Fixed Points Theorems In Three Metric Spaces

Page 1

Manish Kumar Mishra et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES Vol No. 1, Issue No. 2, 130 - 134

Fixed Points Theorems In Three Metric Spaces ( FPTTMS) Manish Kumar Mishra and Deo Brat Ojha

mkm2781@rediffmail.com, deobratojha@rediffmail.com

Department of Mathematics R.K.G.Institute of Technology Ghaziabad,U.P.,INDIA

satisfying integral type inequality.

for all x in X, y in Y and z in Z, where 0≤ c < 1. Then RST has

Mathematics Subject Classification: 54H25

a unique fixed point u in X, TRS has a unique fixed point v in Y and STR has a unique fixed point w in Z. Further Tu = v,

ES

Keywords- Three metric space, fixed point, integral type inequality. .

d ( RSTx, RSy )  c max{d ( x, RSy ), d ( x, RSTx),  ( y, Tx),  ( Sy, STx)}  (TRSy, STz )  c max{ ( y, TRz ),  ( x, TRSy),  ( z, Sy), d ( Rz, RSy)}   ( STRz, STx)  c max{ ( z, STx),  ( z, STRz), d ( x, Rz),  (Tx, TRz)}

T

Abstract— We obtained fixed point theorem on three metric spaces

Sv = w and Rw = u. The next theorem was proved in [3]. Theorem 1.3 : Let ( X , d ) , (Y ,  ) and ( Z ,  ) be complete

I.

INTRODUCTION

metric spaces and suppose T is a continuous mapping of X

The following fixed point theorem was proved by Fisher [1].

into Y, S is a continuous mapping of Y into Z and R is a continuous mapping of Z into X satisfying the inequalities:

complete metric spaces. If S is a continuous

mapping of X into Z, and R is a continuous

d ( RSTx, RSTx ')  c max{d ( x, x '), d ( x, RSTx), d ( x ', RSTx '),  (Tx, Tx '),   STx, STx ' }

IJ A

Theorem 1.1: Let ( X , d ) and ( Z ,  ) be

mapping

of

Z

into

X

satisfying

the

inequalities:

d ( RSx, RSx ')  c max{d ( x, x '), d ( x, RSx), d ( x ' RSx '), d ( Sx, Sx ')}   (SRz, SRz ')  c max{ ( z, z '),  ( z, SRz),  ( z ' SRz '),  ( Rz, Rz ')}

 (TRSy, TRSy ')  c max{ ( y, y '),  ( y, TRSy),  ( y ', TRSy '),  ( Sy, Sy '), d  RSy, RSy ' } 

 (STRz, STRz ')  c max{ ( z, z '),  ( z, STRz),  ( z ', STRz '), d  Rz, Rz '   (TRz, TRz ')}

for all x, x in X, y, y in Y and z, z in Z where 0  c <1. Then RST has a unique fixed point u in X, TRS

for all x, x' in X, and z, z' in Z, where 0≤ c < 1, then

has a unique fixed point v in Y and STR has a unique

RS has a unique fixed point u in X and RS has a

fixed point w in Z. Further, Tu = v, Sv = w and Rw = u. In

unique fixed point w in Z. Further Su = w and Rw =

recently Ansari , Sharma[6] generate Related Fixed Points

u. The next theorem was proved in [2].

Theorems on Three Metric Spaces. Now We obtained related

Theorem 1.2: Let

( X , d ) , (Y ,  ) and ( Z ,  ) be complete

metric spaces and suppose T is a continuous mapping of X into Y, S is a continuous mapping of Y into Z and R is a continuous mapping of Z into X satisfying the inequalities:

ISSN: 2230-7818

fixed point theorem on three metric spaces satisfying integral type inequality. Let

(X,d ) be

 [0,1], f : X  X

@ 2010 http://www.ijaest.iserp.org. All rights Reserved.

a

complete

metric

space,

a mapping such that for each

Page 130


Manish Kumar Mishra et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES Vol No. 1, Issue No. 2, 130 - 134 d ( fx , fy )

x , y X ,

 (t )dt  

d ( x, y )

