GOA COLLEGE OF ENGINEERING

“BHAUSAHEB BANDODKAR TECHNICAL EDUCATION COMPLEX"

FARMAGUDI, PONDA- GOA - INDIA

DISCRETE MATHEMATICS
S.E.(COMPUTER) SEM-IV
RECURRENCE RELATIONS

DEVELOPED BY MATHEMATICS FACULTY

DEPARTMENT OF SCIENCE & HUMANITIES
GOA COLLEGE OF ENGINEERING
 RECURRENCE RELATIONS
TOPICS TO BE COVERED:
➢ INTRODUCTION

➢ FORMULATION OF RECURRENCE RELATION

➢ METHODS OF SOLUTION OF RECURRENCE RELATIONS

➢ LINEAR RECURRENCE RELATIONS WITH CONSTANT COEFFICIENTS

➢ SOLUTION BY METHOD OF CHARACTERISTIC ROOTS

➢ GENERATING FUNCTIONS

➢ SOLUTION BY GENERATING FUNCTIONS

 INTRODUCTION
▪ Sequences are ordered lists of elements.

• 1, 2, 3, 5, 8,…
• 1, 3, 9, 27, 81, …
▪ Sequences arise throughout mathematics, computer science, and in many
other disciplines, ranging from botany to music.

Definition: A sequence is a function from a subset of the integers (usually either

the set {0, 1, 2, 3, 4, …} or {1, 2, 3,4, …}) to a set 𝑆

The notation 𝑎𝑛 is used to denote the image of the integer n. We can think 𝑎𝑛
as equivalent of 𝑓(𝑛) where f is a function from {0, 1, 2, 3, 4, …} to S.
 INTRODUCTION
Definition: A recurrence relation for the sequence 𝑎𝑛 is an equation that expresses 𝑎𝑛
in terms of one or more of the previous terms of the sequence, namely, 𝑎0 , 𝑎1 , 𝑎2 ,…
𝑎𝑛−1 , for all integers n with 𝑛 ≥ 𝑛0 where 𝑛0 is a nonnegative integer.
• A sequence is called a solution of a recurrence relation if its terms satisfy the
recurrence relation.

• Finding a formula for the nth term of the sequence generated by a recurrence
relation is called solving the recurrence relation.

• Such a formula is called a explicit (closed) formula.

• The initial conditions for a sequence specify the terms that precede the first term
where the recurrence relation takes effect.
 INTRODUCTION
• Example: Let 𝑎𝑛 be a sequence that satisfies the recurrence relation
𝑎𝑛 = 𝑎𝑛−1 + 3 for 𝑛 = 1,2,3,4, … and suppose that 𝑎0 = 2. What are
𝑎1 , 𝑎2 and 𝑎3 ?
Solution: 𝑎1 = 𝑎0 + 3 = 2 + 3 = 5
𝑎2 = 𝑎1 + 3 = 5 + 3 = 8
𝑎3 = 𝑎2 + 3 = 8 + 3 =11
• Example: Let 𝑎𝑛 be a sequence that satisfies the recurrence relation
𝑎𝑛 = 𝑎𝑛−1 + 𝑎𝑛−2 ; 𝑛 ≥ 2 and 𝑎0 = 0 and 𝑎1 = 1 (Fibonacci sequence)
Solution: 𝑎2 = 𝑎1 + 𝑎0 = 1 + 0 = 1
𝑎3 = 𝑎2 + 𝑎1 = 1 + 1 = 2
𝑎4 = 𝑎3 + 𝑎2 = 2 + 1 = 3
 INTRODUCTION
• Example: Consider the recurrence relation
𝑎𝑛 = 𝑛𝑎𝑛−1 ; 𝑛 ≥ 1 𝑎𝑛𝑑 𝑎0 = 2.
The solution is 𝑎𝑛 = 𝑛!

• Example:
𝑎𝑛 = 2𝑎𝑛−1 + 1; 𝑛 ≥ 2 and 𝑎1 = 1
The solution is 𝑎𝑛 = 2𝑛 − 1
 FORMULATION
Example 1: A person invest Rs. 10,000 in a savings account @ 12% interest compounded annually. Find
the recurrence relation for the amount 𝑎𝑛 at the end of n years. How much will be there at the end of 15
years?
Solution: Let 𝑎𝑛 be the amount at the end of n years.
The amount at the end of n years is equal to the amount at the end of n-1 years plus the interest on that
amount.
𝑎𝑛 = 𝑎𝑛−1 + 0 ⋅ 12 𝑎𝑛−1 ; 𝑛 ≥ 1, 𝑎0 = 10000
𝑎𝑛 = 1.12 𝑎𝑛−1 ; 𝑛 ≥ 1, 𝑎0 = 10000
𝑎𝑛 = 1.12 𝑎𝑛−1 = 1.12 1.12 𝑎𝑛−2 = 1.12 2 𝑎𝑛−2
= 1.12 3 𝑎𝑛−3
= 1.12 4 𝑎𝑛−4
⋮
= 1.12 𝑛 𝑎0
= 1.12 𝑛 ∙ 10000
The amount at the end of 15 years = 𝑎𝑛 = 1.12 15 ∙ 10000 = 𝑅𝑠. 54735.66
 FORMULATION
Example 2: There are two kinds of particles 𝛼 and 𝛽 in a nuclear reactor. In every second, an 𝛼 particle
will split up into 3 𝛽 particles and a 𝛽 particle will split up into an 𝛼 particle and 2 𝛽 particles. If there is a
single 𝛼 particle in the reactor at time 𝑡 = 0, find the recurrence relation for the number of particles
altogether at the end of n seconds. How many particles are there altogether at 𝑡 = 100 seconds ?
Solution: Let 𝑎𝑛 be total number of particles at the end of n seconds.
Since each of 𝛼 and 𝛽 particles splits into 3 particles at the end of every second,
𝑎𝑛 = 3𝑎𝑛−1 , 𝑛 ≥ 1; 𝑎0 = 1
𝑎𝑛 = 3𝑎𝑛−1
= 32 𝑎𝑛−2
= 33 𝑎𝑛−3
⋮
= 3𝑛 𝑎0 = 3𝑛
𝑎100 = 3100 particles
 FORMULATION
Example 3: Find the recurrence relation 𝑎𝑛 for the number of ways of arranging n
distinct objects in row.

Solution: let 𝑎𝑛 be the number of ways of arranging n distinct objects in row.

The first object can be arranged in n ways and the remaining n-1 objects can be

arranged in 𝑎𝑛−1 ways

Therefore 𝑎𝑛 = 𝑛 ∙ 𝑎𝑛−1 ; , 𝑛 ≥ 1; 𝑎1 = 1 is the required recurrence relation.

 FORMULATION
Example 4: Find the recurrence relation 𝑎𝑛 for the number of subsets of a finite set A containing n
elements.

Solution: let 𝑎𝑛 be the number of subsets of a finite set 𝐴 = 𝑥1 , 𝑥2 , 𝑥3 , ⋯ , 𝑥𝑛 containing n elements.

𝐴 = 𝑥1 , 𝑥2 , 𝑥3 , ⋯ , 𝑥𝑛−1 ∪ 𝑥𝑛

The set 𝑥1 , 𝑥2 , 𝑥3 , ⋯ , 𝑥𝑛−1 has 𝑎𝑛−1 subsets

The element 𝑥𝑛 may or may not belong to the 𝑎𝑛−1 subsets.

Therefore, total number of subsets of A is 𝑎𝑛 = 𝑎𝑛−1 + 𝑎𝑛−1

𝑎𝑛 = 2𝑎𝑛−1 ; 𝑛 ≥ 2; 𝑎1 = 2 is the required recurrence relation.

 FORMULATION
Example 5:Tower of Hanoi
• In the nineteenth century, a game called the Tower of Hanoi became popular in
Europe. This game represents work that is under way in the temple of Brahma.
• There are three pegs, with one peg containing 64 golden disks. Each golden disk is
slightly smaller than the disk below it.
• The task is to move all 64 disks from the first peg to the third peg.

