- Research
- Open access
- Published:
Fixed point theorems for φ-contractions
Journal of Inequalities and Applications volume 2014, Article number: 266 (2014)
Abstract
This paper deals with the fixed point theorems for mappings satisfying a contractive condition involving a gauge function φ when the underlying set is endowed with a b-metric. Our results generalize/extend the main results of Proinov and thus we obtain as special cases some results of Mysovskih, Rheinboldt, Gel’man, and Huang. We also furnish an example to substantiate the validity of our results. Subsequently, an existence theorem for the solution of initial value problem has also been established.
MSC:47H10, 54H25.
1 Introduction and preliminaries
The Banach contraction principle has been extensively used to study the existence of solutions for the nonlinear Volterra integral equations and nonlinear integro-differential equations and to prove the convergence of algorithms in computational mathematics. These applications elicit the significance of fixed point theory. Therefore mathematicians have been propelled to contribute enormously in the field of fixed point theory by finding the fixed point(s) of self-mappings or nonself-mappings defined on several ambient spaces and satisfying a variety of conditions. Among these fixed point theorems only a few have practical importance, i.e., they provide a constructive method for finding fixed point(s). This provides information on the convergence rate along with error estimates. The Banach contraction principle is one of such theorems wherein the proposed iterative scheme converges linearly. Commonly, the iterative procedures serve as constructive methods in fixed point theory. Furthermore, it is also of crucial importance to have prior and posterior estimates for such methods. In this context, Proinov [1] extended the Banach contraction principle with a higher order of convergence. He proposed an iterative scheme for a mapping satisfying a contractive condition which involves a gauge function of order and obtained error estimates as well. His results include as special cases some results of Mysovskih [2], Rheinboldt [3], Gel’man [4], Huang [5], and others. In [6] the authors extended the results of Proinov to the case of multivalued mappings.
For the last few decades fixed point theory has rapidly been evolving, not only in metric structure but also in many different generalized spaces and the b-metric space is one of them. The notion of a b-metric space was initiated in some works of Bourbaki, Bakhtin, Czerwik, and Heinonen. Several papers appeared which deal with the fixed point theory for single valued and multivalued functions in a b-metric space [7–11]etc.
Inspired by the work of Proinov [1] in this paper we investigate whether the consequences of his results hold when the underlying structure is replaced with a b-metric space. We give an affirmative answer to this question. Our results generalize main results of Proinov [1] and thus subsume many results of authors [2–5]. We establish an example to substantiate the validity of our results. Consequently, in Section 3 we also obtain an existence theorem for the solution of an initial value problem.
In the following we recall some auxiliary notions and results in a b-metric space [7–9, 11] which are needed subsequently.
Let X be a nonempty set and be a given real number. A function is said to be a b-metric space if and only if for all the following conditions are satisfied:
(d1) if and only if ;
(d2) ;
(d3) .
The pair is called a b-metric space with the coefficient s.
The following example shows that the class of b-metric spaces is essentially larger than the class of metric spaces.
-
(1)
Let with where . Define as
where , . Then is a b-metric space with coefficient .
-
(2)
Let be the space of all real functions , such that . Define as
Then is a b-metric space with coefficient
A sequence in a b-metric space X is convergent if and only if there exists such that as and we write ; it is Cauchy if and only if as . A b-metric space is complete if every Cauchy sequence in X converges.
Let be a b-metric space; then a convergent sequence has a unique limit; every convergent sequence is Cauchy; and in general the b-metric d is not a continuous functional [13].
Definition 1.3 Let be a b-metric space and A be a nonempty subset of X then closure of A is the set consisting of all points of A and its limit points. Moreover, A is closed if and only if .
In the following the b-metric version of Cantor’s intersection theorem is given, which can easily be established running along the same lines as in the proof of its metric version.
Theorem 1.4 [14]
Let be a complete b-metric space, then every nested sequence of closed balls has a nonempty intersection.
Let and there exist some such that the set . The set is known as an orbit of . We recall that a function G from D into the set of real numbers is said to be f-orbitally lower semi-continuous at if and implies [15].
Throughout this paper let J always denote an interval in containing 0 i.e., an interval of the form , or ( is a trivial interval). Let denote a polynomial of the form and . Let denote the n th iterate of a function .
