A new type of difference sequence spaces Vakeel A. Khan Abstract. In this article we introduce a new sequence space denoted by m(4uv , φ, p). We give some topological properties and inclusion relations on this space. The results herein proved are analogous to those from [1]. M.S.C. 2000: 40C05, 46A45. Key words: Difference sequence; solid spaces; symmetric space. 1. Introduction Let `0 be the set of all complex sequences and l∞ , c and c0 be the sets of all bounded, convergent and null sequences x = (xk ) with complex terms, respectively, normed by ||x||∞ = sup |xk |, where k ∈ N = {1, 2, · · · }. k A sequence space X with linear topology is called a K-space if each of the maps Pk : X → C defined by Pk (x) = xk is continuous for k = 1, 2, · · · . A K- space X is called a BK-space provided X is a Banach space. The idea of difference sequence space was introduced by Kizmaz [12]. In 1981, Kizmaz [12] defined the sequence spaces: l∞ (4) = {x = {xk } ∈ `0 : (4xk ) ∈ l∞ }, c(4) = {x = {xk } ∈ `0 : (4xk ) ∈ c}, and c0 (4) = {x = {xk } ∈ `0 : (4xk ) ∈ c0 }, where 4x = (xk − xk+1 ). These are Banach spaces with the norm ||x||4 = |x1 | + ||4x||∞ . Et and Colak [5] generalized the above sequence spaces to the sequence spaces X(4r ) = {x = {xk } ∈ `0 : 4r xk ∈ X}, for X = l∞ , c and c0 , where r ∈ N, 40 x = (xk ), 4x = (xk − xk+1 ), 4r x = (4r xk − 4r xk+1 ) Applied Sciences, Vol.12, 2010, pp. 102-108. c Balkan Society of Geometers, Geometry Balkan Press 2010. ° A new type of difference sequence spaces and so that r 4 xk = r X i=0 103 · ¸ r (−1) xk+i . i i Difference sequence spaces have been studied by Colak and Et [2], Et [4], Et and Esi [6], Vakeel A. Khan [8,9,10,11] and many others. Let X, Y ⊂ `0 . Then we shall write \ x−1 ∗ Y = {a ∈ `0 : ax ∈ Y for all x ∈ X}. M (X, Y ) = x∈X The set X α = M (X, l1 ) is called Köthe - Toeplitz dual space or α - dual of X(see [16]). Let X be a sequence space. Then X is called (i) solid (or normal), if (αk xk ) ∈ X, whenever (xk ) ∈ X for all sequences of scalars (αk ) with |αk | ≤ 1 for all k ∈ N. (ii) symmetric, if (xk ) ∈ X implies (xπ(k) ) ∈ X, where π(k) is a permutation of N. (iii)perfect, if X = X αα . (iv) sequence algebra, if x.y ∈ X, whenever x, y ∈ X. It is well known that if X is perfect then X is normal (see [7]). Let C denote the space whose elements are finite sets of distinct positive integers. Given any element σ of C, we denote by c(σ) the sequence {cn (σ)} which is such that cn (σ) = 1 if n ∈ σ, cn (σ) = 0 otherwise. Further, ( ) ∞ X cn (σ) ≤ s (cf [13]), Cs = σ ∈ C : n=1 the set of those σ whose support has cardinality at most s, and ½ µ ¶ ¾ φk 0 Φ = φ = {φk } ∈ ` : φ1 > 0, ∆φk ≥ 0 and ∆ ≤ 0 (k = 1, 2, . . .) , k where ∆φk = φk − φk−1 ; and `0 is the set of all real sequences. For φ ∈ Φ, we define the following sequence space, introduce in [14], ! ) ( à 1 X 0 | xk | < ∞ . m(φ) = x = {xk } ∈ ` : sup sup φs s≥1 σ∈Cs k∈σ Recently the space m(φ) was extended to m(φ, p) by Tripathy and Sen [15] as follows: ( ) 1 X 0 p m(φ, p) := x = {xk } ∈ ` : sup sup | xk | < ∞ . s≥1 σ∈Cs φs k∈σ It is easy to see that: à k x km(φ,p) = sup sup s≥1 σ∈Cs 1 X | xk |p φs k∈σ ! p1 . 