• Journal of KIISE:Computer Systems and Theory
• /
• v.32 no.11_12
• /
• pp.541-556
• /
• 2005

The memory access latency of the system has been a primary factor of performance degradation in single-processor system and multi-processor system. The remote memory access latency takes a lot of overhead over the local memory access latency especially in the distributed shared-memory system. To resolve this problem, the multi-level cache architecture that contains a remote cache in the multi-processor system has been proposed. In this paper, we propose a new cache replacement policy that improves the performance of the multi-processor system with the remote cache. If the multi-level cache keeps the multi-level inclusion(MLI) property and uses the LRU(Least Recently Used) cache replacement policy, the LRU information of the higher-level cache(a processor cache) would be different with that of the lower-level cache(a remote cache). In this situation, the replacement of a remote cache line can induce the exchange of a processor cache line that is used by the processor. It is a main factor of performance degradation in a whole system. To alleviate this disadvantage of the LRU replacement polity, the new policy analyses tht processor's remote memory access pattern of each node and uses this information to reduce the number of invalidations of the useful cache line in the higher-level cache. The new replacement policy of the remote cache can improve the performance by $3.5\%$ in maximum and $2.5\%$ in average on SPLASH-2 benchmarks, compared to the general LRU cache replacement policy.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2008.02a
• /
• pp.163-166
• /
• 2008

이 논문에서는 MPEG-4 SVC 신호를 DVB-S2 전송시스템에 적용하여 Ka대역 위성방송서비스를 제공하는 전송방안에 대한 연구를 하였다. SVC신호는 계층 수를 2개로 제한하였으며, 이러한 다계층 신호를 DVB-S2 전송시스템에 적용하기 위하여, 성상도상에서 각 계층의 신호를 동기적으로 합치는 Hierarchical 변조방식과 각 계층의 전송스트림을 RF 주파수 상에서 비동기적으로 합치는 방식인 Layered 변조방식을 사용하였다. 각 전송방안에서 변조방식과 채널 부호율을 결합 소스/채널 부호화 방안에 따라 변화시켜 최적의 전송기법을 도출하였다.

• The Journal of the Korea Contents Association
• /
• v.9 no.12
• /
• pp.515-527
• /
• 2009

IEEE 802.11n is an ongoing next-generation WLAN(Wireless Local Area Network) standard that supports a very high-speed connection with more than 100Mb/s data throughput measured at the MAC(Medium Access Control) layer. Study trends of IEEE 802.11n show two aspects, enhanced data throughput using aggregation among packets in MAC layer, and better data rates adapting MIMO(Multiple-Input Multiple-Output) in PHY(Physical) layer. But, the former doesn't consider wireless channel and the latter doesn't consider aggregation among packets for reality. Therefore, this paper analyzes data throughput for IEEE 802.11n considering MAC and PHY connection. A-MPDU(Aggregation-MAC Protocol Data Unit) and A-MSDU(Aggregation-MAC Service Unit) is adapted considering multi-service in MAC layer, WLAN MIMO TGn channel using SVD(Singular Value Decomposition) is adapted considering MIMO and wireless channel in PHY layer. Consequently, Simulation results shows throughput between A-MPDU and A-MSDU. Also, We use Ns-2(Network simulator-2) for reality.

• Journal of the Korea Society of Computer and Information
• /
• v.13 no.7
• /
• pp.127-138
• /
• 2008

Configuration and data management system for weapon system research and development processes should assist concurrent engineering and collaboration activities for various documents, drawings, part informations and product structure informations. In this thesis we defined the system architecture using the CBD methodology and implemented the component based system development process. Most of the configuration and data management systems consist of three layered system architecture which use the general MVC models. In this thesis we defined four layered system architecture based on EJB structure under J2EE environments. Through the four layered system architecture, we break down the general model layer into two detailed business layers and the ordinary control layer into two system dependent layers. With these four layers, we proposed the methodology that detailed the characteristics of the components. Through the CBD approach and the component based development process, we can get the representation model and implement the actual system of the configuration and data management system for weapon system R&D.

• The KIPS Transactions:PartC
• /
• v.17C no.1
• /
• pp.61-68
• /
• 2010

As the mobile technology advances, file searching among the mobile device users becomes more important. In this paper, we propose the uniform grid, greedy, and MIS P2P systems that have double-layered topology to search files efficiently for mobile ad-hoc networks. In these systems, peers are classified into two groups, super-peers and sub-peers, and each super-peer manages its neighboring sub-peers. In the proposed systems, each super-peer maintains the appropriate information of its sub-peers so that when a peer requests a file, the request is sent to its super-peers and then to neighboring super-peers. Hence the proposed systems could avoid multi-broadcasting and reduce network overheads. The experimental results show that the proposed systems outperform a single-layered P2P system in terms of the average number of messages to find target files. Especially the MISsystem improves by reducing the average number of messages by 48.9% while maintaining the same search accuracy.

