You are on page 1of 6

# Mathematical Communications 10(2005), 105-110

105

A common unique ﬁxed point result in metric spaces involving generalised altering distances
Binayak S. Choudhury∗

Abstract. In this paper we work out a unique common ﬁxed point result for two self-mappings deﬁned on a complete metric space. These mappings are assumed to satisfy a contractive inequality which involves two generalised altering distances. Key words: generalised altering distances, metric space, ﬁxed point AMS subject classiﬁcations: 54H25 Received December 18, 2003 Accepted May 17, 2004

1.

Introduction

Fixed point theory in spaces has a vast literature. In particular, there has been a number of works on ﬁxed points involving Altering Distance Functions. These are control functions which alter the distance between two points in a metric space. Such functions were introduced by M. S. Khan et al. in [4] and used in the same paper for deﬁning and solving a new category of ﬁxed points problems in metric spaces. Deﬁnition 1 [see [4]]. An altering distance function is a function ψ : [0, ∞) → [0, ∞) which is (i) monotone increasing and continuous and (ii) ψ (t) = 0 if and only if t = 0. Afterwards a number of works has appeared in which altering distances has been used. In references [5], [6] and [7], for example, ﬁxed points of single valued mappings and in [1] ﬁxed points of multi-mappings have been obtained by using altering distance functions. Altering distances have been generalised to a twovariable function and in [3] a generalisation to a three-variable function has been introduced and applied for obtaining ﬁxed point results in metric spaces. In this paper we propose a generalisation of altering distances to a three-variable function and with the help of such function we derive a unique common ﬁxed point result for two self-mappings in a complete metric space. We note that speciﬁc ﬁxed
∗ Department of Mathematics, Bengal Engineering and Science University, P.O. - B. Garden, Shibpur, 711 103 Howrah, West Bengal, India, e-mail: binayak12@yahoo.co.in

106 B. y ). T y )) − ψ2 (d(x. ∞] is said to be a generalised altering distance function if (i) ψ (x. x. Proof. we get φ1 (d(Sx2n . z ) = 0 only if x = y = z = 0. (3) (2) . for k > 0. d(x2n . x2n+1 ).mappings such that the following inequality is satisﬁed: φ1 (d(Sx. y. d(x2n+1 . (ii) ψ is monotone increasing in all the three variables and (iii) ψ (x. Putting x = x2n and y = x2n+1 in (1). x2n+1 ). d(y. a2n+1 ) − ψ2 (a2n . d(x2n+1 . Choudhury point results follow by speciﬁc choices of the functional form of ψ . c) = k max{a. . r ≥ 1. T x2n+1 )) = φ1 (d(x2n+1 . T y ) ≥ ψ1 (d(x. c) = (a + α bq )r + cs . φ(x) = 0 if and only if x = 0. Sx). Sx). . ∞)3 → [0. d(x2n+1 . d(x2n+1 . . 2. for all n = 0. a2n . Let(X. . d(x. We deﬁne φ (x) = ψ (x.d) be a complete metric space and S and T two self . Examples of ψ are ψ (a. x) for x ∈ [0. d(x. . d(x2n . r. Then S and T have a common ﬁxed point. x2n+1 ). . q. x. b. ψ (a. Fixed point theorem Theorem 1. for all n = 0. 2. Clearly. S. a2n+1 ). s ≥ 1 and α > 0 Other examples may also be constructed. y. c}. Deﬁnition 2. T x2n+1 )) −ψ2 (d(x2n . 2. d(x2n . b. ψ (a. Let x0 ∈ X be an arbitrary point. Sx2n ). We propose the following deﬁnition. Sx2n ). T x2n+1 )) = ψ1 (d(x2n . c) = ap + bq + cr . a2n . 1. d(y. b. ∞). x2n+1 ). b. d(x2n . x2n+2 )) or by(2). q. x2n+1 ). A function ψ : [0. x2n+2 ) −ψ2 (d(x2n . T y )) (1) where ψ1 and ψ2 are generalised altering distance functions and φ1 (x) = ψ1 (x. x2n+1 = Sx2n and x2n+2 = T x2n+1 let an = d(xn . . xn+1 ). φ1 (a2n+1 ) ≤ ψ1 (a2n . p. 1. x2n+2 )) ≤ ψ1 (d(x2n . . For n = 0. y ). x2n+1 ). x). . 2. 1. z ) is continuous. where p.

