EUROPEAN JOURNAL OF PURE AND APPLIED MATHEMATICS 2025, Vol. 18, Issue 4, Article Number 6526 ISSN 1307-5543 – ejpam.com Published by New York Business Global Bi-Metric Structures and Their Applications in Bitopological Contexts Abdullah Alsoboh1, Jamal Oudetallah2, Ala Amourah3,∗, Raja’a Al-Naimi4, Mohammed Al Hatmi1,∗, Wasim Audeh2, Ahmad Almalkawi5, Tala Sasa6 1 College of Applied and Health Sciences, A’Sharqiyah University, Post Box No. 42, Post Code No. 400, Ibra, Sultanate of Oman 2 Department of Mathematics, University of Petra, Amman, 11196, Jordan 3 Mathematics Education Program, Faculty of Education and Arts, Sohar University, Sohar 311, Oman 4 Emirates Aviation University, Dubai, United Arab Emirates 5 Modern College of Business and Science, Muscat, Sultanate of Oman 6 Department of Mathematics, Faculty of Science, Applied Science Private University, Amman, Jordan Abstract. We introduce a novel mathematical framework for analyzing bitopological spaces through bi-metric structures. Our research establishes the theoretical underpinnings of coupled metric spaces - configurations that inherently embrace bitopological structures while expanding conven- tional metric-based frameworks. We demonstrate key mathematical correspondences linking these bi-metric constructs to their generated topologies and furnish diverse contextual implementations. Our investigation examines completeness properties, stability characteristics, and develops sys- tematic product structures within these frameworks. Furthermore, we identify significant rela- tionships with functional-analytical principles, particularly regarding bi-normed spaces and quasi- metric frameworks. The mathematical architecture we propose offers innovative perspectives on the interrelationships between metric frameworks and bitopological domains with implications for functional transformation theories, including practical applications in computer networks, image processing, and economic modeling. 2020 Mathematics Subject Classifications: 54E35, 54E55, 54E40 Key Words and Phrases: Bitopological spaces, bi-metric structures, transformation theory, stability principles, functional analysis ∗Corresponding author. ∗Corresponding author. DOI: https://doi.org/10.29020/nybg.ejpam.v18i4.6526 Email addresses: abdullah.alsoboh@asu.edu.om (A. Alsoboh), jamal.oudetallah@uop.edu.jo (J. Oudetallah), AAmourah@su.edu.om (A. Amourah), Rajaa.alnaimi@eau.ac.ae (R. Al-Naimi), mohammed.alhatmi@asu.edu.om (M. A. Al Hatmi), waudeh@uop.edu.jo (W. Audeh), ahmad.abdelqader@mcbs.edu.om (A. Almalkawi), t_sasa@asu.edu.jo (T. Sasa) https://www.ejpam.com 1 Copyright: © 2025 The Author(s). (CC BY-NC 4.0) A. Alsoboh et al. / Eur. J. Pure Appl. Math, 18 (4) (2025), 6526 2 of 21 1. Conceptual Framework and Historical Context The notion of spaces characterized by two distinct topological structures was intro- duced by Kelly [1], who termed these constructs bitopological spaces. These mathemati- cal entities have demonstrated substantial utility across varied analytical and topological domains. Concurrently, metric spaces remain foundational in mathematical analysis. The convergence of these areas presents unique investigative opportunities which we explore comprehensively in this work. Recent developments in bitopological theory have expanded significantly since Kelly’s foundational work. Y. Y. Yousif and L. A. Hussain [2] investigated fibrewise IJ-perfect bitopological spaces, establishing new characterizations and properties. GarcÃa-Máynez and Pimienta [3] explored symmetry properties in bitopological spaces, while Chen and Li [4] developed applications in fuzzy topology. The intersection with computer science has been particularly fruitful, with Smyth [5] demonstrating applications in domain theory and denotational semantics. Furthermore, recent work by Kumar and Singh [6] has con- nected bitopological structures to rough set theory, and Martinez et al. [7] have explored applications in data analysis and machine learning. While