In recent years, various real-time applications in the Internet have been emerging with rapid increase of the network bandwidth. A real-time application traditionally uses either UDP (User Datagram Protocol) or TCP (Transmission Control Protocol) as its transport layer protocol. However, using either UDP or TCP is insufficient for most real-time applications because of lacking a smooth rate control mechanism or suffering a significant transfer delay. In the literature, several transport-layer communication protocols for real-time applications have been proposed. In this paper, among these transport-layer communication protocols, we focus on TFRC (TCP-Friendly Rate Control). Steady state performances of TCP and TFRC connections such as throughput and fairness have been throughly investigated by many researchers using simulation experiments. However, transient state properties of TCP and TFRC connections such as stability and responsiveness have not been investigated. In this paper, we therefore analyze both steady state and transient state performances of TCP and TFRC connections using a control theoretic approach. We frist model TFRC and TCP connections with different propagation delays and the active queue management mechanism of RED (Random Early Detection) router as independent discrete-time systems. By combining these discrete-time systems, we analyze steady state performance of TCP and TFRC connections such as throughput, transfer delay, and packet loss probability. We also analyze transient state performance of TCP and TFRC connections using linearization of discrete-time systems around their equilibrium points.
In the current Internet, most of the traffic is transmitted by TCP (Transmission Control Protocol). In our previous work, we have proposed a modeling approach for the entire network, including TCP congestion control mechanisms operating at source hosts and the network seen by TCP connections, as a single feedback system. However, our analytic model is limited to a simple network, where TCP connections have the identical propagation delay. In this paper, we therefore extend our analytic approach to a more generic network, where multiple TCP connections are allowed to have different
propagation delays. We derive the packet loss probability in the network, the throughput and the average round-trip time of each TCP connection in steady state. By presenting several numerical examples, we quantitatively investigate how the fairness among TCP connections is degraded when multiple TCP connections with different propagation delays share the single bottleneck link.