• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.27 no.11C
• /
• pp.1152-1157
• /
• 2002

In this paper, a detailed procedure of development for an AAL2 processor widely used in voice gateway application is introduced. The processor supports CPS and SSCS with voice service and framed mode data service. It provides 4 ATM virtual connections, which include 1020 AAL2 channels. The processor has one UTOPIA Level 1 interface for an ATM cell interface and 4 TDM ports for a voice channel interface. The TDM ports carry PCM/ADPCM voice streams. Most AAL2 processors are implemented as software, or hardware and software, so its latency is large. But this processor has very low latency as to CPS and SSCS because all of them are implemented in hardware. Also, it allows not only loopback and switching of CPS packets, but loopback and switching of TDM channels. The key feature is that the internal structure of the CPS and SSCS in this processor seems like as each software function, so they are called whenever they are required. In addition, they are reusable for another design and are scalable for more channels.

• Journal of Broadcast Engineering
• /
• v.10 no.3
• /
• pp.340-347
• /
• 2005

This paper presents a polyphase down sampling based multiple description coding applied to H.264 video coding standard. For a given macroblock, a residual macroblock is calculated by motion estimation, and before applying DCT, quantization and entrophy coding of the H.264 coding process, the polyphase down sampling is applied to the residual macroblock to code in four separate descriptions. Experiments were performed for all the 9 test sequences of JVT SVC standardization in various packet loss patterns. Experimental results show that the proposed one gives 0.5 to 5 dB enhancement over an error-concealment based on the slice group map technolgoy.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.8
• /
• pp.4077-4091
• /
• 2017

This paper presents the design of a broadcasting scenario and system for an 8K-resolution content. Due to an 8K content is four times larger than the 4K content in terms of size, many technologies such as content acquisition, video coding, and transmission are required to deal with it. Therefore, high-quality video and audio for 8K (ultra-high definition television) service is not possible to be transmitted only using the current terrestrial broadcasting system. The proposed broadcasting system divides the 8K content into four 4K contents by area, and each area is hierarchically encoded by Scalable High-efficiency Video Coding (SHVC) into three layers: L0, L1, and L2. Every part of the 8K video content divided into areas and hierarchy is independently treated. These parts are transmitted over heterogeneous networks such as digital broadcasting and broadband networks after going through several processes of generating signal messages, encapsulation, and packetization based on MPEG media transport. We propose three methods of generating streams at the sending entity to merge the divided streams into the original content at the receiving entity. First, we design the composition information, which defines the presentation structure for displays. Second, a descriptor for content synchronization is included in the signal message. Finally, we define the rules for generating "packet_id" among the packet header fields and design the transmission scheduler to acquire the divided streams quickly. We implement the 8K broadcasting system by adapting the proposed methods and show that the 8K-resolution contents are stably received and serviced with a low delay.

• Journal of IKEEE
• /
• v.26 no.4
• /
• pp.568-576
• /
• 2022

Automatic modulation classification(AMC) is a core technique in Software Defined Radio(SDR) platform that enables smart and flexible spectrum sensing and access in a wide frequency band. In this study, we propose a simple yet accurate deep learning-based method that allows AMC for variable-size radio signals. To this end, we design a classification architecture consisting of two Convolutional Neural Network(CNN)-based models, namely main and small models, which were trained on radio signal datasets with two different signal sizes, respectively. Then, for a received signal input with an arbitrary length, modulation classification is performed by augmenting the input samples using a self-replicating padding technique to fit the input layer size of our model. Experiments using the RadioML 2018.01A dataset demonstrated that the proposed method provides higher accuracy than the existing methods in all signal-to-noise ratio(SNR) domains with less computation overhead.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.21 no.6
• /
• pp.1197-1202
• /
• 2017

Security services that support electric energy service gateway require relatively high reliability. In particular, the application services that accompany communications and data are run organically. Each of the security services should support a secure service platform that supports a secure, scalable life cycle for existing services which should be extends security layer of Universal Middleware. In this convergence platform, it is the study of security transfer modular services that allow independent life cycle management of systems through Universal middleware. First, It is modular in terms of energy consumption service and data, enabling real-time operation, communications, remote management and applications. Second, the life cycle of the secure module to support start, stop and updating of the security module by applying the security policy module layer concept. It is modular system enabling to design of dyanmic models in the smart grid, the service was intended to be standardized and applied to the security service platform.