previous scholarly investigations have examined relationships between metric characteristics and bitopological spaces [8], we identify a substantial theoretical gap: the absence of a comprehensive mathematical architecture specifically addressing metric sys- tems purposefully designed for bitopological environments. We address this deficiency by proposing an innovative conceptualization of metric assessment that naturally accommo- dates multiple topological configurations. Our methodology diverges fundamentally from existing approaches. Rather than an- alyzing independent metric functions applied across identical spaces, we develop an in- tegrated structural framework termed a bi-metric system that inherently captures the multi-dimensional nature of bitopological spaces. This formulation extends and general- izes classical metric theory established by Banach [9]. The practical significance of our theoretical framework extends to numerous real-world applications. In computer networks, routing algorithms often need to optimize for both physical distance and transmission delay. In image processing, quality assessment requires both pixel-wise accuracy and perceptual similarity metrics. Economic models frequently involve multi-criteria optimization where different metrics capture distinct aspects of sys- tem performance. This paper presents a comprehensive framework for bi-metric structures in bitopo- logical contexts. We establish fundamental definitions in Section 2, providing essential mathematical foundations including bitopological spaces, metric spaces, and quasi-metrics. Section 3 introduces bi-metric systems with their structural properties, illustrating these abstract concepts through concrete examples to facilitate comprehension. Section 4 ex- amines the connections between bi-metric systems and functional analysis, particularly exploring relationships with bi-normed spaces and developing frameworks that encompass primal-dual configurations. In Section 5, we investigate advanced theoretical properties including contraction mapping principles and completeness characteristics adapted specif- A. Alsoboh et al. / Eur. J. Pure Appl. Math, 18 (4) (2025), 6526 3 of 21 ically for bi-metric environments. Section 6 expands our analysis to structural properties and applications, addressing density concepts, product structures, characterization theo- rems for bitopological spaces, and relationships with quasi-metrics, while presenting exper- imental results that validate our theoretical findings. The concluding section summarizes contributions. In future research, an intriguing direction lies in exploring the interplay between bi- metric structures and complex analysis, particularly within bitopological frameworks. Bi- metric spaces endowed with two compatible metrics offer a natural setting for extending classical notions of convergence, continuity, and analyticity to more generalized environ- ments. When combined with tools from complex analysis, such as conformal mappings and analytic function theory, these structures could yield new geometric interpretations of bi-univalent and multi-univalent functions [10–15], as well as novel characterizations of analytic mappings between dual topological systems. Moreover, the synthesis of bi- metric geometry with complex analytic methods may provide a foundation for modeling dual phase systems, complex dynamical behaviors, and operator-theoretic generalizations in functional spaces, thus opening new pathways for both pure mathematical theory and applied geometric function research. 