104 Vakeel A. Khan Remark 1. The space m(φ, p) is a Banach space with the norm à k x km(φ,p) = sup sup s≥1 σ∈Cs 1 X | xk |p φs ! p1 . k∈σ Remark 2. As in [14], we have (i) If φn = 1 for all n ∈ N then m(φ, p) = lp . Moreover lp ⊆ m(φ, p) ⊆ l∞ . (ii) If p = 1, then m(φ, p) = m(φ). Also m(φ) ⊆ m(φ, p). ³ ´ φs (iii) m(φ, p) ⊆ m(ψ, p) if and only if sups≥1 ψ < ∞. s 2. Main Results Let u be a fixed positive integer and v = (vk ) be any fixed sequence of non zero complex numbers (see [3]). Now we define the sequence space m(4uv , φ, p) as follows: ( ) 1 X 0 u u p m(4v , φ, p) := x = {xk } ∈ ` : sup sup (|4v xk | ) < ∞, 0≤p<∞ . s≥1 σ∈Cs φs k∈σ where 40v xk = (vk xk ), 4v xk = (vk xk − vk+1 xk+1 ), 4uv xk = (4u−1 xk − 4u−1 xk+1 ) v v such that 4uv xk · ¸ u X i u = (−1) vk+i xk+i . i i=0 It is clear that if u = 0, v = (1, 1, 1, · · · ) and p = 1, we have m(φ), which was defined by Sargent [13]. Theorem 2.1. The sequence space m(4uv , φ, p) is a Banach space for φ ∈ Φ normed by (2.1.1) ||x||41 = u X |xi | + sup sup i=1 s≥1 σ∈Cs 1 X 1/p ((|4uv xk |p )) , φs 1 ≤ p < ∞, k∈σ and a complete p-normed space by p-norm (2.1.2) ||x||42 = u X i=1 |xi |p + sup sup s≥1 σ∈Cs 1 X (|4uv xk |p ) , φs 0 < p < 1. k∈σ Proof. It is clear that m(4uv , φ, p) is a normed linear space normed by (2.1.1) for 1 ≤ p < ∞ and a p - normed space by p - norm (2.1.2) for 0 < p < 1. We need to A new type of difference sequence spaces 105 show that m(4uv , φ, p) is complete. Let {x(l) } be a Cauchy sequence in m(4uv , φ, p) where xl = (xlk )k = (xl1 , xl2 , · · · ) ∈ m(4uv , φ, p) for each l ∈ N. Then for given ² > 0 there exists n0 ∈ N such that l t ||x −x ||41 = u X |xli −xti |+sup sup s≥1 σ∈Cs i=1 ¢¢1/p 1 X ¡¡ u l |4v (xk − xtk )|p < ², for all l, t > n0 . φs k∈σ Now we obtain |xlk − xtk | → 0, as l, t → ∞, for each k ∈ N. Therefore (xlk )l = (x1k , x2k , · · · ) is a Cauchy sequence in C for each k. Since C is complete, it is convergent lim xlk = xk (say) for each k ∈ N. l Taking limit as t → ∞ in (2.1.3), we get u X |xli − xi | + sup sup s≥1 σ∈Cs i=1 ¢1/p 1 X¡ u l |4v (xk − xk )|p < ², φs for all n > n0 . k∈σ Hence (xlk − xi ) ∈ m(4uv , φ, p). We know that m(4uv , φ, p) is a linear space and (xlk ), (xlk − xk ) are in m(4uv , φ, p), it follows that (xk ) = (xlk ) − (xlk − xk ) ∈ m(4uv , φ, p). Hence m(4uv , φ, p) is complete. Similarly, we can show that m(4uv , φ, p) is complete space p-normed by (2.1.1) for 0 < p < 1. ¤ Theorem 2.2. For any φ ∈ Φ the space m(4uv , φ, p) is a K - space. The proof is straightforward. Theorem 2.3. m(4uv , φ) ⊂ m(4uv , φ, p), for any φ ∈ Φ. Proof. Let x ∈ m(4uv , φ). Then there exist a positive number K such that X |4uv xk |p ≤ Kφs , σ ∈ φs for each fixed s. k∈σ Hence X |4uv xk |p < Kφs , σ ∈ φs for each p ¿ 0 and σ ∈ φs . k∈σ Thus x ∈ m(4uv , φ, p). ¤ Theorem 2.4. For any two sequences (φs ) and (ψs ) of real numbers, we have m(4uv , φ, p) ⊂ m(4uv , ψ, p) if and only if µ ¶ φs sup < ∞. ψs s≥1 106 Vakeel A. Khan P Let x ∈ m(4uv , φ, p). Then sups≥1 supσ∈Cs φ1s k∈σ |4uv xk |p < ∞. ³ ´ φs Suppose that sups≥1 ψ < ∞. Then φs ≤ Kψs and so that ψ1s ≤ φKs for some s positive number K and for all s. Therefore we have Proof. 