Definition 1.5 [1]
Let ; a function is said to be a gauge function of order r on J if it satisfies the following conditions:
-
(i)
for all and ,
-
(ii)
for all .
The condition (i) of Definition 1.5 elicits and is nondecreasing on . A gauge function is said to be a Bianchini-Grandolfi gauge function if for all [16].
Subsequently, in this paper let be a b-metric space (unless specified otherwise) with a coefficient . We assume that be an operator and there exist some such that . Let the operator f satisfy the following iterated contractive condition:
where φ is a gauge function of order on an interval J. We establish two convergence theorems for iterative processes of the type
where f satisfies (1.1).
2 b-Bianchini-Grandolfi gauge functions
In [1] Proinov proved his main results by assuming Bianchini-Grandolfi gauge functions and the mapping f satisfying the contractive condition (3.5) when the underlying space is endowed with a metric (see Corollary 3.9). But in the setting of b-metric space for some technical reasons we have to restrict ourselves to the gauge functions satisfying for all where s is the coefficient of b-metric space. Furthermore, taking into account such a crucial condition in order to calculate prior and posterior estimates we consider the gauge functions of the form
where is the coefficient of b-metric d and ϕ is nonnegative nondecreasing function on J such that
Remark 2.1 One can always define a nonnegative nondecreasing function ϕ on J satisfying (2.1) and (2.2) as follows:
where s is the coefficient of b-metric d.
For a fixed , let us consider the following simple examples of gauge functions of order r:
-
(i)
, is a gauge function of order 1 on ;
-
(ii)
(, ) is a gauge function of order r on where .
It is essential to mention here that to establish the fixed point theorem (see Theorem 3.7) we do not necessarily require the gauge functions φ satisfying (2.1), (2.2). But we consider the gauge function such that for all where s is a coefficient of b-metric space.
Lemma 2.2 Let φ be a gauge function of order on J. If ϕ is a nonnegative and nondecreasing function on J satisfying (2.1) and (2.2), then:
-
(1)
for all ,
-
(2)
for all and .
Remark 2.3 When d is a simple metric, then . In such case every gauge function satisfying is of the form where ϕ is nonnegative nondecreasing function on J (see [1]). Thus in such case the condition for all becomes superfluous and directly follows from Lemma 2.2.
The following lemma is fundamental to our main results.
Lemma 2.4 Let φ be a gauge function of order on J. If ϕ is a nonnegative and nondecreasing function on J satisfying (2.1) and (2.2), then for every we have:
-
(1)
for all ,
-
(2)
for all .
Proof (1) Set and let . Then from Lemma 2.2 we obtain . For the case is trivial. We shall prove (1) by using mathematical induction. For the property (1) is trivially satisfied as it reduces to an equality. Let it also hold for any integer , i.e.,
Since φ is nondecreasing on J, we obtain (as because and )
(2) By making use of Lemma 2.2 and monotonicity of ϕ, (1) leads to the following:
which completes the proof. □
Definition 2.5 Let be a fixed real number. A nondecreasing function is said to be a b-Bianchini-Grandolfi gauge function with a coefficient q on J if
We note that a b-Bianchini-Grandolfi gauge function also satisfies the following functional equation:
It is easy to see that every b-Bianchini-Grandolfi gauge function is also a Bianchini-Grandolfi [16] gauge function but the converse may not hold. A b-Bianchini-Grandolfi gauge function having coefficient is also a b-Bianchini-Grandolfi gauge function having coefficient for every .
From now on, we always assume that the coefficient of the b-Bianchini-Grandolfi gauge function is at least as large as the coefficient of the b-metric space.
Lemma 2.6 Every gauge function of order defined by (2.1) and (2.2) is a b-Bianchini-Grandolfi gauge function with coefficient .
Proof It is immediately follows from the first part of Lemma 2.4 and using the fact that for and . □
3 Fixed point theorems
For convenience we define a function by and assume that there exist some such that , so that the condition (1.1) can be put in the form
Lemma 3.1 Suppose is such that . Assume that ; then for all .
Proof Note that are well defined and belong to D. From (3.1) we have
Hence, . Similarly, iterating successively we get for all . □
Definition 3.2 Suppose is such that and . Then for every iterate , we define the closed ball with center at and radius , where is defined by (2.4).
