Perhaps the greatest technological innovation of the next several decades will be universal access and utilization of the Internet. Already congestion is becoming a serious impediment to efficient utilization. We introduce a stochastic equilibrium concept for a general mathematical model of the Internet, and demonstrate that the efficient social welfare maximizing stochastic allocation of Internet traffic can be supported by optimal congestion prices. We further demonstrate via simulation modelling that approximately optimal prices can be readily computed and implemented in a decentralized system. We further propose simulation modeling to study the impact of private strategic pricing and public policies.
