• Journal of KIISE:Information Networking
• /
• v.30 no.6
• /
• pp.764-775
• /
• 2003

Previous works for the assured service in the Diffserv network have no sufficient consideration on the fairness of bandwidth share based on RTTs and the target rates of TCP flows. In this paper, in order to solve these problems, we propose a TRA3CM(Target rate and RTT Aware 3 Color Marking) mechanism. The TRA3CM mechanism provides three color marking and fair transmission rates among aggregate flows by considering RTT and target rate simultaneously. In case of higher target rate than bottleneck bandwidth, the TRA3CM mechanism is able to mitigate the RTT effect and provides fair transmission rates. In the results of comparing the performance among existing mechanisms and the TRA3CM, the TRA3CM mechanism was able to mitigate the RTT effect better than the former. The TRA3CM is shown to provide good performance for transmission rates proportional to various target rates.

• Journal of KIISE:Information Networking
• /
• v.31 no.5
• /
• pp.469-481
• /
• 2004

Previous works for the AF(Assured Forwarding) service in the Diffserv network have no sufficient consideration on the fairness of bandwidth share based on RTTs, the target rates, and the impact of UDP against TCP. In this paper, in order to solve these problems, we propose a FDSA(Fair Differentiated Service Architecture) composed of TRA3CM(Target rate and RTT Aware 3 Color Marking) and TRBD(Target Rate Based Dropping) mechanisms. The TRA3CM and TRBD mechanisms provide three color marking and fair transmission rates among aggregate flows by considering RTT, target rate, and UDP flows simultaneously. In the results of comparing the performance among existing mechanisms and the TRA3CM-TRBD, the TRA3CM-TRBD mechanism was able to mitigate the RTT and UDP effect better than the former. The TRA3CM-TRBD is shown to provide good performance for transmission rates proportional to various target rates.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2015.05a
• /
• pp.649-652
• /
• 2015

Previous marking policy for the AF service of TCP traffic in the Diffserv network have no sufficient consideration on the effect of RTT and target rate. In this paper, in order to improve fairness Index by the effect RTT difference of TCP traffic, we propose the modified TSW3CDM(Time Sliding Window Three Color Dynamic Marker) based on average transfer rate estimation and the flow state. The proposed algorithm is dynamic marking policy that do allocate band width in proportion to transmission rate. To evaluate the performance of the proposed algorithm, We accomplished a computer simulation using NS-2. From simulation results, the proposed TSW3CDM algorithm improves fairness index by comparison with TSW3CM.

• The KIPS Transactions:PartC
• /
• v.12C no.4 s.100
• /
• pp.571-580
• /
• 2005

In fast long-distance networks, TCP's congestion control algorithm has the problem of utilizing bandwidth effectively. Several window-based congestion control protocols for high-speed and large delay networks have been proposed to solve this problem. These protocols deliberate mainly three properties : scalability, TCP-friendliness, and RTT-fairness. These protocols, however, cannot satisfy above three properties at the same time because of the trade-off among them This paper presents a new window-based congestion control algorithm, called EM (Exponential Increase/ Multiplicative Decrease), that simultaneously supports all four properties including fast convergence, which is another important constraint for fast long-distance networks; it can support scalability by increasing congestion window exponentially proportional to the time elapsed since a packet loss; it can support RTT-fairness and TCP-friendliness by considering RTT in its response function; it can support last fair-share convergence by increasing congestion window inversely proportional to the congestion window just before packet loss. We evaluate the performance of EIMD and other algorithms by extensive computer simulations.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.7B
• /
• pp.1260-1269
• /
• 2000

TCP 구현의 하나인 Vegas는 RTT 측정값을 바탕으로 혼잡을 인지하며 윈도우 크기 등 혼잡제어를 위한 주요 인자를 결정한다. 이러한 Vegas가 기존 TCP 구현에 비해 우수한 성능과 높은 공평성을 가짐은 여러 논문을 통해 보고되었다. 그런데 우리는 Vegas의 혼잡회피 방안이 TCP 데이터 흐름의 비대칭적 특성을 제대로 반영하지 못하며, 이것이 양쪽 방향 상태를 반영하는 RTT 측정값을 순방향 링크의 상태 해석에 이용하기 때문임을 발견하였다. 우리는 이를 Vegas RTT Ambiguity 문제라 하고, 네트워크 내의 패킷 스케쥴링 방법인 Small Get Priority Queue, TCP Timestamp Options을 이용한 수정 Vegas 구현 등 두 가지 해결 방안을 제시하였다. 우리는 시뮬레이션을 통해 Vegas RTT Ambiguity 문제와 제시한 해결 방안들을 검증하였다.

