ebook img

Hyers-Ulam-Rassias Stability of n-Apollonius Type Additive - Hikari PDF

18 Pages·2011·0.15 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Hyers-Ulam-Rassias Stability of n-Apollonius Type Additive - Hikari

Int. Journal of Math. Analysis, Vol. 6, 2012, no. 3, 111 - 128 Hyers-Ulam-Rassias Stability of n-Apollonius Type ∗ Additive Mapping and Isomorphisms in C -Algebras Fridoun Moradlou Department of Mathematics Sahand University of Technology, Tabriz, Iran [email protected] Choonkil Park Department of Mathematics, Hanyang University Seoul, 133–791, Republic of Korea [email protected] Jung Rye Lee Department of Mathematics, Daejin University Kyeonggi 487-711, Republic of Korea [email protected] Abstract. In this paper, we prove Hyers-Ulam-Rassias stability of the following functional equation in Banach modules over a unital C∗-algebra: (cid:3) (cid:4) (cid:2)n (cid:2) (cid:2)n 1 1 f(z −x )+ f(x +x ) = nf z − x , i n i j n2 i i=1 1≤i<j≤n i=1 which n is fixed integer and n ≥ 2. As an application, we show that every almost linear bijection h : A → B of a unital C∗-algebra A onto a unital C∗- (cid:5)(cid:5) (cid:6) (cid:6) (cid:5)(cid:5) (cid:6) (cid:6) algebraB is a C∗-algebra isomorphism whenh n2−1 duy = h n2−1 du h(y) n2 n2 for all unitaries u ∈ A, all y ∈ A, and all d ∈ Z. 112 F. Moradlou, C. Park and J. Lee Mathematics Subject Classification: Primary 39B72, 47H10, 46L05, 46B03, 47Jxx Keywords: Hyers-Ulam-Rassias stability, n-Apollonius type additive map- ping, C∗-algebra homomorphism, generalized derivation 1. Introduction and preliminaries The stability problem of functional equations originated from a question of Ulam [36] concerning the stability of group homomorphisms: Let (G ,∗) be a 1 group and let (G ,(cid:6),d) be a metric group with the metric d(·,·). Given (cid:2) > 0, 2 does there exist a δ((cid:2)) > 0 such that if a mapping h : G → G satisfies the 1 2 inequality d(h(x∗y),h(x)(cid:6)h(y)) < δ for all x,y ∈ G , then there is a homomorphism H : G → G with 1 1 2 d(h(x),H(x)) < (cid:2) for all x ∈ G ? If the answer is affirmative, we would say that the equation 1 of homomorphism H(x∗y) = H(x)(cid:6)H(y) is stable. The concept of stability for a functional equation arises when we replace the functional equation by an inequality which acts as a perturbation of the equation. Thus the stability question of functional equations is that how do the solutions of the inequality differ from those of the given functional equation? Hyers [9] gave a first affirmative answer to the question of Ulam for Banach spaces. Let X and Y be Banach spaces. Assume that f : X → Y satisfies (cid:7)f(x+y)−f(x)−f(y)(cid:7) ≤ ε for all x,y ∈ X and some ε ≥ 0. Then there exists a unique additive mapping T : X → Y such that (cid:7)f(x)−T(x)(cid:7) ≤ ε for all x ∈ X. Th.M.Rassias[30]providedageneralizationofHyers’Theoremwhichallows the Cauchy difference to be unbounded. Theorem 1.1. (Th.M.Rassias). Let f : E → E(cid:4) be a mapping from a