TCP having passed the slow start phase, these flows


TCP is the most extensively used transport protocol by
Internet applications owing to its robust and reliable connectivity. Hence, the
performance of TCP has a significant impact on the performance of the overall
Internet. The explosive growth in Internet applications over the past few years
has had researchers focus over the aspect of controlling congestion. Though a
lot of research is underway, a major problem that still persists is the “many-flow
problem”. In other words, when the TCP connections sharing the link are
sufficiently large, some of these connections will be subject to frequent TCP
timeouts. Certain applications, such as real time, require long-lived TCP
connections. Delay introduced by these timeouts may significantly degrade network
performance as perceived by end users.


To overcome the problem posed by timeouts, the TCP flows are divided
into two phases:

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

Ø  Initial Phase – referred to as the slow start phase

Ø  Congestion avoidance phase – After having passed the slow start
phase, these flows need to be kept in the congestion avoidance phase, without
being timed out, until the end of the connection.


            When the number of
competing TCP flow increases, the rate of flow needs to be reduced in order to
avoid congestion in the network. During the congestion avoidance phase, the
rate of TCP flow is determined by the ratio CWND/RTT where CWND is the
Congestion Window Size and RTT is the Round- Trip time. Thus, reduction in flow
rate may be achieved by either

Ø  Decreasing the congestion window size or

Ø  Increasing the Round Trip Time


Currently, TCP exercises a closed loop congestion
control mechanism comprising of Randomized feedback in conjunction with active
queue management strategies such as RED and ECN. They are being deployed in an
effort to provide more effective and early congestion indication to adaptive
TCP flows and thereby reduce packet loss. However, a closer look at these
mechanisms reveals a major drawback of frequent timeouts owing to the fact that
TCP ECN (or the conventional TCP) reduces the congestion window size during the
congestion avoidance phase. Also, when the congestion window size falls



below a certain level it fails to trigger the fast retransmit and recovery
algorithm at the sender making it difficult for the TCP flow to recover from
the packet loss.


A solution to the aforementioned problem would be to
increase the Round Trip Time during the congestion avoidance phase and maintain
the window size above a certain threshold even when the fair share of the link
bandwidth is quite small. This methodology of controlling congestion has been
proposed in this project and is referred to as the Sender based Delay Control
or TCP SDC. It derives its name from the fact that congestion is being
controlled by adding delay to the packets transmitted by the sender.


The performance of TCP SDC is studied by conducting
simulations in NS-2. The simulations conducted have validated the fact that TCP
SDC allows many TCP flows to share a link without experiencing frequent
timeouts. In addition, as the window size is maintained above a certain
threshold during congestion avoidance phase, the sender recovers from a packet
loss at a much faster rate than the conventional TCP ECN.