0

 ; R  R summable,

  (t )dt ,

Where

such

that,

for

each

  0,   (t )dt  0 .

d 2p (TSz ,TSRy )

d3p

0

is a lebesgue integrable mapping which is and

d1p ( SRy , SRTx )

0

0

nonnegative

Then f has a unique common fixed

 max M 1 ( x , y )

 max M 2 ( y , z )

0

 (t ) dt 

 max M 3 ( z , x )

0

Proof. Let

condition by the following

sequences



 (t )dt 

0

and

inequality

J P ( Fx , Fy )

 (t ) dt 

0

IJ A

0

ad ( fx , fy ) d P 1 ( fx , Fx ), ad ( fx , fy ) d P 1 ( fy , Fy ),   P 1 P 1 ad ( fx , Fx ) d ( fy , Fy ), cd ( fx , Fy ) d ( fy , Fx )   (t ) dt

for all

x , y in X ,where p  2 is an integer a  0 and

0  c  1 then f and F have unique common fixed point

x, y

in

0  c  1 then f X

X

,where

and

F

p2

is an integer

a0

 all (t )dt n  N

.

We

will

Taking

spaces and

mappings satisfying the following inequalities:

 (t ) dt 

and

.

Similarly, if

in (1) and (4), we obtain:

{ d1 ( xn , SRy n ), d1 ( xn , SRTxn ), d 2 ( y n , Txn )} p

p

p

0



{ d1p

( xn , xn ), d1p

( xn , xn 1 ), d 2p



{0, d1p ( xn , xn 1 ), d 2p ( yn , yn 1 )}

( xn , xn 1 )

0

 (t )dt  

d1p

( SRyn , SRTxn )

0

( yn , yn 1 )}





 max{0, d1p ( xn , xn 1 ), d 2p ( yn , yn 1 )}



0

 (t )dt  

( yn , yn 1 )

since, if

d1p ( xn , xn1 )

0

 (t )dt  

 (t )dt

F 0

0

d 2p

 (t )dt

 (t ) dt

0

0

 (t ) dt

 (t )dt

F  min M 1 ( xn , yn ) 

 max M1 ( xn , yn )

 (t ) dt

 (t )dt

max M1 ( xn , yn )

0

 (t )dt  

 (t ) dt 

 d1p ( xn , xn1 )

0

d1 ( xn , xn 1 ) p

0

 (t ) dt

then by the inequality

 (t )dt

it follows xn 1  xn since 0    1 .

@ 2010 http://www.ijaest.iserp.org. All rights Reserved.

If

zn  zn1 ,

Thus

ISSN: 2230-7818

and we

xn  , yn  and

0

d1p

0

are three

for all n. Otherwise, if

the latter equality implies that

0

MAIN RESULTS

T : X Y, R :Y  Z, S : Z  X

that

xn  xn1 .

x  xn y  yn

M1 ( xn , yn )

0

be three metric

assume

 zn  are Cauchy sequences.

have unique common fixed point in

( X , d1 ) , (Y , d 2 ) and ( Z , d3 )

. Further,

xn  a, yn1  b and zn1  c

First, we prove that the sequences

Now, we will give and prove our theorem as follows: Theorem 2.1 Let

cZ

yn1  yn2 , zn1  zn2

for some n, then

put

then again

.

II.

SRT

be an arbitrary point. We define three

yn  yn1 , then zn  zn1 and

in X .

for all

then

has a unique fixed point

SRTxn1  SRTxn , i.e. xn  xn1 .

max 

 6

n

could

commutative (OWC) and satisfy the

 (t )dt

xn  , yn   and  zn  with X, Y, Z respectively as

xn  xn1

a

multi-valued map respectively, suppose that f and F are occasionally weakly

x X

ES

f : X  X , F : X  CB( X ) be a single

 5

0   1 ,

xn  xn1 , yn  yn1 and zn  zn1

Let ( X , d ) be a metric space and let

Ojha (2010) [5]

 (t )dt

T

F  min M 3 ( z , x ) 

0

 4

xn   SRT  x0 , yn  Txn1 , zn  Ryn

follows:

1 max{d ( x , y ), d ( x , fx ), d ( y , fy ), [ d ( x , fy )  d ( y , fx )] 2 for 0



 (t )dt

has a unique fixed point

Rhoades(2003)[4], extended this result by replacing the above

d ( fx , fy )

0

, where

Ta  b, Rb  c and Sc  a.

n 

F  min M 2 ( y , z ) 



 (t ) dt

z  X such that for each x  X , lim f x  z. n

F  min M1 ( x , y ) 

0

a  X , TSR

has a unique fixed point

b  Y and RTS

0

 (t ) dt

x  X , y Y , z  Z

for all



 (t ) dt

0

 (t ) dt 

( RTx , RTSz )

0

 (t ) dt 

Page 131


Manish Kumar Mishra et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES Vol No. 1, Issue No. 2, 130 - 134

d1p ( xn , xn1 )

 (t )dt  

0

 d2p ( yn , yn1 )

0

Taking

y  yn , z  zn1

M 2 ( yn , zn 1 )

0

 (t ) dt 

p

p

0



{ d 2p ( yn , yn ), d 2p ( yn , yn 1 ), d 3p ( zn 1 , zn )}



{0, d 2p

( yn , yn 1 ), d3p

 (t )dt  

( zn 1 , zn )}

d 2p (TSzn 1 ,TSRyn )

0





 max{0, d 2p ( yn , yn 1 ), d3p ( zn 1 , zn )}



 d3p ( zn 1 , zn )

0

 (t )dt  

 (t )dt  

0

 max M1 ( xn , b )

0

 (t )dt

F 0

M1 ( xn , b )

0

0



0

0

 (t ) dt 

d 2 ( yn , yn 1 ) p

0

( yn , yn1 )

 (t ) dt

then by the

d1p ( SRb , a )

 (t )dt  

 d3p ( zn1 , zn )

0

8

 (t )dt

0

 (t ) dt 

{ d 3 ( z n , RTxn 1 ), d 3 ( z n , RTSz n ), d1 ( xn 1 , Sz n )} p

p

p

0



{ d3p

( zn , zn ), d3p

( zn , zn 1 ), d1p



{0, d3p ( zn , zn 1 ), d1p ( xn 1 , xn )}

( xn 1 , Szn )}

0

( zn , zn 1 )

0

 (t )dt  

d3p

( RTxn 1 , RTSzn )

0





 max{0, d3p ( zn , zn 1 ), d1p ( xn 1 , xn )}



0

 (t )dt  

0

0

( xn 1 , Szn )

0

Replacing

 (t )dt  

d3p ( zn1 , zn )

n

 d1p ( xn2 , xn1 )

0

 (t )dt

9

 (t )dt

Using (7), (8) and (9) we get

d1p ( xn , xn 1 )

0

 (t ) dt  

 d 2p ( yn , yn 1 )

0



 3 d1p ( xn  2 , xn 1 )

0

 (t ) dt  

 2 d3p ( zn 1 , zn )

0

 (t )dt  

 4 d1p ( xn  4 , xn 3 )

0

  (t )dt  0  .........  3 k p  d1 ( x0 , x1 )   (t ) dt  0  3 k d1p ( x1 , x2 )

M 1 ( a , yn )

 (t )dt

 (t )dt

{ d1p ( a , SRyn ), d1p ( a , SRTa ), d 2p ( yn ,Ta )}

 (t )dt  

0

{ d1p

( a , xn ), d1p

( a , SRTa ), d 2p

( yn ,Ta )}

 (t )dt

d1p ( a , SRTa )

0

 (t )dt  

 max{d1p ( a , a ), d1p ( a , SRTa ), d 2p ( b ,Ta )}

0



 max{d1p ( a , SRTa ), d 2p ( b ,Ta )}

0

 (t )dt

 (t )dt

from which it follows or

d1p ( a , SRTa )

0

or

 (t )dt  

 d1p ( a , SRTa )

0

d1p ( a , SRTa )

0

 (t )dt  

 (t )dt  SRTa  a

 d 2p ( b ,Ta )

0

 (t )dt

which can be also written in the following form

d1p ( a , Sc )

0

 (t )dt  

 d2p ( b ,Ta )

0

 (t )dt

11

since RT a = c.