normed vector space E into a Banach space E(cid:4) subject to the inequality (1.1) (cid:7)f(x+y)−f(x)−f(y)(cid:7) ≤ (cid:2)((cid:7)x(cid:7)p +(cid:7)y(cid:7)p) Stability of the n-Apollonius type additive mapping 113 for all x,y ∈ E, where (cid:2) and p are constants with (cid:2) > 0 and p < 1. Then the limit f(2nx) L(x) = lim n→∞ 2n exists for all x ∈ E and L : E → E(cid:4) is the unique additive mapping which satisfies 2(cid:2) (cid:7)f(x)−L(x)(cid:7) ≤ (cid:7)x(cid:7)p 2−2p for all x ∈ E. Also, if for each x ∈ E the mapping f(tx) is continuous in t ∈ R, then L is R-linear. Theabove inequality (1.1)hasprovided a lotof influence inthedevelopment ofwhatisnowknownasaHyers-Ulam-Rassias stabilityoffunctionalequations. Beginning around the year 1980 the topic of approximate homomorphisms, or the stability of the equation of homomorphism, was studied by a number of mathematicians. G˘avruta [8] generalized the Rassias’ result. The stability problems of several functional equations have been extensively investigated by a number of authors and there are many interesting results concerning this problem (see [2], [5], [7], [13]–[24], [31]–[33]). J.M. Rassias [28] following the spirit of the innovative approach of Th.M. Rassias [30] for the unbounded Cauchy difference proved a similar stability theorem in which he replaced the factor (cid:7)x(cid:7)p+(cid:7)y(cid:7)p by (cid:7)x(cid:7)p·(cid:7)y(cid:7)q for p,q ∈ R with p+q (cid:9)= 1 (see also [29] for a number of other new results). Theorem 1.2. [27, 28, 29] Let X be a real normed linear space and Y a real complete normed linear space. Assume that f : X → Y is an approximately additive mapping for which there exist constants θ ≥ 0 and p ∈ R −{1} such that f satisfies inequality (cid:7)f(x+y)−f(x)−f(y)(cid:7) ≤ θ ·||x||p2 ·||y||2p for all x,y ∈ X. Then there exists a unique additive mapping L : X → Y satisfying θ (cid:7)f(x)−L(x)(cid:7) ≤ ||x||p |2p −2| for all x ∈ X. If, in addition, f : X → Y is a mapping such that the transformation t → f(tx) is continuous in t ∈ R for each fixed x ∈ X, then L is an R-linear mapping. 114 F. Moradlou, C. Park and J. Lee The following functional equation (1.2) Q(x+y)+Q(x−y) = 2Q(x)+2Q(y), is called a quadratic functional equation, and every solution of equation (1.2) is said to be a quadratic mapping. F. Skof [35] proved the Hyers-Ulam-Rassias stability of the quadratic functional equation (1.2) for mappings f : E → E , 1 2 where E is a normed space and E is a Banach space. In [6], S. Czerwik 1 2 proved the Hyers-Ulam-Rassias stability of the quadratic functional equation. C. Borelli and G.L. Forti [4] generalized the stability result as follows: let G be an abelian group, E a Banach space. Assume that a mapping f : G → E satisfies the functional inequality (cid:7)f(x+y)+f(x−y)−2f(x)−2f(y)(cid:7) ≤ ϕ(x,y) for all x,y ∈ G, and ϕ : G×G → [0,∞) is a function such that (cid:2)∞ 1 φ(x,y) := ϕ(2ix,2iy) < ∞ 4i+1 i=0 for all x,y ∈ G. Then there exists a unique quadratic