• Journal of Communications and Networks
• /
• v.4 no.2
• /
• pp.148-157
• /
• 2002

This paper investigates schemes to improve TCP performance in multipath forwarding networks. In multipath routing, packets to the same destination are sent to multiple next-hops in either packet-level or flow-level forwarding mode. Effective bandwidth is increased since we can utilize unused capacity of multiple paths to the destination. In packet-level multipath forwarding networks, TCP performance may not be enhanced due to frequent out-of-order segment arrivals at the receiver because of different delays among paths. To overcome this problem, we propose simple TCP modifications. At the sender, the fast retransmission threshold is adjusted taking the number of paths into consideration. At the receiver, the delayed acknowledgment scheme is modified such that an acknowledgment for an out-of-order segment arrival is delayed in the same way for the in-order one. The number of unnecessary retransmissions and congestion window reductions is diminished, which is verified by extensive simulations. In flow-level multipath forwarding networks, hashing is used at routers to select outgoing link of a packet. Here, we show by simulations that TCP performance is increased in proportion to the number of paths regardless of delay differences.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.7 no.4
• /
• pp.741-746
• /
• 2012

Transmission control protocol(TCP) widely used as a transport protocol in the Internet includes a loss recovery function that detects and recovers packet losses by retransmissions. The loss recovery function consists of the two algorithms; fast retransmit and fast recovery. There have been researches to avoid nonnecessary retransmission timeouts (RTOs), which leads to selective acknowledgement (SACK) option and limited transmit scheme that are standardized by IETF (Internet Engineering Task Force). Recently, a method that covers the case in which a retransmitted packet is lost again has been propsed. The method, however, is not proved in terms of the additive increase multiplicative decrease (AIMD) principle of TCP congestion control. In this paper, therefore, we analyzed the method in terms of the principle by ns-simulations.

• ETRI Journal
• /
• v.27 no.1
• /
• pp.1-12
• /
• 2005

We propose a heuristic buffer management scheme that uses both positive and negative acknowledgments to provide scalability and reliability. Under our scheme, most receiver nodes only send negative acknowledgments to their repair nodes to request packet retransmissions while some representative nodes also send positive acknowledgments to indicate which packets can be discarded from the repair node's buffer. Our scheme provides scalability because it significantly reduces the number of feedbacks sent by the receiver nodes. In addition, it provides fast recovery of transmission errors since the packets requested from the receiver nodes are almost always available in their buffers. Our scheme also reduces the number of additional retransmissions from the original sender node or upstream repair nodes. These features satisfy the original goal of treebased protocols since most packet retransmissions are performed within a local group.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.133.3-133
• /
• 2001

This paper analyses the real-time performance of Ethernet based intranet whether it is applicable to the real-time network. Unpredictability of transmission delay by collision-delay-retransmission mechanism in CAMA/CD(Carrier Sense Multiple Access with Collision Detect) of Ethernet is the major reason making hard to apply to real-time system. Both retransmission mechanism of TCP(Transmission Control Protocol) for reliability and sliding windows algorithm for high utilization make hard to predict transmission delay. Because real-time control network require fast responsibility and bustle of short-periodic messages, global-clock for collision avoidance and UDP(User Datagram Protocol) for high utilization of network are used. The mathematical models for time-delay that can be occured between ...

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.7B
• /
• pp.650-659
• /
• 2004

Domestic wireless network environment is changing rapidly while adapting to meet service requirements of users and growth of market. As a result, reliable data transmission using TCP is also expected to increase. Since TCP assumes that it is used in wired networt TCP suffers significant performance degradation over wireless network where packet losses are not always result of network congestion. Especially RTO imposes a great performance degradation of TCP. In this paper, we propose DAC$^{＋}$ and EFR in order to prevent performance degradation by quickly detecting and recovering loss without RTO during fast recovery. Compared with TCP NewReno, proposed scheme shows improvements in steady-state in terms of higher fast recovery Probability and reduced response time.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.7B
• /
• pp.675-684
• /
• 2004

