• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.12B
• /
• pp.1052-1062
• /
• 2004

Recent studies have pointed out that existing high-speed TCP protocols have a severe unfairness and TCP friendliness problem. As the congestion window achieved by a high-speed TCP connection can be quite large, there is a strong possibility that the sender will transmit a large burst of packets. As such, the current congestion control mechanisms of high-speed TCP can lead to bursty traffic flows in hi인 speed networks, with a negative impact on both TCP friendliness and RTT unfairness. The proposed solution to these problems is to evenly space the data sent into the network over an entire round-trip time. Accordingly, the current paper evaluates this approach with a high bandwidth-delay product network and shows that pacing offers better TCP friendliness and fairness without degrading the bandwidth scalability.

• Proceedings of the Korea Society of Information Technology Applications Conference
• /
• 2005.11a
• /
• pp.211-214
• /
• 2005

There are emerging many eScience applications. More and more scientists want to collaborate on their investigation with international partners without space limitation by using these applications. Since these applications have to analyze the massive raw data, scientists need to send and receive the data in short time. So today's network related requirement is high speed networking. The key point of network performance is transport protocol. We can use TCP and UDP as transport protocol but we use TCP due to the data reliability. However, TCP was designed under low bandwidth network, therefore, general TCP, for example Reno, cannot utilize the whole bandwidth of high capacity network. There are several TCP variants to solve TCP problems related to high speed networking. They can be classified into two groups: loss based TCP and delay based TCP. In this paper, I will compare two approaches of TCP variants and propose a hybrid approach for high speed networking.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.52 no.8
• /
• pp.10-17
• /
• 2015

A satellite communication system is one of viable solutions for Internet applications running in wide areas. However, the performance of TCP can be seriously degraded in the satellite networks due to long round-trip time (RTT) and high bit error rate (BER) over satellite links. Therefore, a performance enhancing proxy(PEP) based TCP splitting connection scheme is used in the satellite link to improve the TCP performance. In this paper, we implement PEP testbed and conduct experiment to evaluate the performance of TCP splitting connection by comparing with high-speed TCP solutions in various environments. In our experimental environment, we consider multiple connections, high packet loss, and limited bandwidth. The experiment results show that PEP improves the TCP throughput than high-speed TCP variants in various environments. However, there is no improvement of the TCP throughput with the limited bandwidth because there is packet loss caused by both the congestion and the channel error.

• The KIPS Transactions:PartC
• /
• v.8C no.3
• /
• pp.335-344
• /
• 2001

최근 인터넷 환경이 반도체, 광통신 그리고 차세대 인터넷 기술의 발달로 고성능화 되어가고 있다. 따라서 고성능 인터넷을 위한 TCP의 성능 향상 연구가 매우 중요해졌다. 그러나 기존 TCP는 수신위도 버퍼의 물리적 크기에 의하여 최대 전송 성능과 대역폭 탐색 기능이 제한을 받는 구조적인 문제점을 갖고 있다. 본 논문에서는 이를 해결하기 위하여 수신 호스트에 가상 윈도 개념을 도입하였다. 이는 송신 호스트가 RTT 동안 균일하게 세그먼트를 분산시켜서 패킷을 전송할 때 세그먼트 간격 시간 동안 수신 호스트의 처리 능력을 가상윈도로 나타내는 것이다. 따라서 가상 윈도의 크기는 수신 호스트의 성능에 비례하기 때문에 수신 호스트가 고성능일 경우 TCP의 전송 능력 성능이 더 높아질 수 있다. 초고속 인터넷일 경우 제안 알고리즘이 기존 TCP보다 전송능력에 있어 1.5∼5배 개선되는 것을 네트워크 시뮬레이션인 NS2를 이용하여 확인하였다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.5 no.4
• /
• pp.706-715
• /
• 2001

TCP typically offers reasonable end-to-end performance to users regardless of the bandwidth and error characteristics of particular network technology, The robustness of TCP has contributed to its success in the internet environment. The role of communications is already well established in the office environment. With the advent of cheap, affordable broadband communications and the increasing complexity of consumer goods, it seems natural to extend the network into homes. In-home networking means a high-speed communication among the digital appliances within a home. Introduction of application over high-speed home network using TCP/IP protocol is increasing. The integrated environment of internet and home network is demanding as well. We have validated TCP model over high speed home network environment, investigated the throughput behavior of TCP over IEEE 1399 home networks, and evaluated a potential solution for high performance of TCP over IEEE 139t home networks. The simulation model has produced several interesting results in the performance of TCP over IEEE 1394 home network.