mapping Q : G → E with the properties (cid:7)f(x)−Q(x)(cid:7) ≤ φ(x,x) for all x ∈ G. Jun and Lee [11] proved the Hyers-Ulam-Rassias stability of the Pexiderized quadratic equation f(x+y)+g(x−y) = 2h(x)+2k(y) for mappings f,g,h and k. In an inner product space, the equality (cid:7) (cid:7) 1 (cid:7) x+y(cid:7)2 (1.3) (cid:7)z −x(cid:7)2 +(cid:7)z −y(cid:7)2 = (cid:7)x−y(cid:7)2 +2(cid:7)z − (cid:7) 2 2 holds, andiscalledtheApollonius’ identity. Thefollowingfunctionalequation, which was motivated by this equation, (cid:8) (cid:9) 1 x+y (1.4) Q(z −x)+Q(z −y) = Q(x−y)+2Q z − , 2 2 is quadratic (see [25]). For this reason, the functional equation (1.4) is called a quadratic functional equation of Apollonius type, and each solution of the functional equation (1.4) is said to be a quadratic mapping of Apollonius type. The quadratic functional equation and several other functional equations are useful to characterize inner product spaces [1]. Stability of the n-Apollonius type additive mapping 115 In [25], C. Park and Th.M. Rassias introduced and investigated a functional equation, which is called the generalized Apollonius type quadratic functional equation. Recently in [26], C. Park and Th.M. Rassias introduced and inves- tigated the following functional equation (cid:8) (cid:9) 1 x+y (1.5) f(z −x)+f(z −y) = − f(x+y)+2f z − 2 4 whichiscalledApollonius type additive functional equation, andwhosesolution is called a Apollonius type additive mapping. In [23], C. Park introduced and investigated a functional equation, which is called the generalized Apollonius- Jensen type additive functional equation and whose solution of the functional equation is said to be a generalized Apollonius-Jensen type additive mapping. In this paper, for a fixed integer n ≥ 2, we introduce the new functional equation, which is called the additive functional equation of n-Apollonius type and whose solution is called an additive mapping of n-Apollonius type, (cid:3) (cid:4) (cid:2)n (cid:2) (cid:2)n 1 1 (1.6) f(z −x )+ f(x +x ) = nf z − x . i n i j n2 i i=1 1≤i<j≤n i=1 This paper is organized as follows: In Section 2, we investigate the Hyers- Ulam-Rassias stability of additive functional equation of n-Apollonius type in Banach modules over C∗-algebras. In Section 3, we investigate C∗-algebra isomorphisims in unital C∗-algebras associated with additive functional equation of n-Apollonius type. 2. Hyers-Ulam-Rassias Stability of n-Apollonius type additive mapping in Banach modules over a C∗-algebra Throughout this section, assume that A is a unital C∗-algebra with norm | · | and unitary group U(A), and that X and Y are Banach modules over a unital C∗-algebra A with norms (cid:7)·(cid:7) and (cid:7)·(cid:7) , respectively. X Y Lemma 2.1. A function f : X → Y satisfies (1.6) for all z,x ,··· ,x if and 1 n only if the function f is additive. Proof. Letting x = ··· = x = z = 0 in (1.6), we get that f(0) = 0. Let j and 1 n k be fixed integers with 1 ≤ j < k ≤ n. Setting