.
 FORMULATION
• The rules for moving the disks are as follows:
1. Only one disk can be moved at a time.
2. The removed disk must be placed on one of the pegs.
3. A larger disk cannot be placed on top of a smaller disk.
• The objective is to determine the minimum number of moves required to transfer the disks
from the first peg to the third peg.
• Find the recurrence relation 𝑎𝑛 for the minimum number of moves required to transfer n
disks from the first peg to the third peg.
Solution: Let 𝑎𝑛 for the minimum number of moves required to transfer n disks from one peg to
another peg.
n – 1 disks can be transferred from peg 1 to peg 2 in 𝑎𝑛−1 moves.
The largest disk can be transferred from peg 1 to peg 3 in one move.
The n – 1 disks from peg 2 can be transferred to peg 3 in 𝑎𝑛−1 moves.
Therefore, 𝑎𝑛 = 𝑎𝑛−1 + 1 + 𝑎𝑛−1
𝑎𝑛 = 2𝑎𝑛−1 + 1; 𝑛 ≥ 2, 𝑎1 = 1 is the required recurrence relation.
 FORMULATION
Example 6: Find the recurrence relation 𝑎𝑛 and give initial conditions for the number of binary sequences(
bit strings) of length n that do not contain the pattern 11.

Solution: Let 𝑎𝑛 be the number of binary sequences( bit strings) of length n that do not contain the
pattern 11.

𝑎𝑛 = 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑏𝑖𝑡 𝑠𝑡𝑟𝑖𝑛𝑔𝑠 𝑠𝑡𝑎𝑟𝑡𝑖𝑛𝑔 𝑤𝑖𝑡ℎ 0 + 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑏𝑖𝑡 𝑠𝑡𝑟𝑖𝑛𝑔𝑠 𝑠𝑡𝑎𝑟𝑡𝑖𝑛𝑔 𝑤𝑖𝑡ℎ 1

If the binary string starts with 0, then we need to append a binary string of length n-1 that do not contain
the pattern 11. There are 𝑎𝑛−1 such bit strings.

If the binary string starts with 1, then the second bit has to be 0. We need to append a binary string of
length n-2 that do not contain the pattern 11. There are 𝑎𝑛−2 such bit strings.

∴ 𝑎𝑛 = 𝑎𝑛−1 + 𝑎𝑛−2 ; 𝑛 ≥ 3, 𝑎1 = 2, 𝑎2 = 3
 FORMULATION
Example 7: Find the recurrence relation 𝑎𝑛 and give initial conditions for the number of binary sequences( bit
strings) of length n in which the pattern 11 appears for the first time at the end of the sequence.
Solution: Let 𝑎𝑛 be the number of binary sequences( bit strings) of length n in which the pattern 11 appears for
the first time at the end of the sequence.
𝑎𝑛 = 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑏𝑖𝑡 𝑠𝑡𝑟𝑖𝑛𝑔𝑠 𝑠𝑡𝑎𝑟𝑡𝑖𝑛𝑔 𝑤𝑖𝑡ℎ 0 + 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑏𝑖𝑡 𝑠𝑡𝑟𝑖𝑛𝑔𝑠 𝑠𝑡𝑎𝑟𝑡𝑖𝑛𝑔 𝑤𝑖𝑡ℎ 1
If the binary string starts with 0, then we need to append a binary string of length n-1 in which the pattern 11
appears for the first time at the end of the sequence. There are 𝑎𝑛−1 such bit strings.
If the binary string starts with 1, then the second bit has to be 0. We need to append a binary string of length
n-2 in which the pattern 11 appears for the first time at the end of the sequence. There are 𝑎𝑛−2 such bit
strings.
∴ 𝑎𝑛 = 𝑎𝑛−1 + 𝑎𝑛−2 ; 𝑛 ≥ 3, 𝑎1 = 0, 𝑎2 = 1
 FORMULATION
Example 8: Find the recurrence relation 𝑎𝑛 and give initial conditions for the number of binary sequences( bit strings) of
length n that do not contain three consecutive zeros.
Solution: Let 𝑎𝑛 be the number of binary sequences( bit strings) of length n that do not contain three consecutive zeros.
𝑎𝑛 = 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑏𝑖𝑡 𝑠𝑡𝑟𝑖𝑛𝑔𝑠 𝑠𝑡𝑎𝑟𝑡𝑖𝑛𝑔 𝑤𝑖𝑡ℎ 0 + 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑏𝑖𝑡 𝑠𝑡𝑟𝑖𝑛𝑔𝑠 𝑠𝑡𝑎𝑟𝑡𝑖𝑛𝑔 𝑤𝑖𝑡ℎ 1
• If the binary string starts with 1, then we need to append a binary string of length n-1 that do not contain three
consecutive zeros. There are 𝑎𝑛−1 such bit strings.
• Consider the binary strings starting with 0.
If the second bit is 1, we need to append a binary string of length n-2 that do not contain three consecutive zeros.
There are 𝑎𝑛−2 such bit strings.
If the second bit is 0, then the third bit has to be 1. Hence we need to append a binary string of length n-3 that do not
contain three consecutive zeros. There are 𝑎𝑛−3 such bit strings.
∴ 𝑎𝑛 = 𝑎𝑛−1 + 𝑎𝑛−2 + 𝑎𝑛−3 ; 𝑛 ≥ 4, 𝑎1 = 2, 𝑎2 = 4, 𝑎3 = 7
 FORMULATION
Example 9: Find the recurrence relation 𝑎𝑛 and give initial conditions for the number of ways of climbing n
stairs if a person climbing the stairs can climb one or two or three stairs at a time.

Solution: Let 𝑎𝑛 be the number the number of ways of climbing n stairs .

• If the person takes a single step in the beginning, then the remaining n-1 steps can be climbed in 𝑎𝑛−1
ways.

• If the person jumps over two steps in the beginning, then the remaining n-2 steps can be climbed in
𝑎𝑛−2 ways.

• If the person jumps over three steps in the beginning, then the remaining n-3 steps can be climbed in
𝑎𝑛−3 ways.

∴ 𝑎𝑛 = 𝑎𝑛−1 + 𝑎𝑛−2 + 𝑎𝑛−3 ; 𝑛 ≥ 4, 𝑎1 = 1, 𝑎2 = 2, 𝑎3 = 4

 LINEAR RECURRENCE RELATIONS WITH CONSTANT COEFFICIENTS

Definition: A linear homogeneous recurrence relation of degree ‘k’ with constant coefficients is
a recurrence relation of the form
𝑐0 𝑎𝑛 + 𝑐1 𝑎𝑛−1 + 𝑐2 𝑎𝑛−2 + ⋯ + 𝑐𝑘 𝑎𝑛−𝑘 = 𝑓(𝑛)
where 𝑐0 , 𝑐1 , 𝑐2 , …, 𝑐𝑘 are real constants and 𝑐𝑘 ≠ 0
• If 𝑓 𝑛 = 0, the recurrence relation is said to be homogenous otherwise non-homogenous
Examples:
1 𝑎𝑛 = (1.11) 𝑎𝑛−1 linear homogeneous recurrence relation of degree 1
2 𝑓𝑛 = 𝑓𝑛−1 + 𝑓𝑛−2 linear homogeneous recurrence relation of degree 2
3 𝑎𝑛 = 2𝑎𝑛−1 + 1 linear non-homogeneous recurrence relation of degree 1
2
4 𝑎𝑛 = 𝑎𝑛−1 + 𝑎𝑛−2 non linear homogeneous recurrence relation of degree 2
2
5 𝑎𝑛 = 𝑛𝑎𝑛−1 non linear , no constant coefficients,homogenous
 METHODS OF SOLUTION OF RECURRENCE RELATIONS
1. Iteration Method: In this method, we apply the recurrence relation repeatedly to obtain the general
expression for 𝑎𝑛 in terms of n.
Examples:
1. Solve the recurrence relation 𝑎𝑛 = 𝑎𝑛−1 + 2, n ≥ 2 , 𝑎1 = 3
Solution: 𝑎𝑛 = 𝑎𝑛−1 + 2
= 𝑎𝑛−2 + 2 + 2 = 𝑎𝑛−2 + 2 ∙ 2
= 𝑎𝑛−3 + 2 + 2 ∙ 2 = 𝑎𝑛−3 + 3 ∙ 2
⋮
= 𝑎1 + (𝑛 − 1) ∙ 2
= 3 + 2(𝑛 − 1)
= 2𝑛 + 1
 ITERATION METHOD
2. Solve the recurrence relation 𝑎𝑛 = 𝑎𝑛−1 + n, n ≥ 2 , 𝑎1 = 1
Solution: 𝑎𝑛 = 𝑎𝑛−1 + n
= 𝑎𝑛−2 + 𝑛 − 1 + n
𝑛(𝑛 + 1)
= 𝑎𝑛−3 + 𝑛 − 2 + 𝑛 − 1 + n 1 + 2 + 3 + ⋯+ 𝑛 =
2
= 𝑎𝑛−4 + 𝑛 − 3 + 𝑛 − 2 + 𝑛 − 1 + n
⋮
= 𝑎1 + 2 + 3 + ⋯ + 𝑛 − 1 + n
= 1 + 2 + 3 + ⋯+ 𝑛
𝑛(𝑛+1)
=
2
 ITERATION METHOD
3. Solve the recurrence relation 𝑎𝑛 = 2𝑎𝑛−1 + 1, n ≥ 2 , 𝑎1 = 1
Solution: 𝑎𝑛 = 2𝑎𝑛−1 + 1
= 2 2𝑎𝑛−2 + 1 + 1 = 22 𝑎𝑛−2 + 2 + 1
= 22 2𝑎𝑛−3 + 1 + 2 + 1 = 23 𝑎𝑛−3 + 22 + 2 + 1
= 24 𝑎𝑛−4 + 23 + 22 + 2 + 1
⋮
= 2𝑛−1 𝑎1 + 2𝑛−2 + ⋯ + 22 + 2 + 1
= 2𝑛−1 + 2𝑛−2 + ⋯ + 22 + 2 + 1
2𝑛 −1
=
2−1
1 − 𝑥 𝑛+1
= 2𝑛 − 1 2 3
1+ 𝑥 + 𝑥 +𝑥 +⋯+ 𝑥 = 𝑛
1−𝑥
𝑖𝑓 |𝑥| < 1

