This paper considers AIMD-based (Additive-Increase Multiplicative-Decrease) congestion control mechanisms that are TCP-compatible (i.e., that compete reasonably fairly with TCP), but that reduce their sending rate less sharply than does TCP in response to a single packet drop. The paper then briefly compares these smoother AIMD-based congestion control mechanisms with TFRC (TCP-Friendly Rate Control), which makes use of equation-based congestion control.

This paper describes the engineering of a new congestion control for TCP motivated by theoretical results [21] which suggest an end-system based flow control protocol can exhibit scalable connection rates, stable operation and a decoupling between the congestion detection and response algorithms. The protocol is designed to be suitable for use in a low-loss and low-delay IP network but it could be adapted for use in current IP networks which have TCP connections that need to scale to high bandwidths on high latency links. It incorporates a packet pacing scheme, a variant of traditional slow-start, and parameter scaling to remove round trip time bandwidth allocation bias. Simulation results, in the context of a low-loss and low-delay IP network, show the strengths and weaknesses of the protocol interconnected with a simple static congestion detection algorithm at routers. It is concluded that the protocol can maintain a low-loss and low-delay packet service model and consistent throughput allocations but that more work is needed to improve link utilizations.

Connection acceptance control is a mechanism which can be used to moderate the load placed on a network by turning away connection requests during times of overload. Traditionally these mechanisms have been implemented by setting up state within a network using signalling protocols, such as the IETF's Resource Reservation Protocol. There have been recent proposals for admission control based on end-systems probing the network to infer network load. These distributed algorithms often have less router state and hence scale more easily. This paper discusses an ECN (Explicit Congestion Notification) probe-based admission control protocol that is fast, scalable and robust. The protocol's performance is then studied through simulation. It is concluded that probe-based admission control is viable in both partitioned and integrated networks, but more research is needed to understand the implications for network policy control. 1.

The Intemet uses a window-based congestion control mechanism in TCP (Transmission Control Protocol). In the literature, there have been a great number of analytical studies on TCP. Most of those studies have focused on the statistical behavior of TCP by assuming a constant packet loss probability in the network. However, the packet loss probability, in reality, changes according to packet transmission rates from TCP connections. In this paper, we explicitly model the interaction between the congestion control mechanism of TCP and the network as a feedback system. Namely, we model the congestion control mechanism of TCP as a dynamic system, where the input to the system is the packet loss probability and the output is the window size. Inversely, we model the network as a dynamic system, where the input is the window size and the output is the packet loss probability. The network is modeled by a M/M/1/m queueing system by assuming an existence of a single bottleneck link. Using our analytic model, the transient behavior of TCP connections is quantitatively evaluated with several numerical examples.

The Internet uses a window-based flow control mechanism in TCP (Transmission Control Protocol). In the literature, there have been a significant number of analytical studies on TCP. Most of those studies, however, have focused on the statistical behavior of TCP by assuming a constant packet loss probability in the network. In our previous work, we have presented an approach for modeling the network as a single feedback system using the fluid flow approximation and the queuing theory. In this paper, by utilizing and extending our previous work, we analyze the steady state behavior and the transient behavior of TCP. We first derive the throughput and the packet loss probability of TCP, and the number of packets queued in the bottleneck router. We then analyze the transient behavior of TCP using a control theoretic approach, showing the influence of the number of TCP connections and the propagation delay on its transient behavior of TCP. Through numerical examples, it is shown that the bandwidth--delay product of a TCP connection significantly affects its stability and transient performance. It is also shown that, contrary to one's intuition, the network becomes more stable as the number of TCP connections and/or the amounts of background traffic increases.