so that a2n+1 ≤ a2n . . 2 . If {x2r }r =1 is not a Cauchy sequence. . a2n ) By an identical argument we obtain a2n+2 ≤ a2n+1 . we obtain for all n = 0. 1. = φ(a2n+1 ) 107 (4) This is due to the fact that ψ1 is monotone increasing in all variables and ψ2 (a2n . Since φ is continuous. x2n(k) ) >∈ and d(x2m(k) . from (9) we obtain φ2 (a) = 0 which implies that a = 0. (10) We next prove that {xn } is a Cauchy sequence. x) φ2 (an+1 ) ≤ φ1 (a0 ) < ∞ n=0 which implies φ2 (an ) → 0 as n → ∞. a2n+1 ) = 0 whenever a2n+1 = 0. or equivalently φ2 (an+1 ) ≤ φ1 (an ) − φ1 (an+1 ). (9) Again from (8) {an } is convergent and let an → a (say) as n → ∞. that is an = d(xn+1 . . xn ) → 0 as n → ∞. (8) Then from (3) and (6). a2n−1 . . x2n(k)−1 ) <∈ (11) . 2. a2n ) − ψ2 (a2n−1 . n = 0. n = 0. for all n = 0. . . x. . a2n . a2n+1 . Summing up in (8) we obtain ∞ where φ2 (x) = ψ2 (x. Thus we arrive at a contradiction. . . n = 0. a2n+1 ).A common unique fixed point result in metric spaces If a2n+1 > a2n . . we obtain φ1 (an+1 ) ≤ φ1 (an ) − φ2 (an+1 ). 2 . 1. 1. 1. d(x2m(k) . 2. then given ∈> 0 we can ﬁnd monotone increasing sequences of natural numbers {2m(k )} and {2n(k )} such that n(k ) > m(k ). . then φ1 (a2n+1 ) < ψ1 (a2n+1 . . (5) Putting x = x2n and y = x2n−1 in (1) we obtain φ1 (a2n ) ≤ ψ1 (a2n−1 . 2. In view of (10) it is suﬃcient to ∞ prove that {x2r }∞ r =1 ⊂ {xn }is a Cauchy sequence. an+1 ≤ an . (7) (6) From (5) and (7). a2n−1 . . 1.

x2n(k)+1 ). x2m(k) ) ≤∈ and ∈≤ lim d(x2n(k)+1 . x2m(k)−1 ) ≤ d(x2n(k) . 0. . Making k → ∞ in the above two inequalities and using (10) and (12) we obtain k→∞ lim d(x2n(k) . . by using f (10) and (12) we have k→∞ (13) (14) lim d(x2n(k)+1 . d(x2n(k) . x2n(k) ) ≤ d(x2m(k) . x2m(k) )) −ψ2 (d(x2n(k) . x2m(k) )) Making k → ∞ in the above inequality and taking into account the continuity of ψ1 and ψ2 . z ) = 0 if and only if x = y = z = 0. x2m(k)−1 ). x2m(k) ). x2m(k) ) ≤ ψ1 (d(x2n(k) . x2n(k)+1 ) + d(x2n(k)+1 . Also for all k = 1. x2n(k)−1 ) + d(x2n(k)−1 . x2m(k) ) + d(x2m(k) . x2m(k) ). x2m(k) . . for all k = 1. x2m(k)−1 ) + d(x2m(k)−1 . (15) and (16) we have φ1 (∈) ≤ ψ1 (∈. . 0) < φ1 (∈) This is due to the fact that ψ1 is monotone increasing in its variables and by property of ψ2 that ψ (x. 0) − ψ2 (∈. y. . we obtain φ1 d(x2n(k)+1 . Making k → ∞ in (13) and (14) respectively. 0. x2n(k) ) Making k → ∞ in the above inequality by virtue of (10) we obtain k→∞ lim d(x2m(k) . . (16) Putting x = x2n(k) and y = x2m(k)−1 in (1). x2n(k) ) =∈ (12) For all k = 1. d(x2n(k) . . (15) For all k = 1. x2m(k) ) =∈ . Choudhury Then by (11) ∈ < d(x2m(k) . d(x2m(k)−1 . x2m(k) ) ≤ d(x2n(k)+1 . . x2m(k) ). . 2. . d(x2n(k)+1 . . d(x2n(k) . 2. 2. d(x2n(k) . x2n(k) ) < ∈ +d(x2n(k)−1 . d(x2m(k)−1 . by virtue of (10).108 B. x2n(k)+1 ). x2m(k)−1 ) =∈ . x2m(k) ) ≤ d(x2n(k) . x2m(k)−1 ) d(x2n(k) . . x2m(k)−1 ). x2m(k) ) k→∞ that is k→∞ lim d(x2n(k)+1 . x2n(k) ) + d(x2n(k) . 2. . x2m(k) ) ≤ d(x2n(k) . S.

