Stability and performance analysis of networks supporting services with rate control-could the Internet be unstable?

We consider the stability and performance of a model for networks supporting services that adapt their transmission to the available bandwidth. Not unlike real networks, in our model connection arrivals are stochastic and have a random amount of data to send, so the number of connections in the system changes over time. In turn the bandwidth allocated to, or throughput achieved by, a given connection, may change during its lifetime due to feedback control mechanisms that react to congestion and thus implicitly to the number of ongoing connections. Ideally, for a fixed number of connections, such mechanisms reach an equilibrium typically characterized in terms of its 'fairness' in allocating bandwidth to users, e.g., max-min fair. We prove the stability of such networks when the offered load on each link does not exceed its capacity. We use simulation to investigate the performance, in terms of average connection delays, for various network topologies and fairness criteria. Finally we pose an architectural problem in TCP/IP's decoupling of the transport and network layer from the point of view of guaranteeing connection level stability, which we claim may explain congestion phenomena on the Internet.