• Proceedings of the Korean Information Science Society Conference
• /
• 2002.10e
• /
• pp.502-504
• /
• 2002

기존 라우터시스템에서의 패킷 포워딩 엔진은 최적화된 룩업 알고리즘을 통한 소프트웨어 방식으로 구현되었으나, 처리해야 할 데이터가 폭주하면서 전용화된 룩업칩을 이용한 포워딩 엔진으로 대체되어 가고 있다. 그러나, 이러한 룩업칩 기반 패킷 포워딩 엔진은 가격이 비싼 것은 물론. 급변하는 통신 프로토콜에 적응하기 힘든 단점을 가지게 되었으며, 이로 인해서 최근에는 Network Processor 기반의 패킷 포워딩 엔진에 대한 개발이 활발해지고 있다. 본 연구소에서도 초고속광가입자망 사업에 적용되는 포워딩 엔진을 IBM의 2.5G Network Processor 인 NP4GS3 칩을 이용하고 있으며, GbE／POS／E-PON 등의 인터페이스를 지원하고 있다. 또한 본 사업에서는 가입자망에서의 Layer2 broadcasting시의 트래픽 폭주와 가입자간의 보안 기능 등을 지원하기 위해 VLAN 서비스를 적용할 예정이다. 본 논문에서는 일반적인 VLAN 기능에 대해서 알아보고, IBM NP4GS3에서 지원되는 VLAN 기능을 분석하였으며, 이를 토대로 NP4GS3 기반 패킷 포워딩 엔진에서 VLAN 구현 방안에 대하여 기술하였다.

• Journal of Communications and Networks
• /
• v.3 no.4
• /
• pp.374-382
• /
• 2001

Nowadays, the commonly used table lookup scheme for IP routing is based on the sc-called classless interdomain routing (CIDR). With CIDR, routers must find out the best matching prefix (BMP) for IP packets forwarding, this complicates the IP lookup. Currently, this process is mainly performed in software and several schemes hale been proposed for hardware implementation. Since the IP lookup performance is a major design issue for the new generation routers, in this article we investigate the properties of the routing table and present a new approach for IP lookups, our approach is not based Gn BMP and significantly reduces the complexity, thus the computation cast of existing schemes can be significantly reduced. We also propose an efficient IP lookup algorithm, with which we improve the binary search on prefixes to 30 millions of packets per second (MPPS) and 5,000 route updates/s under the same experiment setup with an even larger routing table.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.8 no.12
• /
• pp.4389-4410
• /
• 2014

Network coding is a promising technology that increases system throughput by reducing the number of packet transmissions from the source node to the destination node in a saturated traffic scenario. Nevertheless, some packets can suffer from end-to-end delay, because of a queuing delay in an intermediate node waiting for other packets to be encoded with exclusive or (XOR). In this paper, we analyze the delay according to packet arrival rate and propose two network coding schemes, iXOR (Intelligent XOR) and oXOR (Optimal XOR) with Markov Decision Process (MDP). They reduce the average delay, even under an unsaturated traffic load, through the Holding-${\chi}$ strategy. In particular, we are interested in the unsaturated network scenario. The unsaturated network is more practical because, in a real wireless network, nodes do not always have packets waiting to be sent. Through analysis and extensive simulations, we show that iXOR and oXOR are better than the Distributed Coordination Function (DCF) without XOR (the general forwarding scheme) and XOR with DCF with respect to average delay as well as delivery ratio.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.10c
• /
• pp.199-201
• /
• 2003

Ad hoc network은 stationary infrastructure의 도움 없이 이동 노드들이 필요 시 network 형태을 구성하여 통신이 이루어지게 하는 network으로 ad hoc network 환경에 맞는 다양한 라우팅 프로토콜들이 개발되었고 크게는 table­driven, on­demand 방식으로 나눌 수 있는데 on­demand 방식의 AODV 프로토콜은 routing overhead가 적다는 장점이 있는 반면 single path로 data forwarding을 진행하여 중간노드의 이동에 의한 path가 broken되는 경우 local routing을 하거나 새로이 source­initialed route rediscovery을 수행하여 전송 delay 및 control traffic overhead 등을 높이는 결과를 발생 시켰다. 본 논문은 single path로 구성되는 AODV 프로토콜의 route failures시 문제점을 보완한 Associativity Based Redundancy path Routing(ABRR) 및 Alternate Redundancy path Routing(ARR) schemes을 제안한다. 첫째, ABRR은 main path상에 있는 각 노드들은 associativity based stable node 정보를 이용하여 path broken 이전에 local redundancy path을 구성하여 path broken시 local routing없이 route을 복구할 수 있게 하고 둘째, ARR은 source­initialed route discovery에 의해 alternate path을 구성하여 ABRR 그리고 local routing에 의해 main route recovery 실패 시 alternate path을 main path로 전환하여 control traffic overhead 및 전송 delay을 줄이게 한다.