 (t ) dt

Taking,

 (t )dt

z  zn , y  b

above we

for n  2k

obtain:

xn , yn   and  zn 

 (t )dt

Letting n tend to infinity we get

for n  2k  1

Since 0    1 , the sequences

F  min M1 ( a , yn ) 

0

n  1 we obtain:

 (t )dt  

0

 (t )dt  



 (t )dt with

d1p ( SRyn , SRTa )

0

 (t ) dt

F 0

x  a , y  yn we get

 (t )dt  

 max M1 ( a , yn )

0

 (t )dt

0

d1p

SRb  a .

where

 (t )dt

F  min M 3 ( zn , xn 1 ) 

 max M 3 ( zn , xn 1 )

 (t )dt

0

0

 (t ) dt

 (t )dt

d1p ( xn , SRTa )



IJ A

0

d3p

 d1p ( a , SRb )

0

0

Taking x  xn 1 z  zn in (3) and (6) we obtain: M 3 ( zn , xn 1 )

 (t )dt  

Using (4), if we take

Thus 0

 (t )dt

In the same way it can be shown that T Sc = b and RT a = c.

 (t )dt

it follows yn  yn 1 since 0    1 .

 (t )dt

Letting n tend to infinity in the inequality (10) and by the fact that F

from which it follows  d2p

0

d2p ( yn , yn1 )

{d1p ( xn , SRb ), d1p ( xn , xn 1 ), d 2p ( b , yn 1 )}

is continuous in 0 we get

max{d 2p ( yn , yn 1 ), d3p ( zn 1 , zn )}

 (t )dt  

10 

 (t )dt

{ d1p ( xn , SRb ), d1p ( xn , SRTxn ), d 2p ( b ,Txn )}

ES

( yn , yn1 )

0

 (t )dt

F  min M1 ( xn , b ) 

 (t )dt  

 (t )dt  

 (t )dt

inequality

d1p ( SRb , SRTxn )

0

0

d2p

 (t )dt  

0



 (t )dt

since, if

d1p ( SRb , xn1 )

y  b in the inequality (4) we obtain

and

where

0

0

 (t )dt

 (t )dt

0

0

 (t ) dt

 (t )dt

F  min M 2 ( yn , zn 1 ) 

 max M 2 ( yn , zn 1 )

n 

x  xn

Taking

p

0

0

n 

in (2) and (5), we obtain:

0

d 2p ( yn , yn 1 )

lim xn  a  X , lim yn  b  Y , lim zn  c  Z

n 

{ d 2 ( yn , TSz n 1 ), d 2 ( yn , TSRyn ), d 3 ( zn 1 , Ryn )}

have

7

 (t )dt

T

are Cauchy

d2p ( b ,Ta )

0

 (t )dt  

 d3p ( c , Rb )

0

in the inequality (5) in the same way as

 (t )dt

12

In the same way, we obtain:

sequences. Since ( X , d1 ) , (Y , d 2 ) and ( Z , d3 ) are complete metric spaces we

ISSN: 2230-7818

@ 2010 http://www.ijaest.iserp.org. All rights Reserved.

Page 132


Manish Kumar Mishra et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES Vol No. 1, Issue No. 2, 130 - 134

d3p ( c , Rb )

 d1p ( a , Sc )

 (t )dt  

0

0

13

 (t )dt

( a , Sc )

 (t )dt  

0

3

d1p

( a , Sc )

0

Thus, again

d1p ( a , SRTa )

0

d3p ( RTa ', c )

 (t )dt  

0

d3p ( RTa ', c )

0

 (t )dt  Sc  a.

