O'Reilly logo

Mathematical Statistics and Stochastic Processes by Denis Bosq

Stay ahead with the world's most comprehensive technology and business learning platform.

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, tutorials, and more.

Start Free Trial

No credit card required

Chapter 11

Poisson Processes — A Probabilistic andStatistical Study

11.1. Introduction

A Poisson process models a sequence of events that occur at random instants.

Let 0 < T1 < T2 < … < Tn < … be the arrival time of a sequence of events such that limn↑ ∞ Tn = + ∞. For reasons that will be apparent later, we set T0 = 0.

Supposing the Tn to be random variables defined on the same probability space, images is said to be a point process.

Equivalently, the process images may be defined, where Nt denotes the number of events that occur in the interval [0, t]. We say that (Nt) is the counting process associated with (Tn).

(Nt) and (Tn) are connected by the following relations:

N0 =T0 = 0;

images

– {Nt = n} = {Tnt < Tn + 1};

– {Ntn} = {Tnt};

– {s < Tnt} = {Ns < nNt}.

Consequently, (Nt) is entirely determined by (Tn), and vice versa.

11.2. The axioms of Poisson processes

To specify the distribution of the process, we make the following hypotheses:

  A1: (Nt, t ≥ 0) has independent increments, i.e. Nt2Nt1 ,…, NtkNtk−1 are independent for all k > 2, 0 ≤ t1 < t2 < … < tk.

  A2: (Nt, t ≥ 0) has stationary increments, i.e. P(Nt + h − Ns + h) = P(NtNs) for 0 ≤ s < t, h > 0.

THEOREM 11.1.– ...

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, interactive tutorials, and more.

Start Free Trial

No credit card required