• Proceedings of the IEEK Conference
• /
• 1998.10a
• /
• pp.1029-1032
• /
• 1998

In this paper, a comparative study is presented to evaluate the addition architecture of the high-speed checksum module in TCP/IP processing. In order to speed up TCP/IP processing, H/W implementation offers concurrent and parallel processing to yield high speed computation, with respect to S/W implementation. This research aims at comparing two addition architectures of checksum module, which is the major botteleneck in TCP/IP processing. The 16-bit and 8-bit byte-by-byte addition architecture are implemented by the full custom design, and compared, in analytical and experimental manner, from standpoint of space and performance. For LG $0.6\mu\textrm{m}$ TLM process, the 8-bit addition implementation requires the area, 1.3 times larger than the 16-bit one, and it operates at 80MHz while the 16-bit one runs by 66MHz.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2007.06a
• /
• pp.83-86
• /
• 2007

The necessity of the high speed exclusive use server that connect digital measure instrument flag to internet and CPU use rate of measure flag uses other imbedded operating system with many general servers to verification & proofreading system in remotion. This research is objective that embody high speed network module for Site-Based remote verification & proofreading system and through internet digital measure instrument flag to remote verification & proofreading, TCP/IP Offload Engine processing and improved that is Imbedded TCP/IP stack for high speed networking.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.6
• /
• pp.1077-1082
• /
• 2007

The necessity of the high speed exclusive use server that connect digital measure instrument flag to internet and CPU use rate of measure flag uses other imbedded operating system with many general servers to verification & proofreading system in remotion. This research is objective that embody high speed network module for Site-Based remote verification & proofreading system and through internet digital measure instrument flag to remote verification & proofreading, TCP/IP Offload Engine processing and improved that is Imbedded TCP/IP stack for high speed networking.

• Journal of Communications and Networks
• /
• v.12 no.3
• /
• pp.275-284
• /
• 2010

The current multihome-aware protocols (like stream control transmission protocol (SCTP) or parallel TCP for concurrent multipath data transfer (CMT) are not designed for high-capacity and large-latency networks; they often have performance problems transferring large data files over shared long-distance wide area networks. It has been shown that SCTP-CMT is more sensitive to receive buffer (rbuf) constraints, and this rbuf-blocking problem causes considerable throughput loss when multiple paths are used simultaneously. In this research paper, we demonstrate the weakness of SCTP-CMT rbuf constraints, and we then identify that rbuf-blocking problem in SCTP multihoming is mostly due to its loss-based nature for detecting network congestion. We present a simulation-based performance comparison of FAST TCP versus SCTP in high-speed networks for solving a number of throughput issues. This work proposes an end-to-end transport layer protocol (i.e., FAST TCP multihoming as a reliable, delaybased, multihome-aware, and selective ACK-based transport protocol), which can transfer data between a multihomed source and destination hosts through multiple paths simultaneously. Through extensive ns-2 simulations, we show that FAST TCP multihoming achieves the desired goals under a variety of network conditions. The experimental results and survey presented in this research also provide an insight on design decisions for the future high-speed multihomed transport layer protocols.

• Journal of KIISE
• /
• v.42 no.9
• /
• pp.1162-1174
• /
• 2015

These days, the networks have exhibited HBDP (High Bandwidth Delay Product) characteristics. The legacy TCP slowly increases the size of the congestion window and drastically decreases the size of a congestion window. The legacy TCP has been found to be unsuitable for HBDP networks. TCP mechanisms for solving the problems of the legacy TCP can be categorized into the loss-based TCP and the delay-based TCP. Most of the TCP mechanisms use the standard slow start phase, which leads to a heavy packet loss event caused by the overshoot. Also, in the case of congestion avoidance, the loss-based TCP has shown problems of wastage in terms of the bandwidth and RTT (Round Trip Time) fairness. The delay-based TCP has shown a slow increase in speed and low occupancy of the bandwidth. In this paper, we propose a new scheme for improving the over shoot, increasing the speed of the bandwidth and overcoming the bandwidth occupancy and RTT fairness issues. By monitoring the buffer condition in the bottleneck link, the proposed scheme does congestion control and solves problems of slow start and congestion avoidance. By evaluating performance, we prove that our proposed scheme offers better performance in HBDP networks compared to the previous TCP mechanisms.