• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.1
• /
• pp.50-55
• /
• 2021

Multi-path system is a system in which utilizes various networks simultaneously. It is expected that multi-path system can enhance communication speed, reliability, security of network. In this paper, we focus on path selection in multi-path system. To select optimal path, we propose deep reinforcement learning algorithm which is rewarded by the round-trip-time (RTT) of each networks. Unlike multi-armed bandit model, deep Q learning is applied to consider rapidly changing situations. Due to the delay of RTT data, we also suggest compensation algorithm of the delayed reward. Moreover, we implement testbed learning server to evaluate the performance of proposed algorithm. The learning server contains distributed database and tensorflow module to efficiently operate deep learning algorithm. By means of simulation, we showed that the proposed algorithm has better performance than lowest RTT about 20%.

• Proceedings of the Korean Statistical Society Conference
• /
• 2000.11a
• /
• pp.105-108
• /
• 2000

We consider a Markovian retrial queue with waiting space in which the service rates and retrial rates depend on the number of customers in the service facility and in the orbit, respectively. Each arriving customer from outside or orbit decide either to enter the facility or to join the orbit in Bernoulli manner whose entering probability depend on the number of customers in the service facility. In this paper, a stochastic order relation between two bivariate processes (C(t), N(t)) representing the number of customers C(t) in the service facility and N(t) one in the orbit is deduced in terms of corresponding parameters by constructing the equivalent processes on a common probability space. Some applications of the results to the stochastic bounds of the multi-server retrial model are presented.

• Journal of the Korean Statistical Society
• /
• v.29 no.4
• /
• pp.473-488
• /
• 2000

We consider a Markovian retrial queue with waiting space in which the service rates and retrial rates depend on the number of customers in the service facility and in the orbit, respectively. Each arriving customer from outside or orbit decide either to enter the facility or to join the orbit in Bernoulli manner whose entering probability depend on the number of customers in the service facility. In this paper, a stochastic order relation between two bivariate processes(C(t), N(t)) representing the number of customers C(t) in the service facility and one N(t) in the orbit is deduced in terms of corresponding parameters by constructing the equivalent processes on a common probability space. some applications of the results to the stochastic bounds of the multi-server retrial model are presented.

• Journal of Korean Institute of Industrial Engineers
• /
• v.27 no.1
• /
• pp.89-94
• /
• 2001

There are many queueing systems where customers wait for service up to a certain amount of time and leave the system if they are not served during that time. This paper considers a finite capacity multi-server queueing system with Poisson input and Erlang service time, where a customer becomes a lost customer when his service has not begun within an exponential patient time after his arrival. Performance measures such as average queue length, the average number of customers in service, and the proportion of lost customers can be obtained exactly through the proposed numerical solution procedure.

• Journal of KIISE:Software and Applications
• /
• v.30 no.3_4
• /
• pp.231-238
• /
• 2003

This paper presents a virtual grouping method used in multi-user network games, which reduces a response time and losses of response data. The proposed method divides each group into virtual groups and transmits data in them after dividing an overall map on a game into several fixed regions and grouping them. And this paper derives the optimal number of groups minimizing a given cost function. The proposed method if shown to be useful by comparing with a general grouping method.

• Proceedings of the Korea Information Assurance Society Conference
• /
• 2004.05a
• /
• pp.187-198
• /
• 2004

Virus protection infrastructure management is network infrastructure management, traffic route management, virus protection zone expansion, and virus protection management for gateway area. This research paper provides a diagnosis of characteristics and weaknesses of the structure of existing virus protection infrastructure, and recommends an improved multi-level virus protection infrastructure as a measure for correcting these weaknesses. Unproved virus protection infrastructure fitters unnecessary mail at the gateway stage to reduce the toad on server. As a result, number of transmission accumulation decreases due to the reduction in the CPU load on the Virus wall and increase in virus treatment rate.

• Journal of applied mathematics & informatics
• /
• v.33 no.1_2
• /
• pp.77-87
• /
• 2015

Parallel lines are often used to increase production rate in many manufacturing systems where the main line splits into several lines in parallel, and after some operations, they merge into a main line again. Queueing networks with finite buffers have been widely used for modeling and analyzing manufacturing systems. This paper provides an approximation technique for multi-server two-stage networks with merge configuration and blocking which will be a building block for analysis of general manufacturing systems with parallel lines and merge configuration. The main idea of the method is to decompose the original system into subsystems that have two service stations with multiple servers, two buffers and external arrivals to the second stage are allowed. The subsystems are modeled by level dependent quasi-birth-and-death (LDQBD) process.

• Journal of the Korea Society of Computer and Information
• /
• v.20 no.9
• /
• pp.47-53
• /
• 2015

In this paper, we propose a novel tangible interface system whose system does not use the expensive hardware is introduced. This proposed tangible interface is used on the table top capacitive multi touch-screen. The tangible interface apparatus which is called smart puck has sanguine arduino compatible board. The board has a Cds photo-sensing sensor and the EPP8266 WiFi module. The Cds sensor decodes the photometric PWM signals from the system and sends corresponding information to the system via TCP/IP. The system has a server called MT-Server to communicate with the smart pucks. The tangible interface shows reliable operation with fast response that is compatible to the expensive traditional devices in the market.

• Journal of the Korea Society of Computer and Information
• /
• v.20 no.8
• /
• pp.85-92
• /
• 2015

Due to the advancements in the capabilities of smart devices and home networks, we are able to easily access multimedia contents stored in a home server. In this paper, we present a wireless media content sharing mechanism for home networks that utilizes UPnP-based DLNA technology. We also present a novel Peer-to-Peer content sharing system that is able to operate on the client as well as server. Our system supports multiple push/pull services simultaneously via a multi-thread technique, and our intuitive user interface facilitates ease of use. Future studies would explore the feasibility of implementing our system in a multi-hop environment or providing a community-wide service.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.6 no.4
• /
• pp.53-66
• /
• 1996

World Wide Web is a total solution for multi-media data transmission on Internet. Because of its characteristics like ease of use, support for multi-media data and smart graphic user interface, WWW has extended to cover all kinds of applications. The Secure Information Center(SIC) is a data transmission system using conventional cryptography between client and server on WWW. It's main function is to support the encryption of sending data. For encryption of data IDEA(International Data Encryption Algorithm) is used and for authentication mechanism MD5 hash function is used. Since Secure Information Center is used by many users, conventional cryptosystem is efficient in managing their secure interactions. However, there are some restrictions on sharing of same key and data transmission between client and server, for example the risk of key exposure and the difficulty of key sharing mechanisms. To solve these problems, the Secure Information Center provides encryption mechanisms and key management policies.