• Journal of Korea Society of Industrial Information Systems
• /
• v.4 no.4
• /
• pp.149-153
• /
• 1999

This paper proposes a new security system for the next generation mobile telecommunication system The system is based on the network domain security of the 3GPP. The system provides confidentiality and integrity services in the RLC layer, the second layer of 3GPP. Our system has merits in that it can provide security services without any modification to the 3GPP protocol and has low transmission voerhead. Security algorithm and mode are controlled by the RRC layer, the third layer of 3GPP.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.05a
• /
• pp.494-496
• /
• 2022

엣지 디바이스에서 딥러닝을 활용하기 위하여 CNN 경량화 연구들이 진행되고 있다. 경량 CNN 은 대부분 고정 소수점을 사용하며, 계층에 따라 정밀도는 달라진다. 본 논문에서는 경량 CNN 을 지원하기 위하여, 사용 계층에 따라 정밀도를 선택할 수 있는 가변 정밀도 곱셈기를 제안한다. 제안하는 가변 정밀도 곱셈기는 낮은 정밀도 곱셈기를 병합하는 구조로, 정밀도가 낮을 때는 병렬 처리를 통해 효율을 높인다. 제안하는 곱셈기를 Verilog HDL로 설계하고 ModelSim 에서 동작을 확인하였다. 설계된 곱셈기는 계층별로 정밀도가 다른 CNN 가속기에서 효율적으로 적용될 것으로 기대된다.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.16 no.6
• /
• pp.65-73
• /
• 2016

Peer-to-Peer (P2P) systems with super-peer overlay networks show combined advantages of both hybrid and pure P2P systems. Super-peer is a special peer acting as a server to a cluster of generic peers. Organizing a super-peer network is one of important issues for P2P systems with super-peer networks. Conventional P2P systems are based on two-level hierarchies of peers. One is a layer for generic peers and the other is for super-peers. And it is usual that super-peer networks have forms of random graphs. However, for accommodating a large-scale collection of generic peers, the super-peer network has also to be extended. In this paper, we propose a scheme of hierarchically constructing super-peer networks for large-scale P2P systems. At first, a two-level tree, called a simple super-peer network, is proposed, and then a scheme of generalizing and then extending the simple super-peer network to multi-level super-peer network is presented to construct a large-scale super-peer network. We call it an extended super-peer network. The simple super-peer network has several good features, but due to the fixed number of levels, it may have a scalability problem. Thus, it is extended to k-level tree of a super-peer network, called extended super-peer network. It shows good scalability and easy management of generic peers for large scale P2P system.

• Communications for Statistical Applications and Methods
• /
• v.16 no.2
• /
• pp.217-228
• /
• 2009

In this paper, we consider a multi-class queueing system in which different classes of customers have different arrival rates, service rates and class-transition probabilities. We use the fast simulation method to estimate the overflow probability and the expected number of customers of each class at the first time the total number of customers hits a high level. We also discuss the overflow probabilities and the expected number of customers at different loads, respectively.

• The Journal of Information Systems
• /
• v.8 no.2
• /
• pp.111-127
• /
• 1999

본 연구는 글로벌하고 분산된 환경하에서 다중 에이전트에 기반한 공급업자선정과 이를 통한 글로벌 제조를 위한 프레임웍을 구축하고자 하였다. 에이전트 시스템을 위해서는 KQML에 기반을 둔 메시지 전달 시스템에 근거하여 에이전트 통신 계층, 메시지 계층 및 내용서술계층으로 구분하였다. 또한 분산된 객체간에 클라이언트간의 통신을 위해서는 고정된 인터넷 프로토콜의 주소를 없어도 통신이 가능하도록 OLE automation 서버와 ORB 를 이용하여 시스템을 구축하였다. 시스템 구현을 위해서는 스포츠용 신발을 주로 수출하는 중소기업의 업무를 대상으로 하여 정해진 생산계획에 대하여 관련 부품을 생산 납품할 수 있는 외부 공급자를 선정하는 업무에 대하여 프로토 타입을 만들어서 시험하였다. 네트웍 시설이 열악한 동남아 등지에 협력업체를 많이 보유하고 있는 한국의 섬유분야나 신발산업 같은 분야에 있어서 본 시스템의 활용 가능성이 기대된다.