x = 0 for all 1 ≤ i ≤ n,i (cid:9)= j,k i 116 F. Moradlou, C. Park and J. Lee in (1.6), we have (2.1) 1 n−2(cid:5) (cid:6) f(z −x )+f(z −x )+(n−2)f(z) =− f(x +x )− f(x )+f(x ) j k j k j k n n (cid:5) (cid:6) 1 +nf z − (x +x ) n2 j k for all x ,x ,z ∈ X. Replacing x by −x and x by x in (2.1), respectively, j k j j k j we get n−2(cid:5) (cid:6) (2.2) f(z +x )+f(z −x ) = − f(−x )+f(x ) +2f(z) j j j j n for all x ,z ∈ X. Putting z = 0 in (2.2), we conclude that f(−x ) = −f(x ) j j j for all x ∈ X. This means that f is an odd function. Letting x = z = 0 in j k (2.1) and using the oddness of f, we obtain that 1 1 2 2 (2.3) f( x ) = f(x ), f(n x ) = n f(x ) n2 j n2 j j j for all x ∈ X. Letting z = 0 in (2.1), using the oddness of f and (2.3) we have j f(x +x ) = f(x )+f(x ) j k j j for all x ,x ∈ X. Therefore, f : X → Y is a additive mapping. j k The converse is obviously true. For a given mapping f : X → Y and a given u ∈ U(A) and a fixed integer n ≥ 2, we define (cid:2)n (cid:2) 1 D f(z,x ,··· ,x ) := f(uz −ux )+ f(ux +μux ) u 1 n i i j n i=1 1≤i<j≤n (cid:3) (cid:4) (cid:2)n 1 −nuf z − x n2 i i=1 for all z,x ,··· ,x ∈ X. 1 n We investigate the Hyers-Ulam-Rassias stability of a type additive mapping of n-Apollonius type in Banach modules over a C∗-algebra. Stability of the n-Apollonius type additive mapping 117 Theorem 2.2. Let f : X → Y be a mapping satisfying f(0) = 0 for which there exists a function ϕ : An+1 → [0,∞) such that (2.4) (cid:3) (cid:2)∞ (cid:8) n2 (cid:9)j (cid:8)n2 −1(cid:9)j (cid:8)n2 −1(cid:9)j ϕ(cid:10)(z,x) = ϕ z,0,··· ,0, x,··· ,0) < ∞, i n2 −1 n2 (cid:11) n(cid:12)2(cid:13) (cid:14) j=0 i th (2.5) (cid:3) (cid:4) (cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) n2 j n2 −1 j n2 −1 j n2 −1 j lim ϕ z, x ,··· , x = 0, j→∞ n2 −1 n2 n2 1 n2 n (2.6) (cid:7)D f(z,x ,··· ,x )(cid:7) ≤ ϕ(z,x ,··· ,x ) μ 1 n Y 1 n for all u ∈ U(A) and all x,z,x ,··· ,x ∈ X. Then there exists a unique 1 n A-Linear additive mapping L : X → Y of n-Apollonius type such that n (2.7) (cid:7)f(x)−L(x)(cid:7) ≤ ϕ(cid:10)(x,x) Y n2 −1 i for all x ∈ X. Proof. Let u = 1 ∈ U(A). Letting z = x = x and for each 1 ≤ j ≤ n with i j (cid:9)= i, x = 0 in (2.6), we get j (cid:7) (cid:7) (cid:7)n2 −1 (cid:5)n2 −1 (cid:6)(cid:7) (2.8) (cid:7) f(x)−nf x (cid:7) ≤ ϕ(x,0,··· ,0,(cid:11)(cid:12)x(cid:13)(cid:14),0,··· ,0) n n2 Y ith for all x ∈ X. For convenience, set ϕi(z,x) = ϕ(z,0,··· ,0,(cid:11)(cid:12)x(cid:13)(cid:14),0,··· ,0) ith n2 −1 for all x,z ∈ X and all 1 ≤ i ≤ n. Letting α = , we get n (cid:7) (cid:7) (cid:5) (cid:6) (cid:7) α (cid:7) (2.9) (cid:7)αf(x)−nf x (cid:7) ≤ ϕ (x,x) i n Y for all x ∈ X. It follows from (2.9) that (cid:7) (cid:7) (cid:5) (cid:6) (cid:7) n α (cid:7) 1 (2.10) (cid:7)f(x)− f x (cid:7) ≤ ϕ (x,x) i α n Y α 118 F. Moradlou, C. Park and J. Lee (cid:8) (cid:9) α k for all x ∈ X. If we replace x in (2.10) by x and multiply both sides of (cid:8) (cid:9) n n k (2.10) by , then we have α (cid:7)(cid:8) (cid:9) (cid:3)(cid:8) (cid:9) (cid:4) (cid:8) (cid:9) (cid:3)(cid:8) (cid:9) (cid:4)(cid:7) (cid:7) n k α k n k+1 α k+1 (cid:7) (cid:7) f x − f x (cid:7) α n α n Y (2.11) (cid:8) (cid:9) (cid:3)(cid:8) (cid:9) (cid:8) (cid:9) (cid:4) 1 n k α k α k ≤ ϕ x, x i α α n n for all x ∈ X and all non-negative integers k. So we obtain that for all non- negative integers m,l with l > m (cid:3) (cid:4) (cid:3) (cid:4) (cid:7)(cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) (cid:7) (cid:7) n m α m n l+1 α l+1 (cid:7) (cid:7) f x − f x (cid:7) α n α n Y (cid:3) (cid:4) (cid:3) (cid:4) (cid:2)l (cid:7)(cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) (cid:7) (cid:7) n i α i n i+1 α i+1 (cid:7) ≤ (cid:7) f x − f x (cid:7) α n α n Y i=m (cid:3) (cid:4) (cid:2)l (cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) 1 n i α i α i (2.12) ≤ ϕ x, x i α α n n i=m (cid:15) (cid:8) (cid:9) n d for all x ∈ X. So it follows from (2.4) and (2.12) that the sequence (cid:16) α (cid:8) (cid:9) (cid:5) (cid:6) α d f x is Cauchy for all x ∈ X, and thus converges by the completeness n of Y. Thus we can define a mapping L : X → Y by (cid:8) (cid:9) (cid:8) (cid:9) (cid:5) (cid:6) n d α d L(x) = lim f x d→∞ α n for all x ∈ X. Letting m = 0 in (2.12), we obtain (cid:7) (cid:8) (cid:9) (cid:3)(cid:8) (cid:9) (cid:4)(cid:7) (cid:7) n l+1 α l+1 (cid:7) (cid:7)f(x)− f x (cid:7) α n Y (2.13) (cid:2)l (cid:8) (cid:9) (cid:3)(cid:8) (cid:9) (cid:8) (cid:9) (cid:4) 1 n j α j α j ≤ ϕ x, x i α α n n j=0 for all x ∈ X and all l ∈ N. Taking the limit as l → ∞ in (2.13), we obtain the inequality (2.7). It follows from (2.4) and (2.6) that (cid:7) (cid:3) (cid:4)(cid:7) (cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) (cid:7) (cid:7) (cid:7)D1L(z,x1,··· ,xn)(cid:7)Y = lim n d(cid:7)(cid:7)D1f α dz, α dx1,··· , α dxn (cid:7)(cid:7) d→∞ α n n n (cid:3) (cid:4) (cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) n d α d α d α d ≤ lim ϕ z, x ,··· , x = 0. 1 n d→∞ α n n n Stability of the n-Apollonius type additive mapping 119 Therefore, the mapping L : X → Y satisfies the equation (1.6) and hence L is a generalized additive mapping of n-Apollonius type. To prove the uniqueness, let L(cid:4) be another generalized additive mapping of n-Apollonius type satisfying (2.7). Then we have, for any positive integer k (cid:15)(cid:7) (cid:3) (cid:4) (cid:3) (cid:4)(cid:7) (cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) (cid:7) (cid:7) (cid:7)L(x)−L(cid:4)(x)(cid:7) ≤ n k (cid:7)(cid:7)L α kx −f α kx (cid:7)(cid:7) α n n (cid:7) (cid:3) (cid:4) (cid:3) (cid:4)(cid:7)(cid:16) (cid:8) (cid:9) (cid:8) (cid:9) (cid:7) (cid:7) +(cid:7)(cid:7)f α kx −L(cid:4) α kx (cid:7)(cid:7) n n (cid:3) (cid:4) (cid:2)∞ (cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) 2n n k+j α k+j α k+j ≤ ϕ x, x , n2 −1 α i n n j=0 which tends to zero as k → ∞. So we conclude that L(x) = L(cid:4)(x) for all x ∈ X. By the assumption, for each u ∈ U(A), we get (cid:7) (cid:3) (cid:4)(cid:7) (cid:8) (cid:9) (cid:7) (cid:8) (cid:9) (cid:7) n