2. Fundamental Definitions We establish our theoretical foundation with several core definitions aligned with es- tablished topological literature [16], recent advances [17]. Definition 1. A bitopological space consists of a triple (X, T1, T2) where X represents a non-empty set and T1, T2 denote distinct topological structures on X. Definition 2. A metric space comprises a pair (X,ϕ) where X represents a non-empty set and ϕ : X ×X → R+ functions as a mapping satisfying: (i) ϕ(x, y) ≥ 0 for all x, y ∈ X, with equality if and only if x = y (ii) ϕ(x, y) = ϕ(y, x) for all x, y ∈ X (iii) ϕ(x, z) ≤ ϕ(x, y) + ϕ(y, z) for all x, y, z ∈ X Definition 3. We classify a function Q : X ×X → [0,∞) as a quasi-metric on X when: (i) Q(x, y) = 0 if and only if x = y (ii) Q(x, z) ≤ Q(x, y) +Q(y, z) for all x, y, z ∈ X We observe that quasi-metrics typically lack symmetrical properties. Definition 4. A pseudo-metric on a set X is a function p : X ×X → [0,∞) satisfying all metric axioms except that p(x, y) = 0 does not necessarily imply x = y. This concept becomes relevant when considering quotient structures in bi-metric systems. A. Alsoboh et al. / Eur. J. Pure Appl. Math, 18 (4) (2025), 6526 4 of 21 3. Bi-Metric Systems: Definition and Structural Properties We introduce bi-metric systems, a generalized mathematical framework that naturally accommodates bitopological arrangements. This structured approach provides a unified methodology for analyzing spaces with distinct yet interrelated metric measures. Definition 5. A bi-metric system consists of a triple (X,Ψ,⊕) where X represents a non-empty set, and Ψ : X ×X → R+ × R+ functions as a mapping satisfying: (i) Ψ(x, y) = (ϕ1(x, y), ϕ2(x, y)) where ϕi : X ×X → R+ for i ∈ {1, 2} serve as metric functions (ii) Ψ(x, y) = (0, 0) if and only if x = y (iii) Ψ(x, y) = Ψ(y, x) for all x, y ∈ X (iv) ⊕ represents a binary operation ⊕ : (R+ ×R+)× (R+ ×R+) → R+ ×R+ satisfying: (a) Ψ(x, z) ≤comp ⊕(Ψ(x, y),Ψ(y, z)) for all x, y, z ∈ X (b) ⊕ demonstrates monotonicity for both arguments relative to partial ordering ≤comp (c) ⊕((0, 0),(0, 0)) = (0, 0) where ≤comp indicates component-wise partial ordering on R+ × R+. Remark 1. We interpret function Ψ as a bidimensional metric assessment. Operation ⊕ functions as a generalized triangular coordination mechanism that maintains funda- mental metric properties while accommodating the bi-dimensional nature of the structure, extending modular space frameworks developed by Nakano [18]. For enhanced comprehension, we provide illustrative implementations: Example 1. Consider a non-empty set X with two distinct metric functions ϕ1, ϕ2 on X. When we define Ψ(x, y) = (ϕ1(x, y), ϕ2(x, y)) and select ⊕((a1, a2), (b1, b2)) = (a1 + b1, a2 + b2), the resulting structure (X,Ψ,⊕) constitutes a bi-metric system. Example 2 (Network Routing Application). In computer networks, consider nodes X where routing decisions depend on both physical distance and transmission delay. Define: • ϕ1(x, y) = physical cable distance between nodes x and y (in kilometers) • ϕ2(x, y) = average transmission delay between nodes x and y (in milliseconds) With Ψ(x, y) = (ϕ1(x, y), ϕ2(x, y)) and ⊕ as component-wise addition, this bi-metric sys- tem enables routing algorithms to optimize for both distance and latency simultaneously. A. Alsoboh et al. / Eur. J. Pure Appl. Math, 18 (4) (2025), 6526 5 of 21 Example 3. For X = R+ , define Ψ(x, y) = (|x−y|, | tanh(x)−tanh(y)|) with ⊕((a1, a2), (b1, b2)) = (a1 + b1, a2 + b2). We obtain a bi-metric system where both components satisfy traditional triangular coordination principles. Example 4 (Weight-parameterized metric system). Given a metric space (X,ϕ) and a weight function w : X → (0,∞), establishing Ψ(x, y) = (ϕ(x, y), |w(x) − w(y)|) and ⊕((a1, a2), (b1, b2)) = (a1 + b1, a2 + b2) constructs a bi-metric system incorporating both baseline measurements and weight variations. Bi-metric systems naturally generate dual topologies on underlying sets, similar to quasi-gauge spaces previously described in mathematical literature [19]. Theorem 1. Given a bi-metric system (X,Ψ,⊕) with Ψ(x, y) = (ϕ1(x, y), ϕ2(x, y)), for each i ∈ {1, 2}, the topology Ti induced by the corresponding component can be defined as: Ti = {U ⊂ X : ∀x ∈ U, ∃ε > 0 where Vi(x, ε) ⊂ U} in which Vi(x, ε) = {y ∈ X : ϕi(x, y) < ε}. The resulting structure (X, T1, T2) constitutes a bitopological space. Proof. We must establish that T1 and T2 represent legitimate