• Proceedings of the IEEK Conference
• /
• 2002.07c
• /
• pp.1851-1854
• /
• 2002

The Internet with conventional routing scheme cannot meet user demands driven from drastic growth in the Internet user and various service and traffic type. MPLS(Multi Protocol Label Switching) was introduced to the Internet fur solution to resolve this problem. MPLS is a paradigm to integrate higher layer’s software routing functions including layer-3 routing with layer-2 switching. But, the exponential growth of Internet traffic brings out of label space. One scalable solution to cope with this problem is to introduce flow merge technique, i. e. a group of flows is forwarded using the same label. Specially, IETF(Internet Engineering Task Force) recommends that ATM based MPLS system may include VC merge function, so it is scalable to increase of internet traffic. We implemented the MPLS LER system that includes the look-up and forwarding function in incoming path and VC merging function and limited traffic management function in outgoing path. This paper describes the implementation of the LER’s outgoing parts.

• Proceedings of the IEEK Conference
• /
• 2002.06a
• /
• pp.53-56
• /
• 2002

In this paper, we study the network structure that can guarantee the end-to-end quality of service in the Explicit multicast network. In order to guarantee the end-to-end quality of service, we adopt the end-to-end measurement based admission control (EMBAC) scheme. For the effective measurement of the network state and minimum influence on the current data traffic, the queue structure, which the drop precedence in assured forwarding (AF) Per hop behavior (PHB) of the differentiated service (Diffserv) is applied to, is proposed. Through a simulation, we show that EMBAC can make tile controlled load service in tile AF PHB applied Xcast network, and the qualify of the admitted traffic is guaranteed. In addition, the performance of the EMBAC is changing according to the selling parameter of AF PHB.

• Journal of Communications and Networks
• /
• v.10 no.4
• /
• pp.396-402
• /
• 2008

The use of network coding in wireless networks has been proposed in the literature for energy efficient broadcast. However, the decoding complexity of existing algorithms is too high for low-complexity devices. In this work we formalize the all-to-all information exchange problem and shows how to optimize the transmission scheme in terms of energy efficiency. Furthermore, we prove by construction that there exists O(1) -complexity network coding algorithms for grid networks which can achieve such optimality. We also present low-complexity heuristics for random. topology networks. Simulation results show that network coding algorithms outperforms forwarding algorithms in most cases.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2002.04b
• /
• pp.1261-1264
• /
• 2002

개인 통신 서비스에서 주된 당면 문제중의 하나는 위치를 자주 이동하는 많은 이동 단말기들의 위치를 찾는 것이다. 위치 관리를 성취하기 위하여 많은 네트워크 신호처리 트래픽과 데이터베이스 질의가 요구된다. 다수의 정책들은 이동 단말기의 현재 위치를 저장하기 위하여 위치 레지스터 데이터베이스를 사용하고, 그 위치 레지스터의 고장에 대해 약점을 가지고 있다. 이 논문에서, 우리는 위치 레지스터들의 고장을 포용하는 분산 홈 위치 레지스터를 갖는 결함 포용 전방 포인터 방법을 제안한다. 제안하는 방법의 성공을 분석적 모델로 평가하고, Biaz의 우회 전방 정책, 두개 경로 전방 정책과 성능을 비교한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2002.04b
• /
• pp.1297-1300
• /
• 2002

포워딩 엔진은 외부 인터페이스를 통해서 들어오는 패킷에 대해서 IP 주소를 기반으로 목적지로 향하는 다음 홉을 결정한다. 이러한 고속의 패킷 처리를 위한 포워딩 엔진을 설계함에 있어서 IP 주소 탐색은 중요한 요인이다. 본 논문에서는 검색경로 압축 트라이에 기반한 IP 주소 탐색 알고리즘을 병렬화 하는 기법을 제시한다. 제시된 병렬화를 통해서 IP 주소 탐색의 메모리 접근 횟수글 줄일 수 있으며, 고성능의 패킷 포워딩 엔진에 적용할 수 있다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2013.11a
• /
• pp.400-401
• /
• 2013

Mobility in a network causes link disconnections and link recovery is vital for reliability of a network. A link failure affects all the preceding nodes on a damaged routing path; creates communication delay, throughput degradation, and congestion. This paper proposes link recovery mechanisms in CCN based networks. Packet overhearing is used to update neighboring nodes information. The recovery is done by forwarding node resulting in low control overhead, and better efficiency. The proposed mechanisms increase overall performance of a typical CCN and simulation results show that our proposed scheme works very well in densely populated networks with high mobility.