There have been several efforts to avoid unnecessary retransmission timeouts (RTOs), which is the main cause for TCP throughput degradation. Unnecessary RTOs can be classified into three groups according to their cause. RTOs due to multiple packet losses in the same window for TCP Reno, the most prevalent TCP version, can be avoided by TCP NewReno or using selective acknowledgement (SACK) option. RTOs occurring when a packet is lost in a window that is not large enough to trigger fast retransmit can be avoided by using the Limited Transmit algorithm. In this Paper, we comparatively analyze these schemes to cope with unnecessary RTOs by numerical analysis and simulations. On the basis of the results in this paper, TCP performance can be quantitatively predicted from the aspect of loss recovery probability. Considering that overall performance of TCP is largely dependent upon the loss recovery performance, the results shown in this paper are of great importance.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2007.11a
• /
• pp.978-981
• /
• 2007

TCP(Transmission Control Protocol)는 흐름제어와 오류제어 그리고 혼잡제어 등을 통해 목적지까지 효율적으로 데이터를 전송하는 프로토콜이다. 최근 라우터 성능 및 라우팅 알고리즘 등의 발달로 인해 패킷이 비순차적으로 전송되는 패킷 재배치(packet reordering)가 빈번한 발생하여 TCP 성능 저하의 주요 원인이 되고 있다. 패킷 재배치가 TCP 로 하여금 불필요한 재전송(Spurious Retransmission) 및 혼잡제어를 수행하도록 하기 때문이다. 본 논문에서는 TCP 의 신속 재전송(Fast Retransmission)의 중복 응답 임계값(dupthresh)를 동적으로 조절함으로써, 불필요한 재전송 및 혼잡제어의 발생횟수를 효율적으로 줄이는 TCP-DAD(TCP Dynamically Adjusted Dupthresh)기법을 제안한다. 본 논문은 NS-2를 이용한 시뮬레이션을 통해 TCP-DAD 가 다른 TCP 메커니즘보다 평균적으로 약 34%의 성능 향상을 보임을 증명하였다.

• Journal of Communications and Networks
• /
• v.10 no.1
• /
• pp.98-107
• /
• 2008

In mobile ad hoc networks, packet loss is unavoidable due to MAC contention, link failure or the inherent characteristics of wireless link. Since TCP relies on the timely reception of TCP ACK packets to progress the transmission of the TCP DATA packets, ACK loss obviously affects the performance due to two main problems: (a) Frequent occurrence of spurious retransmissions caused by timeout events and (b) impairment of the fast retransmit mechanism caused by the lack of a sufficient number of duplicate ACK packets. In particular, since most reactive routing protocols force the packets buffered over a path to be discarded while performing a route recovery, the performance degradation becomes more serious due to such ACK loss. In this paper, therefore, TCP with two piggybacking schemes (called TCP-pgy) is proposed in order to resolve the above-mentioned problems over reactive routing protocols. Through extensive simulations using the ns-2 simulator, we prove that our proposed schemes contribute to TCP performance improvements.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2004.05a
• /
• pp.1417-1420
• /
• 2004

The combination of Fast-Handoff and Hierarchical Mobile IPv6 (F-HMIPv6) allows the anticipation of the layer3 handoff such that data traffic can be efficiently redirected to the mobile node's new location before it moves there. However, after moving to the new Access Router (NAR), if the mobile node (MN) sends the Local Binding Update (LBU) to the Mobility Anchor Point (MAP) before receiving all of the buffered packet from the NAR, the MN may receive the general packet from the MAP. That is, the MN may simultaneously receive two types of packet which has different sequence number. These cause the confusion in packet order, and the MN sends the dup ack for the packet retransmission to the CN. It results in the degradation of the TCP performance. Therefore, we propose the scheme for minimizing the out-of-sequence packet in F-HMIPv6.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.30 no.3B
• /
• pp.130-137
• /
• 2005

Wireless communication environment is changing rapidly as we use new wireless communication technology such as WiBro to access high speed Internet. As a result, reliable data transmission using TCP is also expected to increase. Since TCP assumes that it is used in wired network, TCP suffers significant performance degradation over wireless network where packet losses are related to non-congestion loss. Especially RTO imposes a great performance degradation of TCP. In this paper, we analyze the loss recovery probabilities based on previous researches, and use simulation results of our algorithm to show that it prevents performance degradation by quickly detecting and recovery losses without RTO during fast recovery.