derive a user’s preferred performance metric. The user interface part plays a role in interfacing with the detailed operations of the other parts. Within the user interface part, the user can set and control the FS-IIoTSim framework. In order to prove the effectiveness of this FS-IIoTSim, experiments were conducted in various industrial application scenarios, and simulation performance was compared and analyzed with a dedicated simulator. The experimental results showed that FS-IIoTSim provided more effective approaches to modeling and performance analysis of real ISN environments, compared to a dedicated simulator.

Regional Density-aware Data Collection Using Unmanned Aerial Vehicle in Large-scale Wireless Sensor Networks

This paper presents a regional density-aware data collection (RDDC) using unmanned aerial vehicles (UAVs) in large-scale wireless sensor networks (WSNs). The RDDC is designed to support fast data collection by adaptively adjusting the value of the minimum contention window (CWmin) of the sensor nodes depending on the regional node density of sensing fields. The operation of RDDC consists of the region-partitioning, node-scanning, and data-collecting phases. In the region-partitioning phase, the RDDC partitions the sensing field into multiple regions and assigns a region ID to each sensor node according to the region to which it belongs. In the node-scanning phase, the RDDC calculates the node density of each region. Finally, in the data-collecting phase, the RDDC assigns different CWmin values to each sensor node considering the regional node density of each region. The experimental simulation results show that the RDDC obtains higher aggregate throughput and shorter delay than the existing data collection method.

Variable-categorized clustering algorithm using fuzzy logic for Internet of things local networks

This paper presents a variable-categorized clustering algorithm (VCCA) using fuzzy logic for Internet of Things (IoT) local networks. The VCCA selects the cluster head (CH) that has the highest network capacity through a classification process of cluster variables in accordance with the characteristics in order to configure a clustered network, which differs for different IoT applications. To achieve this, the VCCA employs a fuzzy inference system (FIS) that calculates an outcome through rule-based variable mapping for low complexity in the CH election and high scalability of cluster variables. In addition, experimental simulations using MATLAB are conducted to evaluate the performance of the VCCA. The simulation results show that the VCCA exhibits better network performance compared to the existing algorithms in terms of throughput, end-to-end latency, network lifetime, and energy consumption.

Categorization-based video streaming for traffic mitigation in content delivery services

This paper presents a categorization-based video streaming approach (CVS) to mitigate the Internet traffic in content delivery services. The CVS uses the statistical information related to the users’ view patterns (i.e., the average view duration in accordance with the playback time of the content) to adjust the request period of content chunks. Therefore, it can reduce the amount of Internet traffic by reducing unnecessary chunk requests and content chunks. The operation of the CVS is based on the average view duration provided by the content provider (CP). However, even if the CP does not provide the average view duration of the content, the CVS can properly predict the average view duration by using the content categorization and adjust the request period of content chunks. The simulation results show that the CVS achieves better performance in terms of the average waste ratio of network resources, the amount of network traffic, and the number of chunk requests.