SUM OF A RANDOM NUMBER OF INDEPENDENT RANDOM VARIABLES

Properties of the Sum of a Random Number of Independent Ran­ dom Variables

Let Xl , X2, . . . be identically distributed random variables with mean E[X] and variance var(X). Let N be a random variable that takes nonnegative in­

teger values. We assume that all of these random variables are independent, and we consider the sum

Y = X1 + · · · + XN·

Then:

E[Y] = E[N] E[X] .

var(Y) = E[N] var(X) + (E[X] ) var(N).

We have

lvly (s) = l\,fN ( log Mx (s)) .

Equivalently, the transform lvly (s) is found by starting with the trans­ form MN(S) and replacing each occurrence of e8 with lvIx (s).

Example 4.34. A remote village has three gas stations. Each gas station is open on any given day with probability 1/2. independent of the others. The amount of gas available in each gas station is unknown and is uniformly distributed between 0 and 1000 gallons. We wish to characterize the probability law of the total amount of gas available at the gas stations that are open.

The number N of open gas stations is a binomial random variable with p= 1/2 and the corresponding transform is

The transform !Ylx {s) associated with the amount of gas available in an open gas station is

elOOOs _ 1

l\/x { s) =

lOs 0 0 .

Sec. 4.5 Sum of a Random Number of Independent Random Variables 243

The transform associated with the total amount Y available is the same as MN (S), except that each occurrence of eS is replaced with Mx (s), i.e.,

My (s) =8 ( ( 1 + ))

1 elOOOS _ 1

1000s

Example 4.35. Sum of a Geometric Number of Independent Exponen­

tial Random Variables. Jane visits a number of bookstores, looking for Great Expectations. Any given bookstore carries the book with probability p, indepen­ dent of the others. In a typical bookstore visited, Jane spends a random amount of time, exponentially distributed with parameter A, until she either finds the book or she determines that the bookstore does not carry it. We assume that Jane will keep visiting bookstores until she buys the book and that the time spent in each is independent of everything else. We wish to find the mean, variance, and PDF of the total time spent in bookstores.

The total number N of bookstores visited is geometrically distributed with pa­ rameter p. Hence, the total time Y spent in bookstores is the sum of a geometrically distributed number N

of independent exponential random variables Xl, X2, We have

E[Y] = E[N] E [ X ] = .

Using the formulas for the variance of geometric and exponential random variables, we also obtain

var(Y) 1 = E[N] var(X)

+ ( )2 E[X] var(N) =p. +

1 I-p

A2 A2 . �= A2p2 ·

In order to find the transform My (s), let us recall that

Mx (s)

Then, My (s) is found by starting with MN (S) and replacing each occurrence of eS with Mx (s). This yields

PA

pMx (s)

My (s)

= A-S

1 - (1 - p)Mx (s) -

1 - ( 1 - p)

­ A-S

which simplifies to

My (s) =

p -s

We recognize this as the transform associated with an exponentially distributed random variable with parameter PA, and therefore,

fy

(y) -P>'Y , pAe

y � o.

244 Further Topics on Random Variables Ch ap. 4 This result can be surprising because the sum of a fixed number n of indepen­

dent exponential random variables is not exponentially distributed. For example, if n=

2, the transform associated with the sum is (>'/(>' - s ) ) 2, which does not

correspond to an exponential distribution. Example 4.36. Sum of a Geometric Number of Independent Geometric

Random Variables. This example is a discrete counterpart of the preceding one. We let N be geometrically distributed with parameter p. We also let each random variable Xi be geometrically distributed with parameter q. We assume that all of these random variables are independent. Let Y = Xl +... + XN . We have

Mx( s ) =

To determine My (s), we start with the formula for MN (S) and replace each occur­ rence of eS with M x (s) . This yields

pMx (s) lY!y (s) =

1 _ (1 - p)Mx (s) '

and, after some algebra,

My ( s) =

1- -

We conclude that Y is geometrically distributed, with parameter pq.