Estimating the number of component processes of a superimposed process
Tapan Kumar Nayak · Biometrika · 1991
We consider a superposition of an unknown number of independent homogeneous Poisson processes with unknown rates and suppose that when an event occurs, the component process in which it occurred can be identified. The number of component processes and their rates are to be estimated by observing the system for some time. In the general case, it is difficult to estimate all of the parameters but we derive the best unbiased estimators of some parametric functions. Inferences for the special case where all the rates are equal are discussed in detail.