Worked example showing the convolution of two sequences, with terms p_n, q_n, and the resulting convolution terms c_n

How Do We Find the Convolution of Sequences?

Convolutions are important in probability, but can be a little confusing.

We consider two sequences. The first is the sequence of odd numbers:

$$1, 3, 5, 7, \dots$$

Let the $n$th term of this sequence be $p_n$, where we start from $n=0$, and $p_n=2n+1$.

The second is the sequence of positive powers of $2$:

$$2,4,8,16, \dots $$

Let the $n$th term of this sequence be $q_n$, where we start from $n=0$, and $q_n=2^{n+1}$.

Now, the convolution of these two sequences is a new sequence, with $n$th term equal to $c_n$, produced from these two initial sequences – but not by just multiplying $p_n$ and $q_n$.

Rather, we use the following formula:

$$c_n = p_0 q_n + p_1 q_{n-1} + p_2 q_{n-2} + \dots + p_n q_0$$

So, $c_n$ is the sum of $n+1$ terms, each of which is a product of two terms, one from each original sequence, whose indices sum to $n$.

For our example, then, we calculate the first four terms of this convolution sequence as follows:

$n$ $p_n$ $q_n$ $c_n$
$0$ $1$ $2$ $1 \times 2 = 2$
$1$ $3$ $4$ $1 \times 4 + 2 \times 3 = 10$
$2$ $5$ $8$ $1 \times 8 + 3 \times 4 + 5 \times 2 = 30$
$3$ $7$ $16$ $1 \times 16 + 3 \times 8 + 5 \times 4 + 7 \times 2 = 74$

Note that to find $c_n$ we need to reference all the rows up to the current one (but no later rows).

Background: