Improving Tcp Performance Over Heterogeneous Networks

Transmission Control Protocol (TCP) is considered one of the most important protocolsin the Internet. An important mechanism in TCP is the congestion controlmechanism which controls TCP sending rate and makes TCP react to congestionsignals. Nowadays in heterogeneous networks, TCP may work in networks with somelinks that have lossy nature (wireless networks for example). TCP treats all packetloss as if they were due to congestion. Consequently, when used in networks thathave lossy links, TCP reduces sending rate aggressively when there are transmission(non-congestion) errors in an uncongested network. One solution to the problem is to discriminate between errors; to deal with congestionerrors by reducing TCP sending rate and use other actions for transmissionerrors. In this work we investigate the problem and propose a solution using anend-to-end error discriminator. The error discriminator will improve the currentcongestion window mechanism in TCP and decide when to cut and how much tocut the congestion window. We have identified three areas where TCP interacts with drops: congestion windowupdate mechanism, retransmission mechanism and timeout mechanism. All ofthese mechanisms are part of the TCP congestion control mechanism. We proposechanges to each of these mechanisms in order to allow TCP to cope with transmissionerrors. We propose a new TCP congestion window action (CWA) for transmissionerrors by delaying the window cut decision until TCP receives all duplicate acknowledgmentsfor a given window of data (packets in flight). This will give TCP a clearimage about the number of drops from this window. The congestion window size isthen reduced only by number of dropped packets. Also, we propose a safety mechanismto prevent this algorithm from causing congestion to the network by usingan extra congestion window threshold (tthresh) in order to save the safe area wherethere are no drops of any kind. The second algorithm is a new retransmission actionto deal with multiple drops from the same window. This multiple drops action(MDA) will prevent TCP from falling into consecutive timeout events by resendingall dropped packets from the same window. A third algorithm is used to calculatea new back-off policy for TCP retransmission timeout based on the network?s availablebandwidth. This new retransmission timeout action (RTA) helps relating thelength of the timeout event with current network conditions, especially with heavytransmission error rates. The three algorithms have been combined and incorporated into a delay basederror discriminator. The improvement of the new algorithm is measured along withthe impact on the network in terms of congestion drop rate, end-to-end delay, averagequeue size and fairness of sharing the bottleneck bandwidth. The results show thatthe proposed error discriminator along with the new actions toward transmissionerrors has increased the performance of TCP. At the same time it has reduced theload on the network compared to existing error discriminators. Also, the proposederror discriminator has managed to deliver excellent fairness values for sharing thebottleneck bandwidth. Finally improvements to the basic error discriminator have been proposed byusing the multiple drops action (MDA) for both transmission and congestion errors. The results showed improvements in the performance as well as decreases in thecongestion loss rates when compared to a similar error discriminator.