Stability of service under time-of-use pricing

摘要: We consider time-of-use pricing as a technique for matching supply and demand of temporal resources with the goal maximizing social welfare. Relevant examples include energy, computing on cloud platform, charging stations electric vehicles, among many others. A client/job in this setting has window time during which he needs service, particular value obtaining it. assume stochastic model demand, where each job materializes some probability via an independent Bernoulli trial. Given per-time-unit resources, any realized will first try to get served by cheapest available resource its and, failing that, find service at next resource, so on. Thus, natural fluctuations have potential lead cascading overload events. Our main result shows that prices optimally handle expected works well: high probability, when actual is instantiated, system stable jobs very close optimal offline algorithm.