Lemma 3.3 Suppose is such that and . Assume that for some ; then and .
Proof Since , Lemma 3.1 implies for all . The condition (2.5) implies for all . We have
Thus . Now let . As so that from (3.1) we have . By making use of (2.5) we get
Hence, . □
Definition 3.4 (Initial orbital point)
We say that a point is an initial orbital point of f if and .
The following lemma is obvious.
Lemma 3.5 For every initial orbital point of f and every we have
Furthermore, if φ is a gauge function of order defined by (2.1) and (2.2), then
where and ϕ is nonnegative nondecreasing on J satisfying (2.1) and (2.2).
Proof By making use of Lemma 3.1 we obtain . Since φ is nondecreasing, it easily follows that . Now from Lemma 2.4(1) we have
By using Lemma 2.4(2) we obtain
□
The following lemma gives bounds for inclusion radii and throughout its proof we will make use of the following facts:
where , and .
Lemma 3.6 Suppose is an initial orbital point of f and φ is a gauge function of order . Let ϕ be nonnegative and nondecreasing on J defined by (2.1) and (2.2). Then for radii , , the following estimates hold:
-
(1)
,
-
(2)
,
-
(3)
,
-
(4)
,
-
(5)
,
where .
-
(1)
Proof From definition of we have
(3.2) -
(2)
From (3.2) we have
-
(3)
By making use of first part of Lemma 3.5 above we have
-
(4)
Now by making use of Lemma 2.4 we have
-
(5)
From (4) we have
□
Now we proceed to formulate the following fixed point theorems.
Theorem 3.7 Let be an operator on a complete b-metric space such that the b-metric is continuous and f satisfies (1.1) with a b-Bianchini-Grandolfi gauge function φ of order on an interval J with coefficient . Then starting from an initial orbital point of f the iterative sequence (1.2) remains in and converges to a point ξ which belongs to each of the closed balls , , where , σ defined in (2.5) and is a coefficient of b-metric space. Furthermore, for each we have
If and the function on D is f-orbitally lower semi-continuous at ξ, then ξ is a fixed point of f.
Proof Since is an initial orbital point of f, from Lemma 3.3 we have
Thus for all . According to the definition of ρ and using Lemma 3.5 we have
Since φ is a b-Bianchini-Grandolfi gauge function, from (3.3) we obtain
which implies that is a nested sequence of closed balls. By Cantor’s theorem (for complete b-metric spaces), we deduce that there exists a unique point ξ such that for all and or equivalently, . From (d3) of Definition 1.1 we have
Letting and since the b-metric is continuous, we obtain
If and is f-lower semi-continuous at ξ, then
which infers . Furthermore, from Lemma 3.5 we obtain the following:
□
Remark 3.8 Theorem 3.7 gives a generalization of [[1], Theorem 4.1] and extends it to the case of b-metric spaces. It reduces to [[1], Theorem 4.1] when . Hence Theorem 3.7 not only extends the result of Proinov [1] but in turn it also includes results of Bianchini and Grandolfi [16] and Hicks [17] as special cases.
Corollary 3.9 [[1], Theorem 4.1]
Let be a complete metric space and be an operator satisfying
where φ is a Bianchini-Grandolfi gauge function on an interval J. Then starting from an initial orbital point of f the iterative sequence remains in and converges to a point ξ which belongs to each of the closed balls , where and . Moreover, if and f is continuous at ξ, then ξ is a fixed point of f.
Proof It follows from Lemma 3.1 that for . Thus from (3.5) we have
Thus Theorem 3.7 yields . Also since the iterative sequence and the mapping f is continuous at point t, we have . Thus
This implies f-orbital lower semi-continuity of at point ξ. Hence the conclusion follows from Theorem 3.7. □
Theorem 3.10 Let be an operator on a complete b-metric space such that the b-metric is continuous and let f satisfy (1.1) with a b-Bianchini-Grandolfi gauge function φ of order and a coefficient s on an interval J. Further, suppose that is an initial orbital point of f, then the following statements hold true.
-
(1)
The iterative sequence (1.2) remains in and converges with rate of convergence at least to a point ξ which belongs to each of the closed balls , , and
(3.7)
where ϕ is nonnegative and nondecreasing function on J satisfying (2.1) and (2.2).
