• Journal of KIISE:Computing Practices and Letters
• /
• v.14 no.4
• /
• pp.369-377
• /
• 2008

Today, Ethernet technology is rapidly developing to have a bandwidth of 10Gbps beyond 1Gbps. In such high-speed networks, the existing method that host CPU processes TCP/IP in the operating system causes numerous overheads. As a result of the overheads, user applications cannot get the enough computing power from the host CPU. To solve this problem, the TCP/IP Offload Engine(TOE) technology was emerged. TOE is a specialized NIC which processes the TCP/IP instead of the host CPU. In this paper, we implemented a high-performance, lightweight TCP/IP(HL-TCP) for the TOE and applied it to an embedded system. The HL-TCP supports existing fundamental TCP/IP functions; flow control, congestion control, retransmission, delayed ACK, processing out-of-order packets. And it was implemented to utilize Ethernet MAC's hardware features such as TCP segmentation offload(TSO), checksum offload(CSO) and interrupt coalescing. Also we eliminated the copy overhead from the host memory to the NIC memory when sending data and we implemented an efficient DMA mechanism for the TCP retransmission. The TOE using the HL-TCP has the CPU utilization of less than 6% and the bandwidth of 453Mbps.

• The KIPS Transactions:PartC
• /
• v.10C no.7
• /
• pp.943-954
• /
• 2003

The mobile telecommunication system has been growing exponentially after 1990s due to the high population in a city and the growth of mobile user. In this time, the current mobile system mainly concentrates on the voice communication. However, in the next generation, mobile users want to get very diverse services via mobile terminal such as the Internet access, web access, multimedia communication, and etc. For this reason, the next generation system, such as the UMTS (Universal Mobile Telecommunication Services) system, has to support the packet data service and it will play the major role in the system. By the way, since the Web service is based on TCP, most of the Internet traffic TCP traffic. Therefore, efficient transmission of TCP traffic will take very important role in the performance of packet data service. There are many researches about improving TCP performance over wireless network. In those schemes, the UMTS system adapts the link layer retransmission scheme. However, there are rarely studies about the exact performance of the link layer retransmission scheme in the face of dynamic changes of wireless environment over the UMTS system. The dynamic changes of wireless environment, such as wireless bandwidth, can degrade TCP performance directly. So, in this paper, we simulate and analyze the TCP performance in the UMTS system with dynamic wireless environments. Then, we propose a simple scheme for minimizing TCP performance degradation. As a result of simulation, we can find that when wireless environment is changed dynamically, the probability of TCP timeout is increased, and the TCP performance is degraded very much. In this situation, the proposed simple scheme shows good performance. It saves wireless resources and reduces the degradation of TCP performance without large overhead of the base station.

• Journal of Satellite, Information and Communications
• /
• v.10 no.3
• /
• pp.53-56
• /
• 2015

In this paper, we want provide improved services with faster transmission, IEEE 802.11n was standardized. A-MPDU (Aggregation MAC Protocol Data UNIT) is a vital function of the IEEE 802.11n standard, which was proposed to improve transmission rate by reducing frame transmission overhead. In this paper, we show the problems of TCP retransmission with A-MPDU and propose a solution utilizing the property of TCP cumulative ACK. If the transmission of an MPDU subframe fails, A-MPDU mechanism allows selective re-transmission of failed MPDU subframe in the MAC layer. In TCP traffic transmission, however, a failed MPDU transmission causes TCP Duplicate ACK, which causes unnecessary TCP re-transmission. Furthermore, congestion control of TCP causes reduction in throughput. By supressing unnecessary duplicate ACKs the proposed mechanism reduces the overhead in transmitting redundant TCP ACKs, and transmitting only a HS-ACK with the highest sequence number. By using the RACK mechanism, through the simulation results, it was conrmed that the RACK mechanism increases up to 20% compared the conventional A-MPDU, at the same time, it tightly assures the fairness among TCP flows.

• 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.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.7
• /
• pp.794-798
• /
• 2010

Assuming that a wireless link is mostly used at the network edge and the wireless NIC driver keeps monitoring the error rate of its link, this paper proposes an enhanced TCP congestion control, TCP-L (TCP Link-Aware). TCP-L predicts true congestion losses occurred inside the wired link area by utilizing the wireless link error rate. As a result, it mitigates performance degradation caused from TCP congestion control actions when segments losses occur in a wireless link. Experimental results show that TCP-L provides better performance and fairness in lossy wireless links than existing TCP congestion control schemes. Our approach utilizing the characteristic of the link at TCP could be well adapted to new wireless environments such as Cognitive Radio and ACK-less IEEE 802.11, where a frame may be delivered with a very long delay or lost in the link.