topologies on X. For each i ∈ {1, 2}: (i) ∅ ∈ Ti through vacuous truth. (ii) X ∈ Ti because for any x ∈ X, Vi(x, ε) ⊂ X for all ε > 0. (iii) For arbitrary collection {Uα}α∈A ⊂ Ti, consider any x ∈ ⋃ α∈A Uα. There exists α0 ∈ A where x ∈ Uα0 . Since Uα0 ∈ Ti, we identify ε > 0 satisfying Vi(x, ε) ⊂ Uα0 ⊂ ⋃ α∈A Uα. This establishes ⋃ α∈A Uα ∈ Ti. (iv) For finite collection U1, U2, . . . , Un ∈ Ti, consider any x ∈ n⋂ j=1 Uj . We determine ε1, ε2, . . . , εn > 0 where Vi(x, εj) ⊂ Uj for each j. Selecting ε = min{ε1, ε2, . . . , εn}, we have Vi(x, ε) ⊂ Vi(x, εj) ⊂ Uj for each j, yielding Vi(x, ε) ⊂ n⋂ j=1 Uj . Therefore, n⋂ j=1 Uj ∈ Ti. Consequently, T1 and T2 constitute legitimate topologies on X, establishing (X, T1, T2) as a bitopological space. A. Alsoboh et al. / Eur. J. Pure Appl. Math, 18 (4) (2025), 6526 6 of 21 Example 5 (Image Quality Assessment). In digital image processing, quality assessment often requires multiple metrics. Consider: • ϕ1(I1, I2) = Mean Squared Error (MSE) between images I1 and I2 • ϕ2(I1, I2) = 1 - SSIM (Structural Similarity Index) The bi-metric system (X,Ψ,⊕) where X is the space of images, allows simultaneous optimization for both pixel-wise accuracy and perceptual quality. Theorem 2. For bi-metric systems (X,ΨX ,⊕X) and (Y,ΨY ,⊕Y ), function f : X → Y demonstrates continuity with respect to the i-th induced topologies precisely when for every x ∈ X and ε > 0, there exists δ > 0 where ϕX,i(x, z) < δ implies ϕY,i(f(x), f(z)) < ε. Proof. (⇒) Assuming f demonstrates continuity with respect to the i-th induced topologies, consider arbitrary x ∈ X and ε > 0. The set VY,i(f(x), ε) = {y ∈ Y : ϕY,i(f(x), y) < ε} forms an open set in the i-th topology of Y . By continuity properties, f−1(VY,i(f(x), ε)) is open in the i-th topology of X. Since x ∈ f−1(VY,i(f(x), ε)), we identify δ > 0 with VX,i(x, δ) ⊂ f−1(VY,i(f(x), ε)). Thus, whenever ϕX,i(x, z) < δ, z ∈ VX,i(x, δ), yielding f(z) ∈ VY,i(f(x), ε) equivalently, ϕY,i(f(x), f(z)) < ε. (⇐) Assuming the stated condition, consider arbitrary open set V in the i-th topology of Y . We must establish that f−1(V ) is open in the i-th topology of X. For any x ∈ f−1(V ), we have f(x) ∈ V . Since V is open, there exists ε > 0 where VY,i(f(x), ε) ⊂ V . By our assumption, we identify δ > 0 where ϕX,i(x, z) < δ implies ϕY,i(f(x), f(z)) < ε. This yields VX,i(x, δ) ⊂ f−1(VY,i(f(x), ε)) ⊂ f−1(V ). Therefore, f−1(V ) is open in the i-th topology of X, confirming f ’s continuity characteristics. Example 6 (Continuity in bi-metric systems). Taking X = Y = R with standard topology, we define bi-metric systems: ΨX(x1, x2) = (|x1 − x2|, |(x1)− (x2)|) ΨY (y1, y2) = (|y1 − y2|, | sinh(y1)− sinh(y2)|) with both ⊕X and ⊕Y implementing component-wise addition. Our continuity approach extends and generalizes quasi-uniformization techniques pre- viously documented in mathematical research [19]. Definition 6. We characterize a sequence {xn} in bi-metric system (X,Ψ,⊕) as bi-Cauchy when for every ε > 0, there exists N ∈ N where for all m,n ≥ N : Ψ(xm, xn) 0, we identify N1 ∈ N where for all m,n ≥ N1, ϕ1(xm, xn) < ε. Setting yn = xn for all n ∈ N, we observe for all m,n ≥ N1, Ψ(ym, yn) = (ϕ1(xm, xn), ϕ2(xm, xn)) 0 (potentially sequence- dependent). For the second component, we identify N2 ∈ N where for all m,n ≥ N2, ϕ2(xm, xn) < ε. Taking N = max{N1, N2}, for all m,n ≥ N , Ψ(ym, yn) n: Ψ(xn, xm) ≤comp ⊕(Ψ(xn, xn+1),⊕(Ψ(xn+1, xn+2), . . . ,Ψ(xm−1, xm))) = (ϕ1(xn, xn+1) + . . .+ ϕ1(xm−1, xm), ϕ2(xn, xn+1) + . . .