𝑥 𝑛+1 − 1
= 𝑖𝑓 |𝑥| > 1
𝑥−1
 METHOD OF CHARACTERISTIC ROOTS
• Used for solving linear recurrence relations with constant coefficients
• In this method the solution is obtained as the sum of two parts 𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝) , the
homogeneous solution 𝑎𝑛 (ℎ) which satisfy the recurrence relation when the RHS = 0 and the
particular solution 𝑎𝑛 (𝑝) which satisfy the recurrence relation with 𝑓 𝑛 on RHS.
• Solving homogeneous recurrence relations
Consider 𝑐0 𝑎𝑛 + 𝑐1 𝑎𝑛−1 + 𝑐2 𝑎𝑛−2 + ⋯ + 𝑐𝑘 𝑎𝑛−𝑘 = 0 -----(1)
Let 𝑎𝑛 = 𝑟 𝑛 be a solution of (1)
∴ 𝑐0 𝑟 𝑛 + 𝑐1 𝑟 𝑛−1 + 𝑐2 𝑟 𝑛−2 + ⋯ + 𝑐𝑘 𝑟 𝑛−𝑘 = 0
Dividing by 𝑟 𝑛−𝑘
𝑐0 𝑟 𝑘 + 𝑐1 𝑟 𝑘−1 + 𝑐2 𝑟 𝑘−2 + ⋯ + 𝑐𝑘 = 0 -----(2)
(2) is called the characteristic equation which is a polynomial equation of degree ‘k’ in r.
A characteristic equation of degree ‘k’ has k characteristic roots 𝑟1 , 𝑟2 , 𝑟3 , ⋯ 𝑟𝑘 .
 METHOD OF CHARACTERISTIC ROOTS
Case I (Distinct roots) : If the k characteristic roots are distinct, then the general form of the solution is
𝑎𝑛 (ℎ) = 𝑏1 𝑟1 𝑛 + 𝑏2 𝑟2 𝑛 + 𝑏3 𝑟3 𝑛 + ⋯ + 𝑏𝑘 𝑟𝑘 𝑛
where 𝑏1 , 𝑏2 , 𝑏3 , ⋯ 𝑏𝑘 are constants which satisfy the initial conditions.

Case ii (Multiple roots): If 𝑟1 is the root of multiplicity m and the remaining 𝑘 − 𝑚 roots of distinct, then
𝑎𝑛 (ℎ) = (𝑏1 +𝑛𝑏2 + 𝑛2 𝑏3 + ⋯ + 𝑛𝑚−1 𝑏𝑚 )𝑟1 𝑛 + 𝑏𝑚+1 𝑟2 𝑛 + 𝑏𝑚+2 𝑟3 𝑛 + ⋯

Examples:
1. If the roots of characteristic equation are 2, 3, -4 , then
𝑎𝑛 (ℎ) = 𝑏1 2𝑛 + 𝑏2 3𝑛 + 𝑏3 (−4)𝑛
2. If the roots of characteristic equation are 2, 2, 2, 3 ,4 then
𝑎𝑛 (ℎ) = (𝑏1 +𝑛𝑏2 + 𝑛2 𝑏3 )2𝑛 + 𝑏4 3𝑛 + 𝑏5 (4)𝑛
3. If the roots of characteristic equation are -1, -1, 3, 3 ,4 then
𝑎𝑛 (ℎ) = (𝑏1 +𝑛𝑏2 )(−1)𝑛 +(𝑏3 +𝑛𝑏4 )3𝑛 + 𝑏5 (4)𝑛
 METHOD OF CHARACTERISTIC ROOTS
1. Solve the recurrence relation 𝑎𝑛 = 5𝑎𝑛−1 , n ≥ 1 , 𝑎0 = 1
Solution: 𝑎𝑛 − 5𝑎𝑛−1 = 0 -----(1)
Let 𝑎𝑛 = 𝑟 𝑛 be a solution of (1)
𝑟 𝑛 − 5𝑟 𝑛−1 = 0
Divide by 𝑟 𝑛−1
⟹r−5=0
⟹r= 5
∴ 𝑎𝑛 = 𝑏1 5𝑛
put n = 0, 𝑎0 = 𝑏1
⟹ 𝑏1 = 1
∴ 𝑎𝑛 = 5𝑛 is the required solution.
 METHOD OF CHARACTERISTIC ROOTS
2. Solve the recurrence relation 𝑎𝑛 = 4(𝑎𝑛−1 − 𝑎𝑛−2 ), n ≥ 2 , 𝑎0 = 0, 𝑎1 = 2
Solution: 𝑎𝑛 − 4𝑎𝑛−1 + 4𝑎𝑛−2 = 0 -----(1)
Let 𝑎𝑛 = 𝑟 𝑛 be a solution of (1)
𝑟 𝑛 − 4𝑟 𝑛−1 + 4𝑟 𝑛−2 = 0
Divide by 𝑟 𝑛−2
⟹𝑟 2 − 4r + 4 = 0
⟹ r = 2, 2
∴ 𝑎𝑛 = (𝑏1 + 𝑛𝑏2 )2𝑛
put n = 0, 𝑎0 = 𝑏1 ⟹ 𝑏1 = 0
put n = 1 𝑎1 = 2𝑏1 + 2𝑏2
⟹ 2 = 2𝑏2
⟹ 𝑏2 = 1
∴ 𝑎𝑛 = 𝑛 ∙ 2𝑛 is the required solution.
 METHOD OF CHARACTERISTIC ROOTS
3. Solve the recurrence relation of the Fibonacci sequence
𝑎𝑛 = 𝑎𝑛−1 + 𝑎𝑛−2 , n ≥ 2 , 𝑎0 = 0, 𝑎1 = 1
Solution: 𝑎𝑛 − 𝑎𝑛−1 − 𝑎𝑛−2 = 0 -----(1)
Let 𝑎𝑛 = 𝑟 𝑛 be a solution of (1)
𝑟 𝑛 − 𝑟 𝑛−1 − 𝑟 𝑛−2 = 0
Divide by 𝑟 𝑛−2
⟹𝑟 2 − r − 1 = 0
1+ 5 1− 5
⟹r = , 2
2
𝑛 𝑛
1+ 5 1− 5
∴ 𝑎𝑛 = 𝑏1 + 𝑏2
2 2
put n = 0, 0 = 𝑏1 + 𝑏2
1+ 5 1− 5
put n = 1 1 = 2
𝑏 1 + 2
𝑏2
1 1
⟹ 𝑏1 = and 𝑏2 = −
5 5
𝑛 𝑛
1 1+ 5 1 1− 5
∴ 𝑎𝑛 = 5 2 − 5 2 is the required solution.
 METHOD OF CHARACTERISTIC ROOTS
Solving non homogeneous recurrence relations:
• Consider 𝑐0 𝑎𝑛 + 𝑐1 𝑎𝑛−1 + 𝑐2 𝑎𝑛−2 + ⋯ + 𝑐𝑘 𝑎𝑛−𝑘 = 𝑓(𝑛)
• Then 𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝)
• 𝑎𝑛 (ℎ) is obtained by solving 𝑐0 𝑟 𝑛 + 𝑐1 𝑟 𝑛−1 + 𝑐2 𝑟 𝑛−2 + ⋯ + 𝑐𝑘 𝑟 𝑛−𝑘 = 0.
• There is no general method to find the particular solution 𝑎𝑛 (𝑝) .
• Depending upon the characteristic roots and 𝑓(𝑛), the following trial sequences are used