d(cid:7) α d (cid:7) (cid:7)DuL(x,0(cid:11),·(cid:12)·(cid:13)· ,0(cid:14))(cid:7) = lim (cid:7)Duf x,0(cid:11),·(cid:12)·(cid:13)· ,0(cid:14) (cid:7) d→∞ α n ntimes ntimes (cid:8) (cid:9) (cid:8) (cid:9) n d α d ≤ lim ϕ( x,0,··· ,0) = 0 (cid:11) (cid:12)(cid:13) (cid:14) d→∞ α n ntimes for all x ∈ X. Hence nL(ux) = nuL(x) for all u ∈ U(A) and all x ∈ X. So L(ux) = uL(x) for all u ∈ U(A) and all x ∈ X. By the same reasoning as in the proofs of [19] and [22], L(ax+by) = L(ax)+L(by) = aL(x)+bL(y) for all a,b ∈ A(a,b (cid:9)= 0) and all x,y ∈ X. And L(0x) = 0 = 0L(x) for all x ∈ X. SotheuniquegeneralizedadditivemappingL : X → Y ofn-Apollonius type is an A-linear mapping. Corollary 2.3. Let (cid:2) ≥ 0 and let p be a real number with p > 1. Assume that a mapping f : X → Y satisfies the inequality (cid:5) (cid:2)n (cid:6) (2.14) (cid:7)D f(z,x ,··· ,x )(cid:7) ≤ (cid:2) (cid:7)z(cid:7)p + (cid:7)x (cid:7)p u 1 n Y X j X j=1 120 F. Moradlou, C. Park and J. Lee for all u ∈ U(A) and all z,x ,··· ,x ∈ X. Then there exists a unique A-linear 1 n additive mapping L : X → Y of n-Apollonius type such that 2n(n2 −1)1−p(cid:2) (cid:7)f(x)−L(x)(cid:7) ≤ ||x||p Y (n2 −1)(2−p) −(n2 −1)n2(1−p) X for all x ∈ X. Proof. It follows from (2.14) that f(0) = 0. Define (cid:5) (cid:2)n (cid:6) ϕ(z,x ,··· ,x ) := (cid:2) (cid:7)z(cid:7)p + ||x ||p , 1 n X j X j=1 and apply Theorem 2.2 to get the result. Theorem 2.4. Let f : X → Y be a mapping satisfying f(0) = 0 for which there exists a function ϕ : Xn+1 → [0,∞) satisfying (2.6) such that (2.15) (cid:3) (cid:4) (cid:2)∞ (cid:8)n2 −1(cid:9)j (cid:8) n2 (cid:9)j (cid:8) n2 (cid:9)j ϕ(cid:10)(z,x) = ϕ z,0,··· ,0, x,0,··· ,0 < ∞ i n2 n2 −1 n2 −1 (cid:11) (cid:12)(cid:13) (cid:14) j=1 i th (2.16) (cid:3) (cid:4) (cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) n2 −1 j n2 j n2 j n2 j lim ϕ z, x ,··· , x = 0 j→∞ n2 n2 −1 n2 −1 1 n2 −1 n for all x,z,x ,··· ,x ∈ X. Then there exists a unique A-linear additive map- 1 n ping L : X → Y of n-Apollonius type such that n (2.17) (cid:7)f(x)−L(x)(cid:7) ≤ ϕ(cid:10)(x,x) Y n2 −1 i for all x ∈ X. Proof. Let u = 1 ∈ U(A). It follows from (2.9) that (cid:7) (cid:7) (cid:8) (cid:9) (cid:5) (cid:6) (cid:7) α n (cid:7) 1 n n (2.18) (cid:7)f(x)− f x (cid:7) ≤ ϕ x, x i n α Y n α α (cid:8) (cid:9) n2 −1 n k for all x ∈ X, where α = . If we replace x in (2.18) by x and n (cid:8) (cid:9) α α k multiply both sides of (2.18) by , then we have n (cid:3) (cid:4) (cid:3) (cid:4) (cid:7)(cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) (cid:8) (cid:9) (cid:7) (cid:7) α k n k α k+1 n k+1 (cid:7) (cid:7) f x − f x (cid:7) n α n α Y (2.19) (cid:8) (cid:9) (cid:3)(cid:8) (cid:9) (cid:8) (cid:9) (cid:4) 1 α k n k+1 n k+1 ≤ ϕ x, x i n n α α

Description:
functional equation (1.4) is said to be a quadratic mapping of Apollonius type. equation, which is called the generalized Apollonius type quadratic functional.
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.