• Journal of KIISE:Information Networking
• /
• v.32 no.6
• /
• pp.732-744
• /
• 2005

Previous works for the AE(Assured Forwarding) service in the Diffserv network have no sufficient consideration on the fairness of bandwidth share based on the target rate and the effect or RTT and UDP. Also Previous works act like Best-effort service in the UPN(under-Provisioned Network) condition. In this paper, in order to solve these problems, we propose the PFDSA(Proportionally Fair Differentiated Service Architecture) composed of tmTRA3CM(tcp-microflow based Target rate and an Aware Three color Marking), um3CM(udp-microflow based Three color Marker), TRBD(Target Rate Based Dropper), and target rate adjusting function. In the results of comparing the performance among existing mechanisms and the PFDSA, the PFDSA was able to mitigate the RTT and UDP effect better than the former. The PFDSA was shown to provide good performance for transmission rates proportional to various target rates in the UPN condition.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.7B
• /
• pp.613-627
• /
• 2004

In this paper, we propose an Aggregate Fairness Maker (ARM) required for an Edge router to improve fairness of throughput among the flows of Assured Service in DiffServ with different round trip time (RTT) and we propose a user flow Three Color Marker (uf-TCM) as a flow marker that marks packets from the flow as green, yellow, or red. A yellow packet is the packet that consumes loss token in uf-TCM as well as that is demoted green packet in AM due to disobey the aggregate traffic profile. The proposed AFH promotes yellow packet to green packet or demotes green packet to yellow packet through the fair method without per-flow management, and it improves the feirness of throughput among the flows as well as link utilization. A yellow packet and a red packet have the same drop precedence at Core Router in our scheme. So we can use the RIO buffer management scheme. We evaluated the performance of our proposed AFM and the REDP Marker that was proposed to improve fairness without per-flow management. Simulation results show that, compared with the REDP marker, proposed AFM can improve performance of throughput fairness among the flows with different RTT and link utilization under the over-provisioning, exact-provisioning, and under-provisioning network environments at Multiple DiffServ domains as well as at Single DiffServ domain.

• 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 Information Processing Society Conference
• /
• 2002.04b
• /
• pp.1345-1348
• /
• 2002

기존의 TCP/IP 네트워크 상에서 혼잡을 제어하는 일반적인 방법은 end-to-end 피드백 방식이다. 혼잡이 발생했는지의 여부는 수신측의 ACK 신호로 예측할 수밖에 없으므로 RTT(round trip time)가 길어질 경우 혼잡처리 시간이 길어지게 된다. Active Network 기술을 이용한 ACC(Active Congestion Control)의 경우 혼잡이 발생한 라우터에서 전송자에게 직접적으로 정보를 전송하기 때문에 보다 빠른 혼잡처리가 가능하다. 비록 코어 라우터에서 패킷을 누락시키는 방법에 있어 상태정보를 필요로 하는 단점이 있지만 모의 실험 결과 향상된 처리량을 보여주었다. 혼잡처리에 있어 새로운 처리 방식은 AQM(Active Queue Management)이다. 이 방식은 RED 를 기본 모델로 변형된 것이 대부분이며 RED 가 갖는 단점을 개선하고 혼잡처리와 공평성(fairness)을 향상 시키는데 목적이 있다. 본 논문에서는 ACC 의 피드백 메커니즘과 AQM 의 융통성 있는 큐 관리 기법을 이용하여 보다 향상된 혼잡제어 기법을 보이겠다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.28 no.11A
• /
• pp.867-875
• /
• 2003

Current end-to-end congestion control depends only on the information of end points (using three duplicate ACK packets) and generally responds slowly to the network congestion. This mechanism can't avoid TCP global synchronization in which TCP congestion window size is fluctuated during congestion period. Furthermore, if RTT(Round Trip Time) is increased, three duplicate ACK packets are not correct congestion signals because congestion might already disappear and the host may send more packets until it receives three duplicate ACK packets. Recently there are increasing interests in solving end-to-end congestion control using AQM(Active Queue Management) to improve the performance of TCP protocols. AQM is a variation of RED-based congestion control. In this paper, we first evaluate the effectiveness of the current AQM schemes such as RED, CHOKe, ARED, FRED and SRED, over traffic with different rates and over traffic with mixed responsive and non-responsive flows, respectively. In particular, CHOKe mechanism shows greater unfairness, especially when more unresponsive flows exist in a shared link. We then propose a new AQM scheme using CHOKe mechanism, called DQC(Double Queue CHOKe), which uses two FIFO queues before applying CHOKe mechanism to adaptive congestion control. Simulation shows that it works well in protecting congestion-sensitive flows from congestion-causing flows and exhibits better performances than other AQM schemes. Also we use partial state information, proposed in LRURED, to improve our mechanism.