𝑓(𝑛) Trial sequence 𝑎𝑛 (𝑝)

1 𝑓(𝑛) is a constant and r≠1 i.e 1 is not a 𝐴, where A is a constant
root of characteristic equation
2 𝑓(𝑛) is a constant and r =1 is a root of 𝐴𝑛𝑚
characteristic equation of multiplicity m
3 𝑓 𝑛 = 𝑐 𝑛 , where r ≠ c i.e c is not a root 𝐴𝑐 𝑛
of characteristic equation
 METHOD OF CHARACTERISTIC ROOTS

𝑓(𝑛) Trial sequence 𝑎𝑛 (𝑝)

4 𝑓 𝑛 = 𝑐 𝑛 , where r = c is a root of characteristic equation 𝐴𝑐 𝑛 𝑛𝑚
of multiplicity m
5 𝑓 𝑛 = 𝑃(𝑛) where 𝐴0 + 𝐴1 𝑛 + 𝐴2 𝑛2 + ⋯ + 𝐴𝑘 𝑛𝑘
𝑃 𝑛 = 𝑐0 + 𝑐1 𝑛 + 𝑐2 𝑛2 + ⋯ + 𝑐𝑘 𝑛𝑘 is a polynomial of
degree k and r≠1(1 is not a root of characteristic equation)
6 𝑓 𝑛 = 𝑃(𝑛) where 𝐴0 + 𝐴1 𝑛 + 𝐴2 𝑛2 + ⋯ + 𝐴𝑘 𝑛𝑘 𝑛𝑚
𝑃 𝑛 = 𝑐0 + 𝑐1 𝑛 + 𝑐2 𝑛2 + ⋯ + 𝑐𝑘 𝑛𝑘
is a polynomial of degree k and r =1 is a root of
characteristic equation of multiplicity m
7 𝑓 𝑛 = 𝑐 𝑛 ∙ 𝑃(𝑛), where r ≠ c (c is not a root of 𝑐 𝑛 𝐴0 + 𝐴1 𝑛 + 𝐴2 𝑛2 + ⋯ + 𝐴𝑘 𝑛𝑘
characteristic equation)

8 𝑓 𝑛 = 𝑐 𝑛 ∙ 𝑃(𝑛), where r = c is a root of characteristic 𝑐 𝑛 𝐴0 + 𝐴1 𝑛 + 𝐴2 𝑛2 + ⋯ + 𝐴𝑘 𝑛𝑘 𝑛𝑚

equation of multiplicity m
 METHOD OF CHARACTERISTIC ROOTS
1. Solve the recurrence relation 𝑎𝑛 − 3𝑎𝑛−1 − 18𝑎𝑛−2 = 20, n ≥ 2 , 𝑎0 = 1, 𝑎1 =2
Solution: 𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝)
To find 𝑎𝑛 (ℎ) , consider the homogenous recurrence relation
𝑎𝑛 − 3𝑎𝑛−1 − 18𝑎𝑛−2 = 0 ----- (1)
Let 𝑎𝑛 = 𝑟 𝑛 be a solution of (1)
𝑟 𝑛 − 3𝑟 𝑛−1 − 18𝑟 𝑛−2 = 0
Divide by 𝑟 𝑛−2
⟹𝑟 2 − 3r − 18 = 0
⟹ r = 6, −3
∴ 𝑎𝑛 (ℎ) = 𝑏1 6𝑛 + 𝑏2 (−3)𝑛
Since 𝑓 𝑛 = 20 # 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 and 1 is not a root of characteristic equation,
𝑎𝑛 (𝑝) = 𝐴
Substitute 𝑎𝑛 (𝑝) = 𝐴 in the non homogenous recurrence relation
 METHOD OF CHARACTERISTIC ROOTS
𝑎𝑛 − 3𝑎𝑛−1 − 18𝑎𝑛−2 = 20
⟹ 𝐴 − 3𝐴 − 18𝐴 = 20
⟹ −20𝐴 = 20
⟹ 𝐴 = −1
∴ 𝑎𝑛 (𝑝) = −1
𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝) = 𝑏1 6𝑛 + 𝑏2 (−3)𝑛 −1 is the general solution
put n = 0, 1 = 𝑏1 + 𝑏2 − 1
⟹ 𝑏1 + 𝑏2 = 2 ----- (i)
put n = 1 2 = 6𝑏1 − 3𝑏2 − 1
⟹ 6𝑏1 − 3𝑏2 = 3
⟹ 2𝑏1 − 𝑏2 = 1 ----- (ii)
Solving (i) and (ii), 𝑏1 = 1 and 𝑏2 = 1
∴ 𝑎𝑛 = 6𝑛 + (−3)𝑛 −1 is the required solution.
 METHOD OF CHARACTERISTIC ROOTS
2. Solve the recurrence relation 𝑎𝑛 − 2𝑎𝑛−1 + 𝑎𝑛−2 = 6, n ≥ 2 , 𝑎0 = 0, 𝑎1 =2
Solution: 𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝)
To find 𝑎𝑛 (ℎ) , consider the homogenous recurrence relation
𝑎𝑛 − 2𝑎𝑛−1 + 𝑎𝑛−2 = 0 ----- (1)
Let 𝑎𝑛 = 𝑟 𝑛 be a solution of (1)
𝑟 𝑛 − 2𝑟 𝑛−1 + 𝑟 𝑛−2 = 0
Divide by 𝑟 𝑛−2
⟹𝑟 2 − 2r + 1 = 0
⟹ r = 1, 1
∴ 𝑎𝑛 (ℎ) = 𝑏1 + 𝑛𝑏2 1𝑛 = 𝑏1 + 𝑛𝑏2
Since 𝑓 𝑛 = 6 # 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 and 1 is a root of characteristic equation of multiplicity 2,
𝑎𝑛 (𝑝) = 𝐴𝑛2
Substitute 𝑎𝑛 (𝑝) = 𝐴𝑛2 in the non homogenous recurrence relation
 METHOD OF CHARACTERISTIC ROOTS
𝑎𝑛 − 2𝑎𝑛−1 + 𝑎𝑛−2 = 6
⟹ 𝐴𝑛2 − 2𝐴 𝑛 − 1 2 + 𝐴 𝑛 − 2 2 = 6
⟹ 𝐴𝑛2 − 2𝐴𝑛2 + 4𝐴𝑛 − 2𝐴 + 𝐴𝑛2 − 4𝐴𝑛 + 4𝐴 = 6
⟹2𝐴 = 6
⟹𝐴 = 3
𝑎𝑛 (𝑝) = 3𝑛2
𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝) = 𝑏1 + 𝑛𝑏2 + 3𝑛2 is the general solution
put n = 0, 0 = 𝑏1
put n = 1 2 = 𝑏1 + 𝑏2 + 3
⟹ 𝑏1 + 𝑏2 = −1
Substituting 𝑏1 = 0 and 𝑏2 = −1
∴ 𝑎𝑛 = −𝑛 + 3𝑛2 is the required solution.
 METHOD OF CHARACTERISTIC ROOTS
3. Solve the recurrence relation 𝑎𝑛 = 5𝑎𝑛−1 − 6𝑎𝑛−2 + 7𝑛 , n ≥ 2 , 𝑎0 = 0, 𝑎1 = 0
Solution: 𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝)
To find 𝑎𝑛 (ℎ) , consider the homogenous recurrence relation
𝑎𝑛 − 5𝑎𝑛−1 + 6𝑎𝑛−2 = 0 ----- (1)
Let 𝑎𝑛 = 𝑟 𝑛 be a solution of (1)
𝑟 𝑛 − 5𝑟 𝑛−1 + 6𝑟 𝑛−2 = 0
Divide by 𝑟 𝑛−2
⟹𝑟 2 − 5r + 6 = 0
⟹ r = 2, 3
∴ 𝑎𝑛 (ℎ) = 𝑏1 2𝑛 + 𝑏2 3𝑛
Since 𝑓 𝑛 = 7𝑛 and 7 is not a root of characteristic equation
𝑎𝑛 (𝑝) = 𝐴7𝑛
Substitute 𝑎𝑛 (𝑝) = 𝐴7𝑛 in the non homogenous recurrence relation
 METHOD OF CHARACTERISTIC ROOTS
𝑎𝑛 − 5𝑎𝑛−1 + 6𝑎𝑛−2 = 7𝑛
⟹ 𝐴7𝑛 − 5𝐴7𝑛−1 + 6𝐴7𝑛−2 = 7𝑛
Divide by 7𝑛
5 6
⟹𝐴− 𝐴+ 𝐴 =1
7 49
20
⟹ 𝐴 =1
49
49
⟹𝐴 =
20
49 𝑛
𝑎𝑛 (𝑝) = 7
20
49 𝑛
𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝) = 𝑏1 2𝑛 + 𝑏2 3𝑛 + 7
20
49
put n = 0, 0 = 𝑏1 + 𝑏2 +
20
49
⟹ 𝑏1 + 𝑏2 = − ----- (i)
20
343
put n = 1 0 = 2𝑏1 + 3𝑏2 +
20
343
⟹ 2𝑏1 + 3𝑏2 = − -----(ii)
20
49 49
Solving (i) and (ii), 𝑏1 = and 𝑏2 = −
5 4
49 𝑛 49 𝑛 49 𝑛
∴ 𝑎𝑛 = 2 − 3 + 7 is the required solution.
5 4 20
 METHOD OF CHARACTERISTIC ROOTS
4. Solve the recurrence relation 𝑎𝑛 − 5𝑎𝑛−1 + 6𝑎𝑛−2 = 2𝑛 , n ≥ 2 , 𝑎0 = 1, 𝑎1 = 0
Solution: 𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝)
To find 𝑎𝑛 (ℎ) , consider the homogenous recurrence relation
𝑎𝑛 − 5𝑎𝑛−1 + 6𝑎𝑛−2 = 0 ----- (1)
Let 𝑎𝑛 = 𝑟 𝑛 be a solution of (1)
𝑟 𝑛 − 5𝑟 𝑛−1 + 6𝑟 𝑛−2 = 0
Divide by 𝑟 𝑛−2
⟹𝑟 2 − 5r + 6 = 0
⟹ r = 2, 3
∴ 𝑎𝑛 (ℎ) = 𝑏1 2𝑛 + 𝑏2 3𝑛
Since 𝑓 𝑛 = 2𝑛 and 2 is a root of characteristic equation of multiplicity 1
𝑎𝑛 (𝑝) = 𝐴𝑛2𝑛
Substitute 𝑎𝑛 (𝑝) = 𝐴𝑛2𝑛 in the non homogenous recurrence relation
 METHOD OF CHARACTERISTIC ROOTS
𝑎𝑛 − 5𝑎𝑛−1 + 6𝑎𝑛−2 = 2𝑛
⟹ 𝐴𝑛2𝑛 − 5𝐴(𝑛 − 1)2𝑛−1 + 6𝐴(𝑛 − 2)2𝑛−2 = 2𝑛
Divide by 2𝑛
5 6
⟹ 𝐴𝑛 − (𝑛 − 1)𝐴 + (𝑛 − 2)𝐴 = 1
2 4
5 5 3
⟹𝐴𝑛 − 𝑛𝐴 + 𝐴+ 𝑛𝐴 − 3𝐴 = 1
2 2 2
1
⟹− 𝐴 = 1 ⟹𝐴 = −2
2
𝑎𝑛 = −2𝑛2𝑛 = −𝑛2𝑛+1
(𝑝)

𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝) = 𝑏1 2𝑛 + 𝑏2 3𝑛 − 𝑛2𝑛+1

put n = 0, 1 = 𝑏1 + 𝑏2 ----- (i)
put n = 1 0 = 2𝑏1 + 3𝑏2 − 4
⟹ 2𝑏1 + 3𝑏2 = 4 ------ (ii)
Solving (i) and (ii), 𝑏1 = −1 and 𝑏2 = 2
∴ 𝑎𝑛 = −2𝑛 + 2 ∙ 3𝑛 − 𝑛2𝑛+1 is the required solution.
 METHOD OF CHARACTERISTIC ROOTS
5. Solve the recurrence relation 𝑎𝑛 − 𝑎𝑛−1 − 2𝑎𝑛−2 = 𝑛2 , n ≥ 2
Solution: 𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝)
To find 𝑎𝑛 (ℎ) , consider the homogenous recurrence relation
𝑎𝑛 − 𝑎𝑛−1 − 2𝑎𝑛−2 = 0 ----- (1)
Let 𝑎𝑛 = 𝑟 𝑛 be a solution of (1)
𝑟 𝑛 − 𝑟 𝑛−1 − 2𝑟 𝑛−2 = 0
Divide by 𝑟 𝑛−2
⟹𝑟 2 − r − 2 = 0
⟹ r = −1, 2
∴ 𝑎𝑛 (ℎ) = 𝑏1 2𝑛 + 𝑏2 (−1)𝑛
Since 𝑓 𝑛 = 𝑛2 is a polynomial of degree 2 and 1 is not a root of characteristic equation
𝑎𝑛 (𝑝) = 𝐴0 + 𝐴1 𝑛 + 𝐴2 𝑛2
Substitute 𝑎𝑛 (𝑝) = 𝐴0 + 𝐴1 𝑛 + 𝐴2 𝑛2 in the non homogenous recurrence relation
 METHOD OF CHARACTERISTIC ROOTS
𝑎𝑛 − 𝑎𝑛−1 − 2𝑎𝑛−2 = 𝑛2
𝐴0 + 𝐴1 𝑛 + 𝐴2 𝑛2 − 𝐴0 + 𝐴1 (𝑛 − 1) + 𝐴2 (𝑛 − 1)2 − 2 𝐴0 + 𝐴1 (𝑛 − 2) + 𝐴2 (𝑛 − 2)2 = 𝑛2
𝐴0 + 𝐴1 𝑛 + 𝐴2 𝑛2 − 𝐴0 − 𝐴1 𝑛 + 𝐴1 − 𝐴2 𝑛2 + 2𝐴2 𝑛 − 𝐴2 − 2𝐴0 − 2𝐴1 𝑛 + 4𝐴1 − 2𝐴2 𝑛2 + 8𝐴2 𝑛 − 8𝐴2 = 𝑛2
−2𝐴0 + 5𝐴1 − 9𝐴2 + −2𝐴1 + 10𝐴2 𝑛 − 2𝐴2 𝑛2 = 𝑛2
Comparing coefficients of like powers of n
1
−2𝐴2 = 1 ⟹ 𝐴2 = − 2
5
−2𝐴1 + 10𝐴2 = 0 ⟹ −2𝐴1 − 5 = 0 ⟹ 𝐴1 = − 2
25 9
−2𝐴0 + 5𝐴1 − 9𝐴2 = 0 ⟹ −2𝐴0 − + = 0 ⟹ 𝐴0 = −4
2 2
5 1
𝑎𝑛 (𝑝) = −4 − 𝑛 − 𝑛2
2 2
5 1
𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝) = 𝑏1 2𝑛 + 𝑏2 (−1)𝑛 −4 − 𝑛 − 𝑛2 is the general solution
2 2
 METHOD OF CHARACTERISTIC ROOTS
6. Solve the recurrence relation 𝑎𝑛 − 4𝑎𝑛−1 + 3𝑎𝑛−2 = 𝑛, n ≥ 2 , 𝑎0 = 0, 𝑎1 = 0
Solution: 𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝)
To find 𝑎𝑛 (ℎ) , consider the homogenous recurrence relation
𝑎𝑛 − 4𝑎𝑛−1 + 3𝑎𝑛−2 = 0 ----- (1)
Let 𝑎𝑛 = 𝑟 𝑛 be a solution of (1)
𝑟 𝑛 − 4𝑟 𝑛−1 + 3𝑟 𝑛−2 = 0
Divide by 𝑟 𝑛−2
⟹𝑟 2 − 4r + 3 = 0
⟹ r = 1, 3
∴ 𝑎𝑛 (ℎ) = 𝑏1 + 𝑏2 3𝑛
Since 𝑓 𝑛 = 𝑛 is a polynomial of degree 1 and 1 is a root of characteristic equation of multiplicity 1
𝑎𝑛 (𝑝) = 𝐴0 + 𝐴1 𝑛 𝑛
Substitute 𝑎𝑛 (𝑝) = 𝐴0 𝑛 + 𝐴1 𝑛2 in the non homogenous recurrence relation
 METHOD OF CHARACTERISTIC ROOTS
𝑎𝑛 − 4𝑎𝑛−1 + 3𝑎𝑛−2 = 𝑛
𝐴0 𝑛 + 𝐴1 𝑛2 − 4 𝐴0 𝑛 − 1 + 𝐴1 𝑛 − 1 2 + 3 𝐴0 𝑛 − 2 + 𝐴1 𝑛 − 2 2 = 𝑛
𝐴0 𝑛 + 𝐴1 𝑛2 − 4𝐴0 𝑛 + 4𝐴0 − 4𝐴1 𝑛2 + 8𝐴1 𝑛 − 4𝐴1 + 3𝐴0 𝑛 − 6𝐴0 + 3𝐴1 𝑛2 − 12𝐴1 𝑛 + 12𝐴1 = 𝑛
−2𝐴0 + 8𝐴1 − 4𝐴1 𝑛 = 𝑛
Comparing coefficients of like powers of n
1
−4𝐴1 = 1 ⟹ 𝐴1 = − 4
−2𝐴0 + 8𝐴1 = 0 ⟹ − 2𝐴0 − 2 = 0 ⟹ 𝐴0 = −1
1
∴𝑎𝑛 (𝑝) = −𝑛 − 4 𝑛2
1
𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝) = 𝑏1 + 𝑏2 3𝑛 − 𝑛 − 𝑛2 is the general solution
4
put n = 0, 0 = 𝑏1 + 𝑏2 ----- (i)
1
put n = 1 0 = 𝑏1 + 3𝑏2 − 1 − 4
5
⟹ 𝑏1 + 3𝑏2 = 4------ (ii)
5 5
