Optimal Pricing in Multi-Agent System

We have been working in the domain Optimal Pricing in Multi-Agent System. Here, agents are the guys who provides the services to their consumers using the vast servers and generate revenue from the services provided. We considered a system in which there are (N+1) servers. One server is maintained by an agent and the others are maintained by the other agent. We called these servers as tagged and non-tagged respectively. Each arrival has its own value and the servers need to be paid a price for every arrival to be processed which is derived from this client specific distribution. Our goal was to choose the tagged price rate in such a way which maximizes its limiting revenue. Using the literature, we derived the limiting fraction of occupied servers and proved that the empirical occupancy fraction converges to this limiting fraction. Also, we derived the aggregate revenue of the tagged server as a function of the limiting fraction and proved that the experimental results matches with the expected ones through simulation.