+ ϕ2(xm−1, xm)) For first component: ϕ1(xn, xn+1) = ϕ1(T (xn−1), T (xn)) ≤ λ1 · ϕ1(xn−1, xn) A. Alsoboh et al. / Eur. J. Pure Appl. Math, 18 (4) (2025), 6526 12 of 21 ≤ λn1 · ϕ1(x0, x1) Similarly, for second component: ϕ2(xn, xn+1) ≤ λn2 · ϕ2(x0, x1) Therefore: ϕ1(xn, xm) ≤ m−1∑ i=n ϕ1(xi, xi+1) ≤ m−1∑ i=n λi1 · ϕ1(x0, x1) = ϕ1(x0, x1) · λn1 · 1− λm−n 1 1− λ1 Similarly: ϕ2(xn, xm) ≤ ϕ2(x0, x1) · λn2 · 1− λm−n 2 1− λ2 As n → ∞, both components approach zero since λ1, λ2 ∈ [0, 1), confirming {xn} as bi-Cauchy. By bi-completeness of the bi-metric system, {xn} converges to some x∗ ∈ X. To establish x∗ as T ’s equilibrium point: Ψ(T (x∗), x∗) ≤comp ⊕(Ψ(T (x∗), T (xn)),Ψ(T (xn), x ∗)) = (ϕ1(T (x ∗), T (xn)) + ϕ1(T (xn), x ∗), ϕ2(T (x ∗), T (xn)) + ϕ2(T (xn), x ∗)) ≤comp (λ1 · ϕ1(x∗, xn) + ϕ1(xn+1, x ∗), λ2 · ϕ2(x∗, xn) + ϕ2(xn+1, x ∗)) As n→ ∞, both components approach zero, yielding Ψ(T (x∗), x∗) = (0, 0), confirming T (x∗) = x∗. For uniqueness, assuming alternative equilibrium point y∗ ̸= x∗: Ψ(x∗, y∗) = Ψ(T (x∗), T (y∗)) ≤comp (λ1 · ϕ1(x∗, y∗), λ2 · ϕ2(x∗, y∗)) Since λ1, λ2 < 1, this creates contradiction unless Ψ(x∗, y∗) = (0, 0), which means x∗ = y∗. Example 12 (Expanded Bi-metric Contraction Application). Consider X = [0, 1] with bi- metric system Ψ(x, y) = (|x−y|, | sinh(x)− sinh(y)|) and ⊕ implementing component-wise addition. Define T : X → X by T (x) = x 2 + 1 4 . Step 1: Verify contraction conditions for first component ϕ1(T (x), T (y)) = ∣∣∣∣x2 + 1 4 − y 2 − 1 4 ∣∣∣∣ = 1 2 |x− y| = 1 2 ϕ1(x, y) A. Alsoboh et al. / Eur. J. Pure Appl. Math, 18 (4) (2025), 6526 13 of 21 Thus λ1 = 1 2 . Step 2: Verify contraction conditions for second component By the Mean Value Theo- rem, for some c ∈ [0, 1]: | sinh(T (x))− sinh(T (y))| = cosh(c)|T (x)− T (y)| = cosh(c) 2 |x− y| Since cosh(c) ≤ cosh(1) for c ∈ [0, 1]: ϕ2(T (x), T (y)) ≤ cosh(1) 2 ϕ2(x, y) where λ2 = cosh(1) 2 ≈ 0.77 < 1. Step 3: Iterative computation starting from x0 = 0 x0 = 0 x1 = T (0) = 1 4 x2 = T ( 1 4 ) = 1/4 2 + 1 4 = 3 8 x3 = T ( 3 8 ) = 3/8 2 + 1 4 = 7 16 ... x∗ = 1 2 (fixed point) Step 4: Verification T ( 1 2 ) = 1/2 2 + 1 4 = 1 2 ✓ This example demonstrates how bi-metric contraction provides tighter convergence bounds when λ1 ̸= λ2, offering advantages over classical single-metric approaches. Example 13 (Function space with dual metrics). Consider function space X = C[0, 1], continuous functions on [0, 1], with: Ψ(f, g) = ( max x∈[0,1] |f(x)− g(x)|, ∫ 1 0 |f(x)− g(x)|2dx ) and ⊕((a1, a2), (b1, b2)) = (a1+b1, a2+b2). This forms a bi-metric system with components representing uniform and L2 metrics, respectively. This bi-metric system induces distinct topologies: uniform convergence topology and L2 convergence topology. A sequence might converge in one topology but not the other. For example, sequence fn(x) = xn converges to zero function in L2 topology but not in uniform topology. A. Alsoboh et al. / Eur. J. Pure Appl. Math, 18 (4) (2025), 6526 14 of 21 Example 14 (Machine Learning Application). In neural network training, consider the space X of network parameters with: • ϕ1(θ1, θ2) = ∥θ1 − θ2∥2 (parameter distance) • ϕ2(θ1, θ2) = |L(θ1)− L(θ2)| (loss function difference) The bi-metric system enables tracking both parameter convergence and loss minimization simultaneously, providing better insights into training dynamics. Our bi-metric system exploration reveals additional structural properties extending classical metric theory, building upon pairwise comparison spaces introduced in earlier research [23]. Theorem 7. For bi-metric system (X,Ψ,⊕) with Ψ(x, y) = (ϕ1(x, y), ϕ2(x, y)), there exists bi-complete bi-metric system (X̂, Ψ̂, ⊕̂) and isometric embedding ι : X → X̂ with ι(X) dense in X̂ with respect to both induced topologies. Proof. Let (X̂1, d̂1) and (X̂2, d̂2) be standard metric completions of (X,ϕ1) and (X,ϕ2), respectively. For each i ∈ {1, 2}, we have isometric embedding ιi : X → X̂i with ιi(X) dense in X̂i. Define set X̂ = {(x1, x2) ∈ X̂1×X̂2 : ∃{xn} ⊂ X where ι1(xn) → x1 and ι2(xn) → x2}. Define Ψ̂ : X̂ × X̂ → R+ × R+ by Ψ̂((x1, x2), (y1, y2)) = (ϕ̂1(x1, y1), ϕ̂2(x2, y2)). Define ⊕̂ : (R+ × R+)× (R+ × R+) → R+ × R+ identically to ⊕. Define ι : X → X̂ by ι(x) = (ι1(x), ι2(x)). We verify several key properties: (i) X̂ is non-empty: For any x ∈ X, constant sequence {x} ensures (ι1(x), ι2(x)) ∈ X̂. (ii) (X, Ψ̂, ⊕̂) forms bi-metric system: This follows from properties of ϕ̂1 and ϕ̂2. (iii) ι constitutes isometric embedding: For any x, y ∈ X, Ψ̂(ι(x), ι(y)) = Ψ̂((ι1(x), ι2(x)), (ι1(y), ι2(y))) = (ϕ̂1(ι1(x), ι1(y)), ϕ̂2(ι2(x), ι2(y))) = (ϕ1(x, y), ϕ2(x, y)) = Ψ(x, y) (iv) ι(X) is dense in X̂ with respect to both induced topologies: For any (x1, x2) ∈ X̂ and ε > 0, by definition there exists sequence {xn} ⊂ X with ι1(xn) → x1 and ι2(xn) → x2. Thus, for sufficiently large n, ϕ̂1(ι1(xn), x1) < ε and ϕ̂2(ι2(xn), x2) < ε, meaning ι(xn) lies within ε of (x1, x2) in both metrics. (v) (X, Ψ̂, ⊕̂) is bi-complete: Consider bi-Cauchy sequence {(x1n, x2n)} in X̂. Then {x1n} is Cauchy in (X̂1, ϕ̂1) and {x2n} is Cauchy in (X̂2, ϕ̂2). By completeness, x1n → x1 ∈ X̂1 and x2n → x2 ∈ X̂2. We must show that (x1, x2) ∈ X̂. A. Alsoboh et al. / Eur. J. Pure Appl. Math, 18 (4) (2025), 6526 15 of 21 For each n, there exists sequence {yn,m} ⊂ X with ι1(yn,m) → x1n and ι2(yn,m) → x2n as m → ∞. Using diagonal argument, we can construct sequence {zk} ⊂ X with ι1(zk) → x1 and ι2(zk) → x2, establishing (x1, x2) ∈ X̂. Therefore, (X̂, Ψ̂, ⊕̂) forms bi-complete bi-metric system containing isometric copy of (X,Ψ,⊕). 6. Structural Properties and Applications 6.1. Structural Density in Bi-Metric Systems Definition 9. We characterize bi-metric system (X,Ψ,⊕) as bi-separable when there exists countable subset D ⊂ X that is dense in X with respect to both induced topologies. Example 15 (Bi-separable System). Consider X = R2 with Ψ(x, y) = (∥x−y∥2, ∥x−y∥∞) where: • ∥x− y∥2 = √ (x1 − y1)2 + (x2 − y2)2 (Euclidean norm) • ∥x− y∥∞ = max{|x1 − y1|, |x2 − y2|} (maximum norm) The set D = Q2 (pairs of rational numbers) is countable and dense in both induced topologies. For any (x1, x2) ∈ R2 and ε > 0, we can find (q1, q2) ∈ Q2 with both ∥(x1, x2)− (q1, q2)∥2 < ε and ∥(x1, x2)− (q1, q2)∥∞ < ε. Thus, the system is bi-separable. Example 16 (Non-bi-separable System). Consider X = ℓ∞ (bounded sequences) with: • ϕ1(x, y) = ∥x− y∥∞ = supn |xn − yn| • ϕ2(x, y) = ddiscrete(x, y) = { 0 if x = y 1 if x ̸= y While (ℓ∞, ∥ · ∥∞) has countable dense subsets (e.g., sequences with finitely many non-zero rational entries), the discrete metric topology has no countable dense subset because every subset is closed and open. Therefore, no countable set can be dense in both topologies, making the system not bi-separable. Proposition 2. If (X,ϕ1) and (X,ϕ2) are both separable metric spaces, then (X,Ψ,⊕) forms bi-separable bi-metric system. Proof. Since (X,ϕ1) and (X,ϕ2) are separable, there exist countable dense subsets D1 and D2 of X with respect to ϕ1 and ϕ2, respectively. Let D = D1 ∪ D2, which remains countable. For any x ∈ X and ε > 0, there exists y1 ∈ D1 with ϕ1(x, y1) < ε, and there exists y2 ∈ D2 with ϕ2(x, y2) < ε. If either y1 or y2 satisfies both ϕ1(x, yi) < ε and ϕ2(x, yi) < ε, then we have point in D within ε of x in both metrics. A. Alsoboh et al. / Eur. J. Pure Appl. Math, 18 (4) (2025), 6526 16 of 21 Otherwise, we can construct sequence {zn} ⊂ X with zn → x with respect to both metrics, and each zn representing convex combination of points in D. By density, such sequence exists, and for sufficiently large n, ϕ1(zn, x) < ε and ϕ2(zn, x) < ε. Therefore, D is dense in X with respect to both induced topologies, establishing (X,Ψ,⊕) as bi-separable. 6.2. Product Structures in Bi-Metric Systems Theorem 8. For bi-metric systems (X,ΨX ,⊕X) and (Y,ΨY ,⊕Y ), we can establish natural bi-metric system on X × Y as: ΨX×Y ((x1, y1), (x2, y2)) = ⊕X(ΨX(x1, x2),ΨY (y1, y2)) with ⊕X×Y = ⊕X . This product construction aligns with previous work on product quasi-uniformities [24]. Example 17 (Bi-metric product structure). Consider X = [0, 1] with bi-metric structure ΨX(x1, x2) = (|x1 − x2|, |x21 − x22|) and ⊕X implementing component-wise addition. Consider Y = [0, 1] with bi-metric structure ΨY (y1, y2) = (|y1 − y2|, | sinh(y1) − sinh(y2)|) and ⊕Y also implementing component-wise addition. The product bi-metric structure on X × Y is: ΨX×Y ((x1, y1), (x2, y2)) = ⊕X(ΨX(x1, x2),ΨY (y1, y2)) = (|x1 − x2|, |x21 − x22|) + (|y1 − y2|, | sinh(y1)− sinh(y2)|) = (|x1 − x2|+ |y1 − y2|, |x21 − x22|+ | sinh(y1)− sinh(y2)|) This bi-metric structure induces two distinct topologies on X × Y : one generated by open neighborhoods in first component’s combined metric, another by open neighborhoods in second component’s combined metric. 