Solving (i) and (ii), 𝑏1 = − 8 and 𝑏2 = 8
5 5 1
∴ 𝑎𝑛 = − 8 + 8 3𝑛 − 𝑛 − 4 𝑛2 is the required solution.
 METHOD OF CHARACTERISTIC ROOTS
7. Solve the recurrence relation 𝑎𝑛 − 3𝑎𝑛−1 = 4𝑛 + 3 ∙ 2𝑛 , n ≥ 2
Solution: 𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝)
To find 𝑎𝑛 (ℎ) , consider the homogenous recurrence relation
𝑎𝑛 − 3𝑎𝑛−1 = 0 ----- (1)
Let 𝑎𝑛 = 𝑟 𝑛 be a solution of (1)
𝑟 𝑛 − 3𝑟 𝑛−1 = 0
Divide by 𝑟 𝑛−1
⟹r−3=0
⟹r= 3
∴ 𝑎𝑛 = 𝑏1 3𝑛
Since 𝑓 𝑛 = 4𝑛 + 3 ∙ 2𝑛 and 2 is not a root of characteristic equation
𝑎𝑛 (𝑝) = (𝐴0 + 𝐴1 𝑛)2𝑛
Substitute 𝑎𝑛 (𝑝) = (𝐴0 + 𝐴1 𝑛)2𝑛 in the non homogenous recurrence relation
 METHOD OF CHARACTERISTIC ROOTS
𝑎𝑛 − 3𝑎𝑛−1 = 4𝑛 + 3 ∙ 2𝑛
𝐴0 + 𝐴1 𝑛 2𝑛 − 3 𝐴0 + 𝐴1 𝑛 − 1 2𝑛−1 = 4𝑛 + 3 ∙ 2𝑛
Divide by 2𝑛
3 3 3
𝐴0 + 𝐴1 𝑛 − 𝐴0 − 𝐴1 𝑛 + 𝐴1 = 4𝑛 + 3
2 2 2
1 1 3
− 𝐴1 𝑛 + − 𝐴0 + 𝐴1 = 4𝑛 + 3
2 2 2
Comparing coefficients of like powers of n
1
− 𝐴1 = 4 ⟹ 𝐴1 = −8
2
1 3 1
− 𝐴0 + 𝐴1 = 3 ⟹ − 𝐴0 − 12 = 3 ⟹ 𝐴0 = −30
2 2 2
𝑎𝑛 (𝑝) = (−30 − 8𝑛)2𝑛
𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝) = 𝑏1 3𝑛 − (30 + 8𝑛)2𝑛 is the general solution
 METHOD OF CHARACTERISTIC ROOTS
8. Solve the recurrence relation 𝑎𝑛 − 6𝑎𝑛−1 + 9𝑎𝑛−2 = 𝑛 ∙ 3𝑛
Solution: 𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝)
To find 𝑎𝑛 (ℎ) , consider the homogenous recurrence relation
𝑎𝑛 − 6𝑎𝑛−1 + 9𝑎𝑛−2 = 0 ----- (1)
Let 𝑎𝑛 = 𝑟 𝑛 be a solution of (1)
𝑟 𝑛 − 6𝑟 𝑛−1 + 9𝑟 𝑛−2 = 0
Divide by 𝑟 𝑛−2
⟹𝑟 2 − 6r + 9 = 0
⟹ r = 3, 3
∴ 𝑎𝑛 (ℎ) = 𝑏1 + 𝑛𝑏2 3𝑛
Since 𝑓 𝑛 = 𝑛 ∙ 3𝑛 and 3 is a root of characteristic equation of multiplicity 2
𝑎𝑛 (𝑝) = (𝐴0 + 𝐴1 𝑛)3𝑛 𝑛2
Substitute 𝑎𝑛 (𝑝) = (𝐴0 𝑛2 + 𝐴1 𝑛3 )3𝑛 in the non homogenous recurrence relation
 METHOD OF CHARACTERISTIC ROOTS
𝑎𝑛 − 6𝑎𝑛−1 + 9𝑎𝑛−2 = 𝑛 ∙ 3𝑛
𝐴0 𝑛2 + 𝐴1 𝑛3 3𝑛 − 6 𝐴0 (𝑛 − 1)2 +𝐴1 (𝑛 − 1)3 3𝑛−1 + 9 𝐴0 (𝑛 − 2)2 + 𝐴1 (𝑛 − 2)3 3𝑛−2 = 𝑛 ∙ 3𝑛
Divide by 3𝑛
𝐴0 𝑛2 + 𝐴1 𝑛3 − 2 𝐴0 𝑛 − 1 2 + 𝐴1 𝑛 − 1 3 + 𝐴0 (𝑛 − 2)2 + 𝐴1 𝑛 − 2 3 = 𝑛
𝐴0 𝑛2 + 𝐴1 𝑛3 − 2𝐴0 𝑛2 + 4𝐴0 n − 2𝐴0 − 2𝐴1 𝑛3 + 6𝐴1 𝑛2 − 6𝐴1 𝑛 + 2𝐴1 + 𝐴0 𝑛2 − 4𝐴0 n + 𝐴1 𝑛3
− 6𝐴1 𝑛2 + 12𝐴1 𝑛 − 8𝐴1 = 𝑛
2𝐴0 − 6𝐴1 + 6𝐴1 𝑛 = 𝑛
Comparing coefficients of like powers of n
1
6𝐴1 𝑛 = 1 ⟹ 𝐴1 =
6
1
2𝐴0 − 6𝐴1 = 0 ⟹2𝐴0 − 1 = 0 ⟹ 𝐴0 =
2
1 2 1
∴𝑎𝑛 (𝑝) = 𝑛 + 𝑛3 3𝑛
2 6
1 1
𝑎𝑛 = 𝑎𝑛 (ℎ) + 𝑎𝑛 (𝑝) = 𝑏1 + 𝑛𝑏2 + 𝑛2 + 𝑛3 3𝑛 is the general solution
2 6
 GENERATING FUNCTIONS
• alternative way of representing a numeric function
• describes an infinite sequence of numbers 𝑎𝑛 by treating them as coefficients of a series
expansion
• Sum of infinite series is the generating function

Definition: The generating function for the sequence 𝑎0 , 𝑎1 , 𝑎2 , ⋯, 𝑎𝑛 , ⋯ of real numbers is

𝐺 𝑥 = 𝑎0 + 𝑎1 𝑥+ 𝑎2 𝑥 2 + ⋯ + 𝑎𝑛 𝑥 𝑛 + ⋯ = σ∞ 𝑛=0 𝑎𝑛 𝑥
𝑛

• x is just a symbol called an indeterminate.