y. z2 ). for all n = 1. z2 )). A comprehensive survey of various types of contractive mappings and related ﬁxed point theorems may be obtained in [8]. y. If d(z. ✷ Let z1 and z2 be two common ﬁxed points of S and T and z1 = z2 . y )]s + k2 [d(x. we obtain φ1 (d(x2n+1 . In an exactly similar way we prove z = Sz. T z )) − ψ2 (0. Here.d). Then d(z1 . In particular. (19) or z = T z. z2 ). for example. This establishes the uniqueness property of the ﬁxed point. (17) Putting x = x2n and y = z in (1). T z ) = 0.A common unique fixed point result in metric spaces 109 The above inequality gives a contradiction so that ∈= 0. Let xn → z as n → ∞. T z ) = 0. (18) Equations (18) and (19) show that z is a common ﬁxed point of S and T . T z )). z is non-zero. z2 )) ≤ ψ1 (d(z1 . z2 ) = 0. . z ). Let S. d) is a complete metric space satisfying [d(Sx. d(z. d(x2n . 2. by using (10) and (17) and continuity of ψ1 and ψ2 we obtain φ1 (d(z. z ) = 0 if and only if x = y = z = 0. 0. This establishes the fact {x2n }∞ n=1 is a Cauchy sequence and hence in view of (10) {xn } is also a Cauchy sequence and hence convergent in (X. T y )]s (20) . we obtain d(z. This is again due to the fact that ψ1 is monotonic increasing in all its variables and ψ (x. d(z. T z )) −ψ2 (d(x2n . T x)]s + k3 [d(y. d(z. y. A number of ﬁxed point results may be obtained by assuming diﬀerent forms for the functions ψ1 and ψ2 . T : X → X where (X. 0. we obtain φ1 (d(z. T z )) ≤ ψ1 (0. 0. ﬁxed point results under various contractive conditions are obtainable from the above theorems. T z )) ≤ ψ1 (d(x2n . d(x2n . T z )) < φ1 (d(z. x2n+1 ). From (1) we obtain φ1 (d(z1 . then using the property that ψ1 and ψ2 are monotone increasing and ψ2 (x. . 0. z ) ≤ 0 if at least one of x. Hence. T z )). Making n → ∞ in the above inequality. x2n+1 ). T y ]s ≤ k1 [d(x. T z )) which is a contradiction. Corollary 1. Contractive mappings are important in ﬁxed point theory. . d(z. we derive the following corollary of our theorem. 0) − ψ2 (d(z1 . z ). 0) < φ1 (d(z1 . The above inequality is a contradiction which shows that z1 = z2 .

S. Choudhury where 0 < k1 + k2 + k3 < 1 and s > 0. Some new principles in ﬁxed point theory. V. M. A. References [1] B. Then S and T have a common ﬁxed point. c) = (1 − k )[k1 as + k2 bs + k3 cs ] with k = k1 + k2 + k3 .S.S. Upadhyay. Generalization of common ﬁxed point theorems for weakly commuting maps by altering distance. R.Soc. V.48. G. Filomat 14(2000). Choudhury. Joponica 35(1990). [6] K. Babu. ✷ Other ﬁxed point results may also be obtained under speciﬁc choices of ψ1 and ψ2 . R.Tmkang J. preprint. G. Choudhury. S.647. R. G. N. Naidu. R. 645 . Fixed point theorems in metric spaces by altering distances between the points. 30(1984). Bull. Babu. P. 243 . 90(1998). V.Math. Bulletin of Pure and Applied Science 19 E(2000). R. S.Soc. 641 . Taskovic. . Proof. Sastry. Then (20) is implied by (1). A.Calcutta Math.182. Sastry. Babu.pure Appl. 1-9. S. Dutta.Math. Khan.250. 31(2000). P. [2] B.110 B. R. 529-533. Fixed point theorem by altering distances between the points. Indian J. Math. G. 43 . Sessa. Sastry. [3] B. Math. [7] K. Swaleh. Choudhury. b. P. R. Bull. 175 . S. V. N. The corollary then follows by an applying of Theorem 1. Dutta. c) = k1 as + k2 bs + k3 cs ψ2 (a. P. [4] M. A uniﬁed ﬁxed point result in metric spaces involving a two variable functuion. Fixed point theorems in metric spaces by altering distances. 30(1999). A uniﬁed approach to ﬁxed points of selfmappings in metric spaces. We make particular choices of ψ1 and ψ2 as the follows: ψ1 (a. P. Naidu. [8] M. b. [5] K.Austral. R. As a ﬁnal remark we observe that there is no continuity assumption on the functions S and T .666.