Stochastic Modeling of TCP in Networks
with Abrupt Delay Variations
Alhussein A. Abouzeid, Sumit Roy
An analytical model of TCP (Transport Control Protocol) over an end-to-end path with random abrupt round-trip time (RTT) changes is
presented. Modeling the RTT as a stochastic process, we analytically quantify and compare between the degree of degradation of the steady-
state average throughput and window size due to spurious retransmissions for the different versions of TCP (Reno/NewReno versus Tahoe).
The modeling methodology in this paper is used for evaluating different design alternatives for TCP for highly dynamic networks.
Transport Control Protocol, Wireless Networks, Mobile Ad-Hoc Networks, Satellite Networks, Performance Analysis.
TCP, the Internet Transport Control Protocol , is currently the most widely used transport protocol in packet
networks with primary responsibility for congestion control. In the absence of any explicit information about the
network configuration, TCP achieves its congestion control objective by attempting to drive the network to the point
of full utilization (by increasing the rate at which it releases packets to the network) while continuously monitoring
the network for signs of congestion and lowering the rate if congestion is detected. Packet loss is traditionally1
as a signal of congestion; however, this information to the TCP sender is only implicitly provided by the network,
and hence TCP employs schemes to infer packet loss; either by the expiry of a retransmission timer leading to a