-
(2)
For all the following prior estimate holds:
(3.8) -
(3)
For all the following posterior estimate holds:
(3.9) -
(4)
We have
(3.10)
for all .
-
(5)
If and the function on D is f-orbitally lower semi-continuous at ξ, then ξ is a fixed point of f.
Proof (1) From Theorem 3.7 it follows that the iterative sequence (1.2) converges to and further for all . Moreover, estimate (1) of Lemma 3.6 implies
-
(2)
For ,
where . Keeping n fixed and letting we get
Since
we have
Thus it implies
Hence from (3.11) we obtain
-
(3)
From (3.11) we have for ,
Setting , , and we have
Setting again and gives
From Lemma 2.4(2) we obtain
which implies
Thus from (3.13) and (3.15) we deduce for ,
-
(4)
We have
-
(5)
Its proof runs along the same lines as the proof of Theorem 3.7. □
Remark 3.11 For , Theorem 3.10 reduces to [[1], Theorem 4.2]. It also generalizes (taking and , ) results of Ortega and Rheinboldt [[18], Section 12.3.2], Kornstaedt [[19], Satz 4.1], Hicks and Rhoades [15], and Park [[20], Theorem 2]. The first two conclusions of Theorem 3.10 are due to Gel’man [[4], Theorem 3] (taking and , , ). It also yields some results of Hicks [[17], Theorem 3].
Corollary 3.12 Let be an operator on a complete b-metric space such that the b-metric is continuous. Further, assume that f satisfies
where φ is a b-Bianchini-Grandolfi gauge function of order on an interval J and with coefficient . Assume that is such that . Then the following statements hold.
-
(i)
The iterative sequence (1.2) converges to a fixed point ξ of f.
-
(ii)
The operator f has a unique fixed point in .
-
(iii)
The estimates (3.7)-(3.10) are valid.
Proof From (3.16) we have
which gives the continuity of f in b-metric space . Thus conclusions (i) and (iii) follow immediately from Theorem 3.10. Let be another fixed point of f in S; then . It follows from (3.16) that , which yields . □
Remark 3.13 For when the b-metric space under consideration is a simple metric space, the above corollary coincides with [[1], Corollary 4.4]. Thus the conclusions of Corollary 3.12 are consequences of the results of Matkowski [21].
4 Application and illustrative example
The following example illuminates the degree of generality of our result.
Example 4.1 Let . Define a function as
where is any positive integers. It is an easy exercise to see that d is a b-metric with coefficient . Define as
Setting on then φ is a b-Bianchini-Grandolfi gauge function with coefficient having order 2. Moreover, it is easily seen that all conditions of Theorem 3.7 are satisfied.
On the other hand, assume that , , are real numbers and the set is endowed with the Euclidean metric . For each gauge function φ defined on some interval one can find for some . In such a case, identifying , , , we assume f as defined above; then w.r.t. Euclidean metric we have
which contradicts the definition of φ. Hence, one cannot invoke the main results of Proinov [[1], Theorems 4.1, 4.2, Corollary 4.4].
Theorem 4.2 Consider the following initial value problem:
Assume that the following conditions hold:
-
(i)
f is continuous;
-
(ii)
f satisfies the condition
(4.2) -
(iii)
f is bounded on R, i.e.,
(4.3)
where , and .
Then the initial value problem (4.1) has a unique solution on the interval .
Proof Let be the space of all continuous real valued functions on I where with the usual supremum metric, i.e.,
Integrating (4.1) gives
Indeed finding the solution of initial value problem (4.1) is equivalent to finding the fixed point of the self-mapping defined by
where ; then X is a closed subspace of . We see that if , then and gives . Thus , and since f is continuous on R, the integral in (4.5) exists and T is defined for each . To see this, observe that T maps X to itself. We use (4.5) to write
Now by using (4.2) we have
Thus (4.6) implies
We take . Thus it suffices to take for , ; then φ is a gauge function of order . Also, for we have
Thus from (4.7) we obtain for all and . Further, for any it is easily seen that , which yields for . Therefore, all the conditions of Corollary 3.12 are satisfied. Hence the iterative sequence ; converges to the unique fixed point t of T at a rate of convergence . On the other hand, Picard’s iterations converge to the solution linearly. □
5 Conclusion
In Section 3 we have established two convergence theorems in the setting of a b-metric such that the self-mapping satisfies a contraction condition involving a gauge function of order . The gauge function φ has to satisfy the condition where is the coefficient of the underlying b-metric space. An example has been furnished to assess the degree of generality of our results. In Section 4 we established an existence theorem for the solution of an initial value problem, which not only gives the unique solution but also locates the domain for the solution.