 (t )dt  

 d1p ( a , a ')

 (t )dt  

16 

 (t )dt

0

or 

By (11), (12), (13) it follows: d1p

 d3p ( RTa ', c )

0

17 

 (t )dt

By (17) it follows that RTa '  c since 0    1 .

d1p ( a , Sc )

0

 (t )dt  0

By (14), (15) and (16) we obtain

d1p ( a , a ')

0

or SRTa  a

0

and by

So, we proved that a is a fixed point of SRT .

 d2p (Ta ',b )

 (t )dt  

d1p ( a , a ')

 (t )dt  

 (t )dt  

0

 2 d3p ( RTa ', c )

0

 3 d1p ( a , a ')

0

 (t )dt  

 3 d1p ( a , a ')

0

 (t )dt

 (t )dt where 0    1 , it follows

a'  a .

In the same way it can be shown that b is a fixed point of T SR and c

that

is a fixed point of RT S.

Thus we proved that a is the unique fixed point of SRT .

Further, we showed that T a = b, Rb = c, Sc = a.

In the same way, it can be shown that b is the unique fixed point of T

is another fixed point of SRT ,

x  a, y  Ta ' , we get

different from a. Using (4), if we take

d1p ( a , a ')

0

 (t )dt  

d1p ( SRTa ', SRTa )

0



 max M1 ( a ,Ta ')

0

 (t )dt  

 (t )dt

F  min M1 ( a ,Ta ') 

0

 (t )dt

0

{ d1p ( a , SRTa '), d1p ( a , SRTa ), d 2p (Ta ',Ta )}

0



{ d1p ( a , a '), d1p ( a , a ), d 2p (Ta ', b )}

0

Z  Y , d1  d2 ), and the mapping

( a , a ')

0

 (t )dt  

(Ta ', b )

 (t )dt

 (t )dt  

14

in the inequality (5) we obtain:

d2p (TSc ,TSRTa ')

0

 (t )dt



 max M1 (Ta ', c )

0

F  min M 2 (Ta ', c ) 

 (t )dt  

0

 (t )dt

where

M 2 (Ta ', c )

0

 (t )dt  

{ d 2p (Ta ',TSc ), d 2p (Ta ',TSRTa '), d3p ( c , RTa ')}

 (t )dt

0



{ d 2p

(Ta ', b ), d 2p

(Ta ',Ta '), d3p

( c , RTa ')}

0

 (t )dt

Then

d2p ( b ,Ta ')

0

d3p ( c , RTa ')

0

Taking

 (t )dt  

z  c, y  Ta '

d3p ( RTa ',c )

0

 (t )dt  

 max M 3 ( c , a ')

0

 (t )dt  

15

 (t )dt

in the inequality (6) we obtain:

d3p ( RTa ', RTSc )

0



F  min M 2 ( c , a ') 

0

 (t )dt

M 3 ( c , a ')

0

 (t )dt  

{d3p ( c , RTa '), d 2p ( c , RTSc ), d1p ( a ', Sc )}

0



{ d3p ( c , RTa '),0, d1p ( a ', a )}

0

d1p ( Sy , STx )

0

 (t )dt

 (t ) dt 

M1 ( x , y )

0

d2p (TSy ,TSy )

0

apply

the

inequalities

(4)(5)

and

(6)

for

as the identity mapping.

R

 max M 1 ( x , y )

0

 (t ) dt 

p { d1

 (t ) dt

p ( x , Sy ), d1

F  min M1 ( x , y ) 

0

p ( x , STx ), d 2

( y , Tx )}

0

 (t ) dt 

 max M 2 ( y , y )

0

 (t ) dt 

 (t )dt

 (t ) dt

F  min M 2 ( y , y ) 

0

 (t )dt

The inequality (6) takes the form:

d2p (Tx ,TSy )

 (t ) dt 

0

 max M 3 ( y , x )

0

 (t ) dt 

F  min M 3 ( y , x ) 

0

 (t )dt

where

M3 ( y , x )

0

 (t ) dt 

{ d 2 ( y , Tx ), d 2 ( y , TSy ), d1 ( x , Sy )} p

p

p

0

x  X , y Y

for all

 (t ) dt

. In sequel,

is denoted with

M 3 ( y, x)

M 2 ( y, x) .

