We propose an optimization approach to flow control where the objective is to maximize the aggregate source utility over their transmission rates. We view network links and sources as processors of a distributed computation system to solve the dual problem using gradient projection algorithm. In this system sources select transmission rates that maximize their own benefits, utility minus bandwidth cost, and network links adjust bandwidth prices to coordinate the sources' decisions. We allow feedback delays to be different, substantial and time-varying, and links and sources to update at different times and with different frequencies. We provide asynchronous distributed algorithms and prove their convergence in a static environment. We present measurements obtained from a preliminary prototype to illustrate the convergence of the algorithm in a slowly time-varying environment.

Distributed Algorithmic Mechanism Design (DAMD) combines theoretical computer science's traditional focus on computational tractability with its more recent interest in incentive compatibility and distributed computing. The Internet's decentralized nature, in which distributed computation and autonomous agents prevail, makes DAMD a very natural approach for many Internet problems. This paper first outlines the basics of DAMD and then reviews previous DAMD results on multicast cost sharing and interdomain routing. The remainder of the paper describes several promising research directions and poses some specific open problems.

We propose a duality model of congestion control and apply it to understand the equilibrium properties of TCP and active queue management schemes. Congestion control is the interaction of source rates with certain congestion measures at network links. The basic idea is to regard source rates as primal variables and congestion measures as dual variables, and congestion control as a distributed primal-dual algorithm carried out over the Internet to maximize aggregate utility subject to capacity constraints. The primal iteration is carried out by TCP algorithms such as Reno or Vegas, and the dual iteration is carried out by queue management such as DropTail, RED or REM. We present these algorithms and their generalizations, derive their utility functions, and study their interaction.

In this paper we study a previously developed linearized model of TCP and AQM. We use classical control system techniques to develop controllers well suited for the application. The controllers are shown to have better theoretical properties than the well known RED controller. We present guidelines for designing stable controllers subject to network parameters like load level, propogation delay etc. We also present simple implementation techniques which require a minimal change to RED implementations. The performance of the controllers are verified and compared with RED using ns simulations. The second of our designs, the Proportional Integral (PI) controller is shown to outperform RED significantly.

We present a framework for designing end-to-end congestion control schemes in a network where each user may have a different utility function and may experience non-congestion-related losses. We first show that there exists an additive increase-multiplicative decrease scheme using only end-to-end measurable losses such that a socially-optimal solution can be reached. We incorporate round-trip delay in this model, and show that one can generalize observations regarding TCP-type congestion avoidance to more general window flow control schemes. We then consider explicit congestion notification (ECN) as an alternate mechanism (instead of losses) for signaling congestion and show that ECN marking levels can be designed to nearly eliminate losses in the network by choosing the marking level independently for each node in the network. While the ECN marking level at each node may depend on the number of flows through the node, the appropriate marking level can be estimated using only aggregate flow measurements, i.e., per-flow measurements are not required. 1

Under the assumption that queueing delays will eventually become small relative to propagation delays, we derive stability results for a fluid flow model of end-to-end Internet congestion control. The theoretical results of the paper are intended to be decentralized and locally implemented: each end system needs knowledge only of its own roundtrip delay. Criteria for local stability and rate of convergence are completely characterized for a single resource, single user system. Stability criteria are also described for networks where aH users share the same roundtrip delay. Numerical experiments investigate extensions to more general networks. Through simulations, we are able to evaluate the relative importance of queueing delays and propagation delays on network stability. Finally, we suggest how these results may be used to design network resources.

We consider a resource allocation problem where individual users wish to send data across a network to maximize their utility, and a cost is incurred at each link that depends on the total rate sent through the link. It is known that as long as users do not anticipate the effect of their actions on prices, a simple proportional pricing mechanism can maximize the sum of users' utilities minus the cost (called aggregate surplus). Continuing previous efforts to quantify the effects of selfish behavior in network pricing mechanisms, we consider the possibility that users anticipate the effect of their actions on link prices. Under the assumption that the links' marginal cost functions are convex, we establish existence of a Nash equilibrium. We show that the aggregate surplus at a Nash equilibrium is no worse than a factor of 4 # 2 - 5 times the optimal aggregate surplus; thus, the efficiency loss when users are selfish is no more than approximately 34%.

Weconsider a service provider (SP) who provides access to a communication network or some other form of on-line services. Users access the network and initiate calls that belong to a set of diverse service classes, differing in resource requirements, demand pattern, and call duration.

We describe how a packet network with a simple pricing mechanism and no connection acceptance control may be used to carry a telephony-like service with low packet loss and some call blocking. The packet network uses packet marking to indicate congestion and endsystems are charged a fixed small amount per mark received. The end-systems are thus provided with the information and the incentive to use the packet network efficiently. We demonstrate that algorithms embedded in the end-systems are able to synthesize a telephony-like service by blocking calls at times when the load in the packet network is high. 1.

This paper describes a framework for admission control for a packet-based network where the decisions are taken by edge devices or end-systems, rather than resources within the network. The decisions are based on the results of probe packets that the end-systems send through the network, and require only that resources apply a mark to packets in a way that is load dependent. One application example is the Internet, where marking information is fed back via an ECN bit, and we show howthis approach allows a rich QoS framework for ows or streams. Our approach allows networks to be explicitly analysed, and consequently engineered.