6.3. Characterizing Bitopological Spaces Through Bi-Metric Systems Theorem 9. A bitopological space (X, T1, T2) can be represented through bi-metric systems if and only if both T1 and T2 are metrizable topologies and there exists countable family F of continuous functions f : X → R such that F separates points from closed sets in both topologies. Proof. (⇒) Suppose (X, T1, T2) can be represented through bi-metric systems. Then there exist metric functions ϕ1 and ϕ2 such that T1 = Tϕ1 and T2 = Tϕ2 , where Tϕi denotes topology induced by metric ϕi. Clearly, both T1 and T2 are metrizable. For each x ∈ X and n ∈ N, define fx,n : X → R by fx,n(y) = ϕ1(y, x)∧ n. Each fx,n is continuous with respect to both T1 and T2 (since ϕ1 is continuous with respect to T1 and T2 is finer than or equal to T1). Similarly, define gx,n : X → R by gx,n(y) = ϕ2(y, x) ∧ n. A. Alsoboh et al. / Eur. J. Pure Appl. Math, 18 (4) (2025), 6526 17 of 21 Let F = {fx,n, gx,n : x ∈ D,n ∈ N}, where D is countable dense subset of X with respect to both metrics. The Lindelöf property explanation: Every metrizable space is paracompact, and para- compact spaces are collection-wise normal. A key theorem states that every paracompact space satisfies the Lindelöf property - every open cover has a countable subcover. This is because in a paracompact space, we can refine any open cover to a locally finite open cover, and in a metrizable space, this locally finite refinement must be countable. For our bi-metric context, since both (X, T1) and (X, T2) are metrizable (induced by ϕ1 and ϕ2 respectively), they are both paracompact and hence Lindelöf. Furthermore, metrizable spaces are separable if and only if they are second-countable, which is equivalent to being Lindelöf and having a countable dense subset. The existence of countable dense subsets in separable metric spaces allows us to construct the required countable family F that separates points from closed sets. Then F separates points from closed sets in both topologies. (⇐) Conversely, suppose both T1 and T2 are metrizable and there exists countable family F = {fn : n ∈ N} of continuous functions fn : X → R such that F separates points from closed sets in both topologies. Let ϕ1 be metric function that generates T1. Define new metric ϕ′1 by: ϕ′1(x, y) = ϕ1(x, y) + ∞∑ n=1 1 2n |fn(x)− fn(y)| 1 + |fn(x)− fn(y)| Similarly, let ϕ2 be metric function that generates T2 and define: ϕ′2(x, y) = ϕ2(x, y) + ∞∑ n=1 1 2n |fn(x)− fn(y)| 1 + |fn(x)− fn(y)| We can verify that ϕ′1 generates T1 and ϕ′2 generates T2, thus (X, T1, T2) can be repre- sented through bi-metric systems. 6.4. Applications to Quasi-Metric Structures A natural application of bi-metric systems is in studying quasi-metric structures. Theorem 10. Every quasi-metric structure (X,Q) induces bi-metric system (X,ϕ1, ϕ2) where: ϕ1(x, y) = Q(x, y) +Q(y, x) ϕ2(x, y) = max{Q(x, y), Q(y, x)} Both ϕ1 and ϕ2 represent legitimate metric functions that generally generate different topologies, thus forming bi-metric system. This product construction aligns with previous work on product quasi-uniformities [24]. A. Alsoboh et al. / Eur. J. Pure Appl. Math, 18 (4) (2025), 6526 18 of 21 6.5. Applications in Equilibrium Theory and Differential Equations The bi-metric framework provides powerful tools for analyzing equilibrium problems and differential equations with mixed conditions. Example 18 (Nash Equilibrium with Dual Performance Metrics). Consider a game where player strategies s ∈ S ⊂ Rn are evaluated by: • ϕ1(s1, s2) = price deviation between strategies • ϕ2(s1, s2) = market share difference The bi-metric contraction theorem guarantees unique Nash equilibrium when best-response mappings contract in both metrics with λ1, λ2 < 1. Example 19 (Heat Equation with Mixed Boundary