• KIPS Transactions on Computer and Communication Systems
• /
• v.6 no.6
• /
• pp.261-270
• /
• 2017

With the development of smart on-devices and communication technology, demand for non-stop services is increasing. Therefore, the high availability for continuously providing services in the event of system failure has been spotlighted. Meanwhile, because most internet-based services are provided by using TCP, an efficient TCP session recovery technique for providing non-stop services is required. However, the existing TCP session recovery techniques are inefficient because it has a high recovery cost or does not support failover operation, To solve these problems, in this paper, we propose a TCP session recovery technique for high availability in smart on-devices. For this, we first recover the TCP session without re-establish the TCP session by correcting a sequence number and a acknowledgment number. Second, we synchronize the TCP session recovery data between the master and the server, and then we operate the failover operation when master server fails. Finally, we provide the non-stop service to peer by using the virtual IP number and the transmission of GARP (Gratuitous ARP) packet.

• Journal of Institute of Control, Robotics and Systems
• /
• v.19 no.7
• /
• pp.653-660
• /
• 2013

This paper deals with the problem of unfairness among uplink TCP (Transmission Control Protocol) flows associated with frame aggregation employed in IEEE 802.11n WLANs (Wireless Local Area Networks). When multiple stations have uplink TCP flows and transmit TCP data packets to an AP (Access Point), the AP has to compete for channel access with stations for the transmission of TCP ACK (acknowledgement) packets to the stations. Due to this contention-based channel access, TCP ACKs tend to be accumulated in the AP's downlink buffer. We show that the frame aggregation in the MAC (Medium Access Control) layer increases TCP ACK losses in the AP and leads to the serious unfair operation of TCP congestion control. To resolve this problem, we propose the TAC (TCP ACK Compression) mechanism operating at the top of the AP's interface queue. By exploiting the properties of cumulative TCP ACK and frame aggregation, TAC serves only the representative TCP ACK without serving redundant TCP ACKs. Therefore, TAC reduces queue occupancy and prevents ACK losses due to buffer overflow, which significantly contributes to fairness among uplink TCP flows. Also, TAC enhances the channel efficiency by not transmitting unnecessary TCP ACKs. The simulation results show that TAC tightly assures fairness under various network conditions while increasing the aggregate throughput, compared to the existing schemes.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.2B
• /
• pp.130-142
• /
• 2009

In this paper, we propose a self-adaptation framework that selects a TCP variant adapted to current end-to-end path among available TCP variants. There is no single version of TCP that is suitable to all network environments since the causes for performance degradation are different one another according to characteristics of network environments. Thus, determining that which TCP variants should be selected in order to get best performance is very important. To enable adaptation through such determination, we integrate the existing network estimation schemes and some TCP variants into our framework then make light-weight performance knowledge database for TCP selection. Through implementing and evaluating the proposed framework we show that our solution can help TCP get high and stable performance on the various types of network environments by pure end-to-end.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.2B
• /
• pp.124-132
• /
• 2007

Where a wireless LAN and a cellular network coexist, a mobile node has to experience vertical handoffs to move between them. Immediately after the vertical handoffs, TCP must need adaptation latency to adjust its congestion window to the proper size at a newly arrived network to use full of a new end-to-end available bandwidth. Even though SACK TCP has the best performance among other regular TCPs in the previous studies, it still cannot use full of the new available bandwidth quickly due to its inefficient increasing way of congestion window. BIC TCP, that becomes a popular TCP in long fat networks, has great feature working well against vertical handoffs by increasing congestion window exponentially with TCP connection sustained. In this paper, we derive adaptation latency of SACK TCP and BIC TCP numerically, and verify them by simulations. We also find that the shorter adaptation latency of BIC TCP produces higher throughput than SACK TCP on vertical handoffs. Consequently, to get higher performance on vertical handoff situations, we propose to use BIC TCP.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.28 no.4B
• /
• pp.288-297
• /
• 2003

The fast retransmit and fast recovery algorithm of TCP Reno, when multiple packets in the same window are lost, cannot recover them without RTO (Retransmission Timeout). TCP New-Reno can recover multiple lost packets by extending fast recovery using partial acknowledgement. If the retransmitted packet is lost again during fast recovery, however, RTO cannot be avoided. In this paper, we propose an algorithm called "Duplicate Acknowledgement Counting(DAC)" to alleviate this problem. DAC can detect the retransmitted packet loss by counting duplicate ACKs. Conditions that a lost packet can be recovered by loss recovery of TCP Reno, TCP New-Reno and TCP New-Reno using DAC are derived by modeling loss recovery behavior of each TCP. We calculate the loss recovery probability for random packet loss probability numerically, and show that DAC can improve loss recovery behavior of TCP New-Reno.