Thus, the following theorem (Theorem 1[1]) is obtained: ( X , d1 )

,

(Y , d 2 )

be

T : X Y, S :Y  X  (t )dt



which holds always since the left hand is zero.

two

metric

spaces

d1p ( Sy , STx )

0

 (t ) dt 

M1 ( x , y )

0

 max M 1 ( x , y )

0

 (t ) dt 

p { d1

0

 (t ) dt

p ( x , Sy ), d1



F  min M1 ( x , y ) 

0

p ( x , STx ), d 2

( y , Tx )}

 (t )dt

 (t ) dt

and

ISSN: 2230-7818

@ 2010 http://www.ijaest.iserp.org. All rights Reserved.

and

are two mappings satisfying the

following inequalities:

where

Then

where

Let

 (t ) dt

where

We

The inequality (5) takes the form:

IJ A

0

Proof.

Z  Y , d3  d2 , z  y and

 (t )dt

z  c, y  Ta '

d2p ( b ,Ta ')

as the identity mapping of

The inequality (4) takes the form:

0

Taking

 d2p

R

Y  Ry  y, y  Y  , then we obtain Theorem 1[1].

 (t )dt

form which it follows d1p

the same with the metric space (Y , d 2 ) , (that is

ES

 (t )dt  

Corollary 2.2 If we consider in Theorem 2.1 the metric space and

( Z , d3 )

where M1 ( a ,Ta ')

SR and c is the unique fixed point of RT S.

T

a ' X

Let assume now that

Page 133


Manish Kumar Mishra et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES Vol No. 1, Issue No. 2, 130 - 134

d2p (Tx ,TSy )

0

 (t ) dt 

 max M 2 ( y , x )

0

 (t ) dt 

F  min M 2 ( y , x ) 

0

 (t )dt

where

M2 ( y, x)

0

 (t ) dt 

{ d 2 ( y , Tx ), d 2 ( y , TSy ), d1 ( x , Sy )} p

p

0

p

 (t ) dt

,

for all x  X , y  Y and 0    1 , then ST has a unique fixed point

a X

and T S has a unique fixed point

b Y

. Further,

T a = b and Sb = a. It is clear that in the Theorem 1[1] the functions

F1

and

replaced by F such that F  t   max F1  t  , F2  t  and

can be

c1 , c2

can be

c  max c1 , c2  .

Corollary 2.3 For

p  1 and F  t   0 for all t  R  , by Theorem

2.1 we obtain a theorem which extends the result of Theorem 1([2]) to three metric spaces. Corollary 2.4 For

p2

and F  t   0 for all

t  R  , we obtain a

REFERENCES [1]

S. C.Nesi, “A fixed point theorem in two metric spaces”, Bull. Math. Soc. Sci. Math. Roumanie (N.S.) 44, No. 92 (2001), 253257. B. Fisher, “Fixed points on two metric spaces”, Glasnik Mat.16, No. 36(1981), 333-337. V. Popa, Fixed points on two complete metric spaces, Zb. Rad. Prirod.-Mat. Fak. (N.S.) Ser. mat. 21, No. 1 (1991), 83-93. R. K. Jain, H. K. Sahu, B. Fisher,Related fixed point theorems for three metric spaces, Novi Sad J. Math. 26, No.1 (1996), 1117. Luljeta Kikina, “Fixed Points Theorems on Three Metric Spaces”, Int. Journal of Math. Analysis, Vol. 3, 2009, no. 13, 619 – 626. Deo Brat Ojha, Manish Kumar Mishra and Udayana Katoch,A Common Fixed Point Theorem Satisfying Integral Type for Occasionally Weakly Compatible Maps, Research Journal of Applied Sciences, Engineering and Technology 2(3): 239-244, 2010. Rhoades, B.E.,Two fixed point theorem for mapping satisfying a general contractiv condition of integral type. Int. J. Math. Sci., 3: 2003 4007-4013.

IJ A

[2]

ES

generalization of Theorem 2[3], extended to three metric spaces.

T

replaced by

F2

[3] [4]

[5]

[6]

[7]

ISSN: 2230-7818

@ 2010 http://www.ijaest.iserp.org. All rights Reserved.

Page 134


Turn static files into dynamic content formats.

Create a flipbook
Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.