Abstract-Current Internet congestion control protocols operate independently on a per-flow basis. Recent work has demonstrated that cooperative congestion control strategies between flows can improve performance for a variety of applications, ranging from aggregated TCP transmissions to multiple-sender multicast applications. However, in order for this cooperation to be effective, one must first identify the flows that are congested at the same set of resources. In this paper, we present techniques based on loss or delay observations at end hosts to infer whether or not two flows experiencing congestion are congested at the same network resources. Our novel result is that such detection can be achieved for unicast flows, but the techniques can also be applied to multicast flows. We validate these techniques via queueing analysis, simulation, and experimentation within the Internet.In addition, we demonstrate preliminary simulation results that show that the delay-based technique can determine whether two TCP flows are congested at the same set of resources. We also propose metrics that can be used as a measure of the amount of congestion sharing between two flows.

We present a technique for computing upper bounds on the distribution of individual per-session performance measures such as delay and buffer occupancy for networks in which sessions may be routed over several “hops.” Our approach is based on first stochastically bounding the distribution of the number of packets (or cells) which can be generated by each traffic source over various lengths of time and then “pushing” these bounds (which are then shown to hold over new time interval lengths at various network queues) through the network on a per-session basis. Session performance bounds can then be computed once the stochastic bounds on the arrival process have been characterized for each session at all network nodes. A numerical example is presented and the resulting distributional bounds compared with simulation as well as with a point-valued worst-case performance bound.

In this paper we develop a simple analytic characterization of the steady state throughput, as a function of loss rate and round trip time for a bulk transfer TCP flow, i.e., a flow with an unlimited amount of data to send. Unlike the models in [6, 7, lo], our model captures not only the behavior of TCP's fast retransmit mechanism (which is also considered in [6, 7, lo]) but also the effect of TCP's timeout mechanism on throughput.Our measurements suggest that this latter behavior is important from a modeling perspective, as almost all of our TCP traces contained more timeout events than fast retransmit events. Our measurements demonstrate that our model is able to more accurately predict TCP throughput and is accurate over a wider range of loss rates.

Many de nitions of fairness for multicast networks assume that sessions are single-rate, requiring that each m ulticast session transmits data to all of its receivers at the same rate. These de nitions do not account for multi-rate approaches, such a s l a yering, that permit receiving rates within a session to be chosen independently. W e i d e n tify four desirable fairness properties for multicast networks, derived from properties that hold within the max-min fair allocations of unicast networks. We extend the de nition of multicast max-min fairness to networks that contain multi-rate sessions, and show that all four fairness properties hold in a multirate max-min fair allocation, but need not hold in a single-rate max-min fair allocation. We then show t h a t m ulti-rate max-min fair rate allocations can be achieved via intra-session coordinated joins and leaves of multicast groups. However, in the absence of coordination, the resulting max-min fair rate allocation uses link bandwidth ine ciently, and does not exhibit some of the desirable fairness properties. We e v aluate this ine ciency for several layered multi-rate congestion control schemes, and nd that, in a protocol where the sender coordinates joins, this ine ciency has minimal impact on desirable fairness properties. Our results indicate that sender-coordinated layered protocols show promise for achieving desirable fairness properties for allocations in largescale multicast networks.

No abstract