Gulzar. World Journal of Engineering Research and Technology
ON ZEROS OF POLYNOMIALS
M. H. Gulzar*
Department of Mathematics, University of Kashmir, Srinagar, India.
Article Received on 24/01/2017 Article Revised on 14/02/2017 Article Accepted on 09/03/2017
ABSTRACT
In this paper we find ring-shaped regions containing all or a specific
number of zeros of a polynomial. Many important results follow easily
from our results.
Mathematics Subject Classification: 30C10, 30C15. KEYWORDS AND PHRASES: Polynomial, Region, Zeros.
INTRODUCTION
A classical result on the location of zeros of a polynomial is the following known as the
Enestrom-Kakeya Theorem:[2,3]
Theorem A: Let
n
j j jz
a z
P
0 )
( be a polynomial of degree n such that
0
... 1 0
1
a a a
an n .
Then all the zeros of P(z) lie in z 1.
Another classical result giving a region containing all the zeros of a polynomial is the
following known as Cauchy’s Theorem:[2,3]
Theorem B: All the zeros of the polynomial
n
j j jz
a z
P
0 )
( of degree n lie in the circle
M
z 1 , where
n j n j
a a M max0 1 .
The above theorems have been generalized in various ways by the researchers.
World Journal of Engineering Research and Technology
WJERT
www.wjert.org
SJIF Impact Factor: 4.326*Corresponding Author
M. H. Gulzar
Department of
MAIN RESULTS
In this paper we prove the following:
Theorem 1: Let
n
j j jz
a z
P
0 )
( be a polynomial of degree n and
0 0 1 2
1
1 a a ... a a a
a a
L n n n n .
Then all the zeros of P(z) lie in
n n
n
a L z a L a R
a
]
[ 0
1 0
for R1
and in
n
n a
L z a L a R
a
]
[ 0
0
for R1, provided an L.
Further the number of zeros of P(z) in , 1
]
[ 0
1
0
c
c R z a L a R
a
n
n does not exceed
0
0 1
0 [ ]
log log
1
a
a L a R a c
n
n
for R1 and the number of zeros of P(z) in
1 , ]
[ 0
0
c c
R z a L a R
a
n
does not exceed
0
0
0 [ ]
log log
1
a
a L a R a c
n
for R1.
Remark 1: If an an1...a1 a0 0, then L an and it follows from Theorem 1 that all the zeros of P(z) lie in z 1, which is Theorem A i.e. the Enestrom-Kakeya Theorem.
If we take R=1 in Theorem 1, we get the following result:
Corollary 1: Let
n
j j jz
a z
P
0 )
( be a polynomial of degree n and
L anan1 an1an2 ... a1a0 a0 .
Then the number of zeros of P(z) in , 1
0
0
c c
R z a L a
a
n
does not exceed
0 ) (
log log
1
a a L c
n
.
Instead of proving Theorem 1, we prove the following more general result:
Theorem 2: Let
n
j j jz
a z
P
0 )
( be a polynomial of degree n with
n j
a
aj) j,Im( j) j, 0,1,2,...,
. ... ... 0 0 1 2 1 1 0 0 1 2 1 1 n n n n n n n n M L
Then all the zeros of P(z) lie in
n n n a M L z M L a R a ]
[ 0 0
1
0
for R1
and in n n a M L z M L a R a ]
[ 0 0
0
for R1, provided an LM.
Further the number of zeros of P(z) in , 1,
]
[ 0 0
1
0
c c R z M L a R a n
n does not
exceed
0
0 0 1
0 [ ]
log log 1 a M L a R a c n
n
for R1 and the number of zeros of P(z) in
, 1 , ]
[ 0 0
0
c c
R z M L a R a
n
does not exceed
0
0 0
0 [ ]
log log 1 a M L a R a c
n
for R1.
Remark 2: Taking aj real i.e. j 0,j0,1,2,...,n, Theorem 2 reduces to Theorem 1. If we take R=1 in Theorem 2, we get the following result:
Corollary 2: Let
n j j jz a z P 0 )
( be a polynomial of degree n with
n j
a
aj) j,Im( j) j, 0,1,2,...,
Re( and
. ... ... 0 0 1 2 1 1 0 0 1 2 1 1 n n n n n n n n M L
Then the number of zeros of P(z) in 1, 1,
0 0
0
L M z c c
a
a
n
does not exceed
0 0 0 0 log log 1 a M L a a c
n
LEMMAS
For the proof of Theorem 2, we need the following results:
Lemma 1: Let f(z) (not identically zero) be analytic for z R,f(0)0 and f(ak)0,
n
k 1,2,..., . Then
n
j j
i
a R f
d f
1 2
0 log (Re log (0) log
2
1
.
Lemma 1 is the famous Jensen’s Theorem (see page 208 of [1]).