References
Proinov PD: A generalization of the Banach contraction principle with high order of convergence of successive approximations. Nonlinear Anal. 2007, 67: 2361–2369. 10.1016/j.na.2006.09.008
Mysovskih IP: On the convergence of L.V. Kantorovic’s method of solution of functional equations and its applications. Dokl. Akad. Nauk SSSR 1950, 70: 565–568. (in Russian)
Rheinboldt WC: A unified convergence theory for a class of iterative processes. SIAM J. Numer. Anal. 1968, 5: 42–63. 10.1137/0705003
Gel’man A: A certain fixed-point principle. Dokl. Akad. Nauk SSSR 1950, 198: 506–508. (in Russian); Sov. Math., Dokl. 12, 813–816 (1971) (Engl. transl.)
Huang ZD: On Newton’s method under Holder continuous derivative. J. Math. Anal. Appl. 2002, 270: 332–339. 10.1016/S0022-247X(02)00060-4
Kiran Q, Kamran T: Nadler’s type principle with higher order of convergence. Nonlinear Anal. 2008, 69: 4106–4120. 10.1016/j.na.2007.10.041
Bakhtin IA: The contraction mapping principle in almost metric spaces. 30. In Functional Analysis. Ul’yanovsk Gos. Ped. Inst., Ul’yanovsk; 1989:26–37.
Heinonen J: Lectures on Analysis on Metric Spaces. Springer, Berlin; 2001.
Bourbaki N: Topologie générale. Hermann, Paris; 1974.
Boriceanu M, Bota M, Petruşel A: Multivalued fractals in b -metric spaces. Cent. Eur. J. Math. 2010,8(2):367–377. 10.2478/s11533-010-0009-4
Czerwik S: Nonlinear set-valued contraction mappings in b -metric spaces. Atti Semin. Mat. Fis. Univ. Modena 1998, 46: 263–276.
Berinde V: Generalized contractions in quasimetric spaces. Seminar on Fixed Point Theory 1993, 3–9.
Czerwik S: Functional Equations and Inequalities in Several Variables. World Scientific, Singapore; 2002.
Bota M, Molnar A, Varga C: On Ekeland’s variational principle in b -metric spaces. Fixed Point Theory 2011,12(1):21–28.
Hicks TL, Rhoades BE: A Banach type fixed point theorem. Math. Jpn. 1979, 24: 327–330.
Bianchini RM, Grandolfi M: Trasformazioni di tipo contrattivo generalizzato in uno spazio metrico. Atti Accad. Naz. Lincei, Rend. Cl. Sci. Fis. Mat. Nat. 1968, 45: 212–216.
Hicks TL: Another view of fixed point theory. Math. Jpn. 1990, 35: 231–234.
Ortega JM, Rheinboldt WC: Iterative Solutions of Nonlinear Equations in Several Variables. Academic Press, New York; 1970.
Kornstaedt HJ: Funktionalungleichungen und iterationsverfahren. Aequ. Math. 1975, 13: 21–45. 10.1007/BF01834116
Park S: A unified approach to fixed points of contractive maps. J. Korean Math. Soc. 1980, 16: 95–105.
Matkowski J: Integrable solutions of functional equations. Diss. Math. 1975, 127: 1–68.
Author information
Authors and Affiliations
Corresponding author
Additional information
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
All authors contributed equally to the writing of this paper. All authors read and approved the final manuscript.
Rights and permissions
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0), which permits use, duplication, adaptation, distribution, and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
About this article
Cite this article
Samreen, M., Kiran, Q. & Kamran, T. Fixed point theorems for φ-contractions. J Inequal Appl 2014, 266 (2014). https://doi.org/10.1186/1029-242X-2014-266
Received:
Accepted:
Published:
DOI: https://doi.org/10.1186/1029-242X-2014-266