• Journal of KIISE:Computing Practices and Letters
• /
• v.13 no.4
• /
• pp.205-212
• /
• 2007

In this paper, we present a prototype of view-switchable high-definition (HD) multi-view video transmission system. One of the major bottlenecks for the multi-view broadcasting system has been the hardware cost and transmission bandwidth. The proposed system focuses on software-based design, transmission over IP multicast networks, and flexible system configuration to address aforementioned problems. In the proposed system, we implement software stereo HD multiplexing, demultipiexing and decoding, and take advantage of high-speed broadband convergence networks to deliver HD video in real-time. Moreover, the proposed system can be scalable and flexible in terms of the number of views. Furthermore, in order to display any multiview video on 3D display monitor, a face tracking system is integrated to our system. Therefore, users can watch the different stereoscopic video at its related locations.

• The Journal of Society for e-Business Studies
• /
• v.24 no.3
• /
• pp.87-98
• /
• 2019

As artificial intelligence is actively studied, it is being applied to various fields such as image, video and natural language processing. The natural language processing, in particular, is being studied to enable computers to understand the languages spoken and spoken by people and is considered one of the most important areas in artificial intelligence technology. In natural language processing, it is a complex, but important to make computers learn to understand a person's common sense and generate results based on the person's common sense. Knowledge graphs, which are linked using the relationship of words, have the advantage of being able to learn common sense easily from computers. However, the existing knowledge graphs are organized only by focusing on specific languages and fields and have limitations that cannot respond to neologisms. In this paper, we propose an intelligent chatbotsystem that collects and analyzed data in real time to build an automatically scalable knowledge graph and utilizes it as the base data. In particular, the fine-tuned BERT-based for relation extraction is to be applied to auto-growing graph to improve performance. And, we have developed a chatbot that can learn human common sense using auto-growing knowledge graph, it verifies the availability and performance of the knowledge graph.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.05a
• /
• pp.94-96
• /
• 2021

Software-Defined Networking (SDN), which has emerged to overcome the limitations of existing network architectures, makes routing management simpler and more efficient through a central controller. SR (Segment Routing) is a flexible and scalable way of doing source routing, and defines the information path of the network through a list of segments arranged in the packet header. In an SDN environment, the performance of each router is almost the same, but packets tend to be concentrated on routes that are frequently used depending on routing algorithms. Routers in that path have a relatively high frequency of failure and are more likely to become bottlenecks. In this paper, we propose a routing algorithm that allows the router, which is a resource in the network, to evenly process packets in the SDN with SR, so that the administrator can utilize the resources in the network without idle routers, and at the same time facilitate the management of the router.

• ETRI Journal
• /
• v.35 no.5
• /
• pp.786-796
• /
• 2013

In this paper, we propose a hybrid time-division multiplexing and dense wavelength-division multiplexing scheme to implement a cost-effective and scalable long-reach optical access network (LR-OAN). Our main objectives are to increase fiber plant utilization, handle upstream and downstream flow through the same input/output port, extend the reach, and increase the splitting ratio. To this end, we propose the use of an arrayed waveguide grating (AWG) and an erbium-doped fiber amplifier (EDFA) in one configuration. AWG is employed to achieve the first and second objectives, while EDFA is used to achieve the third and fourth objectives. The performance of the proposed LR-OAN is verified using the Optisystem and Matlab software packages under bit error rate constraints and two different approaches (multifiber and single-fiber). Although the single-fiber approach offers a more cost-effective solution because service is provided to each zone via a common fiber, it imposes additional losses, which leads to a reduction in the length of the feeder fiber from 20 km to 10 km.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.9
• /
• pp.3885-3906
• /
• 2020

Hybrid fuzzing which combines fuzzing and concolic execution, has proved its ability to achieve higher code coverage and therefore find more bugs. However, current hybrid fuzzers usually suffer from inefficiency and poor scalability when applied to complex, real-world program testing. We observed that the performance bottleneck is the inefficient cooperation between the fuzzer and concolic executor and the slow symbolic emulation. In this paper, we propose a novel solution named EPfuzzer to improve hybrid fuzzing. EPfuzzer implements two key ideas: 1) only the hardest-to-reach branch will be prioritized for concolic execution to avoid generating uninteresting inputs; and 2) only input bytes relevant to the target branch to be flipped will be symbolized to reduce the overhead of the symbolic emulation. With these optimizations, EPfuzzer can be efficiently targeted to the hardest-to-reach branch. We evaluated EPfuzzer with three sets of programs: five real-world applications and two popular benchmarks (LAVA-M and the Google Fuzzer Test Suite). The evaluation results showed that EPfuzzer was much more efficient and scalable than the state-of-the-art concolic execution engine (QSYM). EPfuzzer was able to find more bugs and achieve better code coverage. In addition, we discovered seven previously unknown security bugs in five real-world programs and reported them to the vendors.