1 X u p K X u p |4v xk | ≤ |4v xk | ψs φs k∈σ for each s. k∈σ Now 1 X u p 1 X u p |4v xk | ≤ K sup sup |4v xk | . s≥1 σ∈Cs ψs s≥1 σ∈Cs φs k∈σ k∈σ P Hence sups≥1 supσ∈Cs ψ1s k∈σ |4uv xk |p < ∞. Therefore x ∈ m(4uv , ψ, p). ³ ´ φs Conversely, let m(4uv , φ, p) ⊆ m(4uv , ψ, p) and suppose that sup ψ = ∞. Then s s≥1 ³ ´ φ there exists aN increasing sequence (si ) of natural numbers such that lim ψssi = ∞. sup sup i Now for every B ∈ R+ (the set of positive real numbers), there exists i0 ∈ N such φ that ψssi > B for all si ≥ i0 . Hence ψ1s > ψBs and i i i B X u p 1 X u p |4v xk | > |4v xk | ψsi ψsi k∈σ k∈σ for all si ≥ i0 . Now taking supremum over si ≥ i0 and σ ∈ Cs we get (2.2.1) sup sup si ≥i0 σ∈Cs 1 X u p 1 X u p |4v xk | > B sup sup |4v xk | . ψsi si ≥i0 σ∈Cs φsi k∈σ k∈σ Since (2.2.1) holds for all B ∈ R+ (we may take B sufficiently large ) we have sup sup si ≥i0 σ∈Cs 1 X u p |4v xk | = ∞ ψsi k∈σ when x ∈ m(4uv , φ, p) with 0 < sup sup si ≥i0 σ∈Cs 1 X u p |4v xk | < ∞. φsi k∈σ u u u Therefore ³ ´ x 6∈ m(4v , ψ, p). This contradicts to m(4v , φ, p) ⊆ m(4v , ψ, p), whence φs ¤ sup ψs < ∞. s≥1 From Theorem 2.4, we get the following result. Corollary 2.1. m(4uv , φ, p) = m(4uv , ψ, p) if and only if µ ¶ µ ¶ φs φs 0 < inf ≤ sup < ∞. s≥1 ψs ψs s≥1 Theorem 2.5. m(4u−1 , φ, p) ⊂ m(4uv , φ, p) and the inclusion is strict. v A new type of difference sequence spaces 107 Proof. The proof follows from the following inequality and Minkowski’s inequality |4uv x| = |4u−1 xk − 4u−1 xk+1 | ≤ |4u−1 xk | + |4u−1 xk+1 |. v v v v To show that the inclusion is strict consider the following example. Example 2.1. Let φn = 1 for all n ∈ N, x = (k u−1 ) and v = (1, 1, 1, · · · ), then ). x ∈ lp (4uv ) \ lp (4u−1 v Theorem 2.6. The sequence space m(4uv , φ, p) is not sequence algebra, is not solid and is not symmetric, for u ≥ 1. Proof. For the proof of this theorem, consider the following examples: Example 2.2. Let x = (k u−1 ), y = (k u−1 ) and v = (1, 1, 1, · · · ). Then x, y ∈ m(4uv , φ, p), but x.y ∈ / m(4uv , φ, p). Hence m(4uv , φ, p) is not sequence algebra. Example 2.3. Let x = (k u−1 ), v = (1, 1, 1, · · · ) and αk = (−1)k . Then x = u−1 (k ) ∈ m(4uv , φ, p), but (αk xk ) ∈ / m(4uv , φ, p) for α = (αk ) = (−1)k . Hence m(4uv , φ, p) is not solid. Example 2.4. Let x = (k u−1 ) and v = (1, 1, 1, · · · ). Let (yk ) be a arrangement of (xk ) which is defined as follows : yk = {x1 , x2 , x4 , x3 , x9 , x5 , x16 , x6 , x25 , x7 , x36 , x8 , x49 , x10 , · · · }. Then y ∈ / m(4uv , φ, p). Hence m(4uv , φ, p) is not