ABSTRACT
The model we study in this chapter is a generalization of the model introduced in Section 1.2 of the introductory chapter. We observed there that many of the salient features of the optimal arrival-rate model with deterministic reward and linear waiting cost do not depend on the system being an M/M/1 queue operating in steady state. For example, the individually optimal arrival rate λe is an upper bound on the socially optimal arrival rate λs for any queueing system satisfying the following conditions:
1. W (λ) is strictly increasing in 0 ≤ λ < µ ; 2. W (λ) ↑ ∞ as λ ↑ µ ; 3. W (0) = 1/µ .