Lemma 2: Let f (z) be analytic for z R,f(0)0 and f(z) M for z R. Then the number of zeros of f(z) in ,c1
c R
z does not exceed
) 0 ( log log
1
f M
c .
Lemma 2 is a simple deduction from Lemma 2.
PROOFS OF THEOREMS
Proof of Theorem 2: Consider the polynomial
) ( ) 1 ( )
(z z P z
F
a z a a z
a z
a a z
a
a z a z
a z a z
n n n n
n
n n n n
) (
) (
... )
(
) ...
)( 1 (
1 1
1 1
1
0 1 1
1
0 0
1 )
(
... a a za
... )
{( )
( ... )
( 1 1 0 0 1
1
n
n n n
n n n
nz z z i z
a
} )
(10 0
z
For z 1so that j n
z j
,..., 2 , 1 , 1
1
, we have, by using the hypothesis
... ...
{ )
( 1 1 0 0 1
1
n
n n n
n n n
n z z z z
a z
F
} 0 0
1
z
1 0
1 0 1 2
1
1 . ...
{
[
n n n n n n
n n n
n
z z
z z
a
z
}]
... 1 10 0
1
n n
n n
z z
z
1 0
0 1 2
1
1 . ...
{
[
n n n n n n n
n z a
z
}]
...
0
)] (
[
zn an z L M
if
n
a M L z
provided an LM.
This shows that those zeros of F(z) whos modulus is greater than 1 lie in
n
a M L z .
Since the zeros of F(z) whose modulus is less than or equal to 1 already satisfy the above
inequality, it follows that all the zeros of F(z) and hence all the zeros of P(z) lie in
n
a M L z .
On the other hand, we have
... )
{( ) (
... )
( )
( 0 1 1 0 1
1
n
n n n
n n n
nz a z z i z
a z
F
} ) (1 0 z
) ( 0 G z
a
Where G(z)anzn1(n n1)zn...(10)zi{(nn1)zn ...
} ) (10 z
.
For z R, we have, by using the hypothesis
z z
z z
z a z
G n n n
n n n n
n 1 1 0 1 1 0
1
... ...
)
(
] [
] ...
... [
... ...
0 0 1
0 1 1
0 1 1
1
0 1 1
0 1 1
1
M L a R
a R
R R
R R
R a
n n
n n n
n n n
n n n n
n n n n
for R1. For R1,
] [
)
(z R a LM 0 0
G n .
Since G(z) is analytic for z R,G(0)0, it follows by Schwarz Lemma that in z R,
z M
L a R z
G n
n
] [
)
( 0 0
1
for R1 and
z M
L a R z
Hence for z R,
) ( )
(z a0 G z
F
z M
L a R a
z G a
n n
] [
) (
0 0 1
0 0
for R1 and
F(z) a0 R[an LM0 0 ]z
for R1.
Thus for R1, F(z) 0if
]
[ 0 0
1
0
M L a R
a z
n
n
and for R1, F(z) 0if
]
[ 0 0
0
M L a R
a z
n
.
In other words, all the zeros of F(z) lie in
]
[ 0 0
1
0
M L a R
a z
n
n for R1 and in
]
[ 0 0
0
M L a R
a z
n
for R1.
Since the zeros of F(z) are also the zeros of P(z), it follows that all the zeros of P(z) lie in
]
[ 0 0
1
0
M L a R
a z
n
n for R1 and in
]
[ 0 0
0
M L a R
a z
n
for R1.
Again , for z R, we have, by using the hypothesis
F(z) an zn1 a0 n n1 zn ...1 0 z n n1 zn ... 10 z
0 1 ... 1 0 1 ...
1
n
n n n
n n n
nR a R R R
a
10 R
1[ 1 ... 1 0 1 ...
0
n n n
n n
n a
R
a
10]
a0 Rn1[an LM 0 0]
for R1 ,
F(z) a0 R[an LM0 0].
Hence, by using Lemma 2 and the above observations, it follows that the number of zeros of
F(z) and therefore P(z) in , 1,
]
[ 0 0
1
0
c
c R z M
L a R
a
n
n does not exceed
0
0 0 1
0 [ ]
log log
1
a M L a R a c
n
n
for R1 and the number of zeros of P(z) in
, 1 , ]
[ 0 0
0
c c
R z M
L a R
a
n
does not exceed
0
0 0
0 [ ]
log log
1
a M L a R a c
n
for R1.
That completes the proof of Theorem 2.
REFERENCES
1. Ahlfors L. V. Complex Analysis, 3rd edition, Mc-Grawhill.
2. Marden M, Geometry of Polynomials, Mathematical Surveys Number 3, Amer. Math.
Soc. Providence, RI, 1966.
3. Q. I. Rahman and G. Schmeisser, Analytic Theory of Polynomials, Oxford University