symmetric. The following result is a consequence of Theorem 2.6. Corollary 2.2. The sequence space m(4uv , φ, p) is not perfect. Theorem 2.7. lp (4uv ) ⊆ m(4uv , φ, p) ⊆ l∞ (4uv ). Proof. Since m(4uv , φ, p) = lp (4uv ) for φn = 1, for all nN, then lp (4uv ) ⊆ m(4uv , φ, p). Now suppose that x ∈ m(4uv , φ, p). Then we have sup sup s≥1 σ∈Cs 1 X u p |4v xk | < ∞, φs k∈σ and hence |4uv xk | < Kφ1 , for all n ∈ N and for some positive integer K. Thus x ∈ l∞ (4uv ). This completes the proof of Theorem. Corollary 2.3. If 0 < p < q, then m(4uv , φ, p) ⊆ m(4uv , φ, q). Proof. For the proof of this theorem follows from the following inequality à n ! q1 à n ! p1 X X q p |xk | ≤ |xk | , (0 < p < q). k=1 k=1 ¤ 108 Vakeel A. Khan References [1] R. Colak and M. Et, On some difference sequence sets and their topological properties, Bull. Malaysian Math.Sci.Soc. 2 (2005), 125-130. [2] R. Colak and M. Et, On some generalized difference sequence spaces and related matrix transformations, Hokkoido Mathematical Journal 26, 3 (1997), 483-492. [3] A. Esi and M.Isik, Some generalized difference sequence spaces, Thai Journal of Mathematics, 3 (2005), 241-247. [4] M. Et, On some topological properties of generalized difference sequence spaces, Internat. J. Math. and Math. Sci. 24, 11 (2000), 785-791. [5] M. Et and R. Colak, On some generalized difference sequence spaces, Soochow J. of Math. 21 (1995), 377-386. [6] M. Et and A. Esi, On Köthe Toeplitz duals of generalized difference sequence spaces, Bull. Malaysian Math. Sci. Soc. 24, 11 (2000), 25-32. [7] P. Kampthan and M. Gupta, Sequence Spaces and Series, Marcel Dekker Inc. 1980. [8] V. A. Khan, Some inclusion relations between the difference sequence spaces defined by sequence of moduli, Indian Mathematical Society 73 (2006), 77-81. [9] V. A. Khan and Q.M.D. Lohani, Difference sequence spaces defined by a sequence of moduli, Southeast Asian Bull. Math. 30 (2006), 1061-1067. [10] V.A. Khan and Q.M.D. Lohani, New lacunary strong convergence difference sequence spaces defined by sequence of moduli, Kyungpook Mathematical Journal 46 (2006), 591-595. [11] V.A. Khan, New lacunary strongly summable difference sequences and 4m v - lacunary almost statistical convergence, Vietnam Journal of Mathematics 36 (2008), 405-413. [12] H. Kizmaz, On certain sequence spaces, Canad. Math. Bull. 24, 2 (1981), 169176. [13] M. Mursaleen, Some geometric properties of a sequence space related to lp , Bull. Aust. Math. Soc. 67, 2 (2003), 343-347. [14] W.L.C. Sargent, Some sequence spaces related to the `p spaces, J. London Math. Soc. 35 (1960), 161-171. [15] B.C. Tripathy and M. Sen, On a new class of sequences related to the space lp , Tamkang J. Math. 33, 2 (2002), 167-171. [16] A. Wilansky, Summability Through Functional Analysis, North Holland Mathematics Studies 85, North Holland, 1984. Author’s address: Vakeel A. Khan Department of Mathematics, A.M.U. Aligarh, India. E-mail: vakhan@math.com