Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Klenke defines the Poisson process as a family of non-negative integer valued random variables with independent increments, whose increments are distributed according to the Poisson distribution (Definition 5.33). He proceeds to prove that the Poisson process can be characterized by five properties (Theorem 5.34, the five properties P1-P5). In order to prove the direction "If a process satisfies P1-P5, it is a Poisson process" (proof), he appeals to the Poisson approximation theorem. This theorem assumes a collection of numbers $\left(p_{n,k}\right)\subseteq\left[0,1\right]$ that satisfy two conditions: $$\left(i\right)\space\space \lim_{n\rightarrow\infty}\sum_{k=1}^\infty p_{n,k}\in\left(0,\infty\right)$$ $$\left(ii\right)\lim_{n\rightarrow\infty}\sum_{k=1}^\infty p_{n,k}^2=0$$ Accordingly, in the proof of Theorem 5.34 Klenke defines a collection of numbers $\left(p_{n,k}\right)=\left(p_n\right)$ and shows that they satisfy condition $\left(i\right)$ (this takes place five lines from the end, where it is shown that $\alpha t=\lim_{n\rightarrow\infty}p_n2^n=\lim \sum_{k=1}^{2^n} p_n$). However, condition $\left(ii\right)$ isn't verified, which brings me to wonder: why does condition $\left(ii\right)$ hold?

share|cite|improve this question
up vote 2 down vote accepted

We have, as $p_{n,k}=\frac{\lambda}n$ when $k\leq n$ and $0$ otherwise, $$\sum_{k=1}^{+\infty}p_{n,k}^2=\sum_{k=1}^n\frac{\lambda^2}{n^2}=\frac{\lambda^2}n,$$ which gives the wanted result, unless I am missing something.

share|cite|improve this answer
Thanks, but what's $\lambda$? Are you referring to Theorem 5.34's proof or to the comment preceding the Poisson approximation theorem? I'm interested in understanding the former. – Evan Aad Sep 9 '12 at 21:43
How is $p_{n,k}$ defined in the former? – Davide Giraudo Sep 9 '12 at 21:58
It is defined as $p_{n,k}:=\mathrm{P}\left(N_{\left(0,t/2^n\right]}\geq1\right)$ if $k\in\left\{1,2,\dots,2^n\right\}$ and $p_{n,k}:=0$ otherwise. – Evan Aad Sep 10 '12 at 3:08
OK, i think i get it now. We define $\lambda_n:=2^np_n$, so that $p_n=\lambda_n/2^n$. It is shown in the proof that $\lim_{n\rightarrow\infty}\lambda_n=\alpha t\in\mathbb{R}$. So $$\lim_{n\rightarrow\infty}\sum_{k=1}^{2^n} p_{n,k}^2=\lim_{n\rightarrow\infty}\frac{\lambda_n^2}{2^n}=0$$ Thanks, Davide! – Evan Aad Sep 10 '12 at 5:30

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.