Examples: Find the generating functions corresponding to the following sequences

1. 1,1,1,1, ⋯
1
𝐺 𝑥 = 1+ 𝑥+ 𝑥 2 + 𝑥 3 + ⋯ =
1−𝑥
2. 𝑘, 𝑘, 𝑘, 𝑘, ⋯
𝑘
𝐺 𝑥 = k+k𝑥+k𝑥 2 + 𝑘𝑥 3 + ⋯ = 𝑘 1+ 𝑥+ 𝑥 2 + 𝑥 3 + ⋯ =
1−𝑥
 GENERATING FUNCTIONS
3. 1,2, 22 , 23 , 24 , ⋯
𝟏
𝐺 𝑥 = 1+2𝑥+22 𝑥 2 + 23 𝑥 3 + ⋯ = 1+(2𝑥)+(2𝑥)2 + 2𝑥 3
+⋯=
𝟏−𝟐𝒙
4. 1,2,3,4, ⋯ , 𝑛 + 1 , ⋯
2 3 𝟏
𝐺 𝑥 = 1+2𝑥+3𝑥 + 4𝑥 + ⋯ =
(𝟏−𝒙)𝟐
Properties of Generating functions:
1. Let 𝑎𝑛 and 𝑏𝑛 be two sequences and 𝐺1 𝑥 and 𝐺2 𝑥 be their corresponding generating
functions.
Then the generating function of
a. 𝑐𝑎𝑛 is c ∙ 𝐺1 𝑥
b. 𝑎𝑛 + 𝑏𝑛 is 𝐺1 𝑥 + 𝐺2 𝑥
c. 𝑐 𝑛 𝑎𝑛 is 𝐺1 𝑐𝑥
d. 𝑎𝑛 ∗ 𝑏𝑛 is 𝐺1 𝑥 ∙ 𝐺2 𝑥
where 𝑎𝑛 ∗ 𝑏𝑛 = σ𝑛𝑚=0 𝑎𝑚 𝑏𝑛−𝑚 is called the convolution of the sequences 𝑎𝑛 and 𝑏𝑛 .
 GENERATING FUNCTIONS
Example: 2𝑛 ∗ 3𝑛 = σ𝑛𝑚=0 2𝑚 3𝑛−𝑚
= 3𝑛 σ𝑛𝑚=0 2𝑚 3−𝑚
𝑛+1
𝑛 σ𝑛 2 𝑚 1 + 𝑥 + 𝑥2 + 𝑥3 + ⋯ + 𝑥𝑛 =
1 − 𝑥
𝑖𝑓 |𝑥| < 1
= 3 𝑚=0 1−𝑥
3

𝑛 2 2 2 2 3 2 𝑛
=3 1 + + + + ⋯+
3 3 3 3

2 𝑛+1
𝑛
1− 3
𝑛+1 2 𝑛+1
= 3 2 =3 1−
1−3 3

= 3𝑛+1 − 2𝑛+1
 GENERATING FUNCTIONS
2. Shifting property:
If 𝐺(𝑥) = σ∞ 𝑎
𝑛=0 𝑛 𝑥 𝑛
generates the sequence 𝑎0 , 𝑎1 , 𝑎2 , 𝑎3 , ⋯, then
a. 𝑥 ∙ 𝐺(𝑥) generates 0, 𝑎0 , 𝑎1 , 𝑎2 , 𝑎3 , ⋯
b. 𝑥 2 ∙ 𝐺(𝑥) generates 0, 0, 𝑎0 , 𝑎1 , 𝑎2 , 𝑎3 , ⋯
c. In general , 𝑥 𝑘 ∙ 𝐺(𝑥) generates 0, 0, 0, ⋯ , 0 , 𝑎0 , 𝑎1 , 𝑎2 , 𝑎3 , ⋯
𝑘 𝑡𝑖𝑚𝑒𝑠
1
Examples: 𝐺(𝑥) = generates the sequence 1,1,1,1, ⋯ , then
1−𝑥
𝑥
a. 1−𝑥
generates 0,1,1,1,1, ⋯
𝑥2
b. 1−𝑥
generates 0, 0, 1,1,1,1, ⋯
 GENERATING FUNCTIONS
3. If 𝐺(𝑥) = σ∞ 𝑛=0 𝑛 𝑎 𝑥 𝑛 generates the sequence 𝑎 , 𝑎 , 𝑎 , 𝑎 , ⋯, then
0 1 2 3
a. 𝐺 𝑥 − 𝑎0 = σ∞ 𝑛=1 𝑛𝑎 𝑥 𝑛 generates 0, 𝑎 ,𝑎 , 𝑎 , ⋯
1 2 3
b. 𝐺 𝑥 − 𝑎0 − 𝑎1 𝑥 = σ∞ 𝑎
𝑛=2 𝑛 𝑥 𝑛
generates 0, 0, 𝑎2 , 𝑎3 , ⋯
c. 𝐺 𝑥 − 𝑎0 − 𝑎1 𝑥 − 𝑎2 𝑥 2 − ⋯ − 𝑎𝑘−1 𝑥 𝑘−1 = σ∞ 𝑛=𝑘 𝑛 𝑎 𝑥 𝑛
generates
0, 0, 0, ⋯ , 0 , 𝑎𝑘 , 𝑎𝑘+1 , 𝑎𝑘+2 , 𝑎𝑘+3 , ⋯
𝑘 𝑡𝑖𝑚𝑒𝑠

4. If 𝐺(𝑥) = σ∞ 𝑎
𝑛=0 𝑛 𝑥 𝑛
generates the sequence 𝑎0 , 𝑎1 , 𝑎2 , 𝑎3 , ⋯, then
𝐺 𝑥 −𝑎0
a. = σ∞ 𝑎
𝑛=1 𝑛 𝑥 𝑛−1 generates 𝑎 ,𝑎 ,
1 2 𝑎3 , ⋯
𝑥
𝐺 𝑥 −𝑎0 −𝑎1 𝑥
b. 𝑥2
generates 𝑎2 , 𝑎3 , 𝑎4 , 𝑎5 ⋯
 GENERATING FUNCTIONS
Example: Find the generating functions of the following sequence:
1. 0, 0,2,2,2,2, ⋯
2 2𝑥 2
Since generates 2,2,2,2,2, ⋯, G 𝑥 = generates 0, 0,2,2,2,2, ⋯
1−𝑥 1−𝑥
2. 1,1, 0, 1,1,1, 1 ⋯
1
G 𝑥 = − 𝑥2
1−𝑥
3. 3, −3, 3, −3, 3, −3 ⋯
3
G 𝑥 =
1+𝑥
4. 1, 0, −1, 0, 1, 0, −1 , 0, 1, 0, −1 , 0, ⋯
G 𝑥 = 1 + 0𝑥 − 𝑥 2 + 0𝑥 3 + 𝑥 4 + 0𝑥 5 − 𝑥 6 + 0𝑥 7 + 𝑥 8 − ⋯
= 1 − 𝑥2 + 𝑥4 − 𝑥6 + 𝑥8 − ⋯
= 1 − 𝑥 2 + (𝑥 2 )2 −(𝑥 2 )3 +(𝑥 2 )4 − ⋯
1
=
1+𝑥 2
 GENERATING FUNCTIONS
Solution of Recurrence relations using Generating functions:
1. 𝑎𝑛 = 3𝑎𝑛−1 + 2 , n ≥ 1, 𝑎0 = 1 ----- (1)
Solution: Let 𝐺(𝑥) = σ∞ 𝑛
𝑛=0 𝑎𝑛 𝑥 be the generating function of sequence {𝑎𝑛 }
Multiply each term of (1) by 𝑥 𝑛 and take sum from 1 to ∞
σ∞ 𝑎
𝑛=1 𝑛 𝑥 𝑛
= 3 σ ∞
𝑎
𝑛=1 𝑛−1 𝑥 𝑛
+ 2 σ ∞
𝑛=1 𝑥 𝑛 ∞

1 ෍ 𝑎𝑛 𝑥 𝑛 = 𝐺 𝑥 − 𝑎0
𝐺 𝑥 − 𝑎0 = 3𝑥𝐺 𝑥 + 2 −1 𝑛=1
1−𝑥
1−(1−𝑥)
𝐺 𝑥 − 1 = 3𝑥𝐺 𝑥 + 2 ∞
1−𝑥 ෍ 𝑎𝑛−1 𝑥 𝑛 = 𝑥𝐺 𝑥
𝑥
𝐺 𝑥 − 3𝑥𝐺 𝑥 = 1 + 2 𝑛=1
1−𝑥
1+𝑥 ∞
𝐺 𝑥 (1 − 3𝑥) = ෍ 𝑥𝑛 =
1
1−𝑥
1−𝑥
1+𝑥 𝑛=0
𝐺 𝑥 =
1−𝑥 (1−3𝑥)
 GENERATING FUNCTIONS
𝐴 𝐵
𝐺 𝑥 = + ∞
1−3𝑥 (1−𝑥) 1
2 1 = ෍ 𝑎𝑛 𝑥 𝑛
= − 1 − 𝑎𝑥
(1−3𝑥) 1−𝑥 𝑛=0

σ∞ 𝑛 ∞ 𝑛 𝑛
𝑛=0 𝑎𝑛 𝑥 = 2 σ𝑛=0 3 𝑥 − σ𝑛=0 𝑥
∞ 𝑛

= σ∞ 𝑛=0 (2 ∙ 3𝑛 −1) 𝑥 𝑛

∴𝑎𝑛 = 2 ∙ 3𝑛 − 1

2. 𝑎𝑛 = 2𝑎𝑛−1 + 1 , n ≥ 1, 𝑎0 = 1 ----- (1)

Solution: Solution: Let 𝐺(𝑥) = σ∞ 𝑛
𝑛=0 𝑎𝑛 𝑥 be the generating function of sequence {𝑎𝑛 }
Multiply each term of (1) by 𝑥 𝑛 and take sum from 1 to ∞
σ∞ 𝑛 ∞ 𝑛
𝑛=1 𝑎𝑛 𝑥 = 2 σ𝑛=1 𝑎𝑛−1 𝑥 + σ𝑛=1 𝑥
∞ 𝑛
1
𝐺 𝑥 − 𝑎0 = 2𝑥𝐺 𝑥 + −1
1−𝑥
1−(1−𝑥)
𝐺 𝑥 − 1 = 2𝑥𝐺 𝑥 +
1−𝑥
 GENERATING FUNCTIONS
𝑥
𝐺 𝑥 − 2𝑥𝐺 𝑥 = 1 +
1−𝑥
1
𝐺 𝑥 (1 − 2𝑥) =
1−𝑥
1
𝐺 𝑥 =
1−𝑥 (1−2𝑥)
𝐴 𝐵
𝐺 𝑥 = +
1−2𝑥 (1−𝑥)
2 1
= −
(1−2𝑥) 1−𝑥

σ∞ 𝑛 ∞ 𝑛 𝑛 ∞
𝑛=0 𝑎𝑛 𝑥 = 2 σ𝑛=0 2 𝑥 − σ𝑛=0 𝑥
𝑛

= σ∞
𝑛=0(2 ∙ 2 𝑛 −1) 𝑥 𝑛

∴𝑎𝑛 = 2 ∙ 2𝑛 − 1
= 2𝑛+1 − 1
 GENERATING FUNCTIONS
3. 𝑎𝑛 − 9𝑎𝑛−1 + 20𝑎𝑛−2 = 0, n ≥ 2, 𝑎0 = −3, 𝑎1 = −10 ----- (1)
Solution: Let 𝐺(𝑥) = σ∞ 𝑛
𝑛=0 𝑎𝑛 𝑥 be the generating function of sequence {𝑎𝑛 }

Multiply each term of (1) by 𝑥 𝑛 and take sum from 2 to ∞

σ∞ σ∞ 𝑛
𝑛=2 𝑎𝑛 𝑥𝑛 − 9 σ∞ 𝑎
𝑛=2 𝑛−1 𝑥 𝑛 + 20 σ∞ 𝑎
𝑛=2 𝑛−2 𝑥 𝑛 =0 𝑛=2 𝑎𝑛 𝑥 = 𝐺 𝑥 − 𝑎0 − 𝑎1 x

𝐺 𝑥 − 𝑎0 − 𝑎1 𝑥 − 9𝑥 𝐺 𝑥 − 𝑎0 + 20𝑥 2 𝐺(𝑥) = 0 ∞

෍ 𝑎𝑛−1 𝑥 𝑛 = 𝑥(𝐺 𝑥 − 𝑎0 )
𝐺 𝑥 + 3 + 10𝑥 − 9𝑥𝐺 𝑥 − 27𝑥 + 20𝑥 2 𝐺(𝑥) = 0 𝑛=2

𝐺 𝑥 − 9𝑥𝐺 𝑥 + 20𝑥 2 𝐺 𝑥 = −3 + 17𝑥 ∞

෍ 𝑎𝑛−2 𝑥 𝑛 = 𝑥 2 𝐺 𝑥
𝐺 𝑥 (1 − 9𝑥 + 20𝑥 2 ) = −3 + 17𝑥 𝑛=2
−3+17𝑥 −3+17𝑥
𝐺 𝑥 = =
(1−9𝑥+20𝑥 2 ) (1−5𝑥)(1−4𝑥)
 GENERATING FUNCTIONS
𝐴 𝐵
𝐺 𝑥 = +
1−5𝑥 (1−4𝑥)
2 5
= −
(1−5𝑥) 1−4𝑥

σ∞ 𝑎
𝑛=0 𝑛 𝑥 𝑛
= 2 σ ∞
𝑛=0 5𝑛 𝑛
𝑥 − 5 σ ∞ 𝑛 𝑛
𝑛=0 4 𝑥

= σ∞ 𝑛 𝑛
𝑛=0(2 ∙ 5 −5 ∙ 4 ) 𝑥
𝑛

∴𝑎𝑛 = 2 ∙ 5𝑛 − 5 ∙ 4𝑛

4. 𝑎𝑛 = 8𝑎𝑛−1 + 10𝑛−1 , n ≥ 1, 𝑎0 = 1 ----- (1)

Solution: Solution: Let 𝐺(𝑥) = σ∞ 𝑛
𝑛=0 𝑎𝑛 𝑥 be the generating function of sequence {𝑎𝑛 }
Multiply each term of (1) by 𝑥 𝑛 and take sum from 1 to ∞
σ∞ 𝑛 ∞ 𝑛
𝑛=1 𝑎𝑛 𝑥 = 8 σ𝑛=1 𝑎𝑛−1 𝑥 + σ𝑛=1 10
∞ 𝑛−1 𝑥 𝑛
𝑥
𝐺 𝑥 − 𝑎0 = 8𝑥𝐺 𝑥 +
1−10𝑥
 GENERATING FUNCTIONS
𝑥
𝐺 𝑥 − 1 = 8𝑥𝐺 𝑥 +
1−10𝑥
𝑥
𝐺 𝑥 − 8𝑥𝐺 𝑥 = 1 +
1−10𝑥
1−9𝑥
𝐺 𝑥 (1 − 8𝑥) =
1−10𝑥
1−9𝑥
𝐺 𝑥 =
1−10𝑥 (1−8𝑥)
𝐴 𝐵
𝐺 𝑥 = +
1−10𝑥 (1−8𝑥)
1 1 1
= +
2 (1−10𝑥) 1−8𝑥
1 ∞ 1 ∞
σ∞
𝑛=0 𝑎𝑛 𝑥𝑛 = σ𝑛=0 10𝑛 𝑥 𝑛 + σ𝑛=0 8𝑛 𝑥 𝑛
2 2
1
= σ∞ 𝑛 𝑛
𝑛=0 (10 + 8 ) 𝑥
𝑛
2
1
∴𝑎𝑛 = (10𝑛 + 8𝑛 )
2
 Important Series Expansion
1
1) 1 + 𝑥 + 𝑥 2 + 𝑥 3 + ⋯ = , |𝑥| < 1
1−𝑥
1
2) 1 − 𝑥 + 𝑥 2 − 𝑥 3 + ⋯ = , |𝑥| < 1
1+𝑥
1
3) 1 + 2𝑥 + 3𝑥 2 + 4𝑥 3 + ⋯ + ∞ = (1−𝑥)2 , |𝑥| < 1
1
4) 1 − 2𝑥 + 3𝑥 2 − 4𝑥 3 + ⋯ ∞ = (1+𝑥)2 , |𝑥| < 1
𝑥2 𝑥3 𝑥4
5) 𝑥 + + + + ⋯ ∞ = −log(1 − 𝑥) , |𝑥| < 1
2 3 4
𝑥2 𝑥3 𝑥4
6) 𝑥 − + − + ⋯ ∞ = log(1 + 𝑥) , |𝑥| < 1
2 3 4

𝑥 𝑥2 𝑥3 𝑥4
7) 1 + + + + + ⋯ ∞ = 𝑒 𝑥 , ∀𝑥
1! 2! 3! 4!
𝑥3 𝑥5 𝑥7
8) 𝑥 − 3!
+
5!
−
7!
+ ⋯ ∞ = sin 𝑥 , ∀ 𝑥
𝑥2 𝑥4 𝑥6
9) 1 − + − + ⋯ ∞ = cos 𝑥 , ∀𝑥
2! 4! 6!
THANK YOU!