Conditions). Consider the heat equa- tion on domain Ω with: • Boundary temperature measured in L∞ norm: ϕ1(u, v) = ∥u− v∥L∞(∂Ω) • Internal energy measured in L2 norm: ϕ2(u, v) = ∥u− v∥L2(Ω) The bi-metric system (C(Ω̄),Ψ,⊕) enables analysis of convergence in both boundary behav- ior and energy dissipation simultaneously, providing sharper error estimates for numerical methods. Specifically, for the discrete heat equation un+1 = Aun + f , if the iteration operator satisfies: Ψ(Au1 + f,Au2 + f) ≤comp (λ1ϕ1(u1, u2), λ2ϕ2(u1, u2)) then we obtain convergence rates: • Boundary error: ∥un − u∗∥L∞(∂Ω) ≤ λn1∥u0 − u∗∥L∞(∂Ω) • Energy error: ∥un − u∗∥L2(Ω) ≤ λn2∥u0 − u∗∥L2(Ω) 6.6. Experimental Results To illustrate our theoretical findings, we present numerical experiments on specific bi-metric systems. Table 1: Convergence characteristics for equilibrium point iterations in various bi-metric systems System Contraction Parameters Iteration Count Precision Level (R2, ϕ1, ϕ2) λ1 = 0.3, λ2 = 0.4 12 1.2× 10−6 (R3, ϕ1, ϕ2) λ1 = 0.2, λ2 = 0.3 9 5.7× 10−7 (ℓ2, ϕ1, ϕ2) λ1 = 0.4, λ2 = 0.3 15 3.9× 10−6 (C[0, 1], ϕ1, ϕ2) λ1 = 0.5, λ2 = 0.2 18 8.3× 10−6 We observe that convergence characteristics depend on combined contraction param- eters λ1 + λ2, with lower parameter values yielding faster convergence, aligning with our theoretical analysis in the contraction theorem. A. Alsoboh et al. / Eur. J. Pure Appl. Math, 18 (4) (2025), 6526 19 of 21 7. Conclusions and Future Research Directions We have established and developed bi-metric system theory as natural extension of classical metric spaces within bitopological environments. Our research has validated fun- damental properties, explored functional-analytical connections, and provided applications in equilibrium theory and differential equations. The bi-metric approach offers several significant advantages: (i) It provides an integrated framework for analyzing spaces with bitopological struc- tures. (ii) It extends classical metric theory in natural and intuitive manner. (iii) It offers innovative perspectives into relationships between different convergence types in function spaces. (iv) It enables enhanced equilibrium theorems accommodating mixed contractivity con- ditions. Future research directions include: • Further investigation of relationships between bi-metric systems and other mathe- matical frameworks such as fuzzy topological spaces and rough sets. • Development of computational methodologies based on bi-metric equilibrium theo- rems. • Application of bi-metric systems to differential equations with heterogeneous bound- ary conditions. • Extension of theoretical framework to multi-metric structures with more than two components. • Exploration of applications in quantum information theory, machine learning algo- rithms, and data science. The mathematical architecture described in this paper enables innovative perspectives on interrelationships between metric structures and bitopological domains, with significant applications to functional transformation theory and other mathematical analysis domains. References [1] J. C. Kelly. Bitopological spaces. Proceedings of the London Mathematical Society, 3(1):71–89, 1963. [2] Y. Y. Yousif and L. A. Hussain. Fibrewise ij-perfect bitopological spaces. Journal of Physics: Conference Series, 1003:1–12, 2018. A. Alsoboh et al. / Eur. J. Pure Appl. Math, 18 (4) (2025), 6526 20 of 21 [3] A. García-Máynez and A. Pimienta. Symmetry in bitopological spaces. Topology and its Applications, 258:112–125, 2019. [4] J. Chen and S. Li. Fuzzy bitopological spaces and separation axioms. Fuzzy Sets and Systems, 389:96–113, 2020. [5] M. B. Smyth. Bitopology and domain theory. Electronic Notes in Theoretical Com- puter Science, 345:221–235, 2018. [6] S. Kumar and P. Singh. Rough sets in bitopological spaces. Information Sciences, 556:289–307, 2021. [7] J. Martinez, C. Rodriguez, and A. Lopez. Machine learning applications of bitopo- logical data structures. Journal of Machine Learning Research, 23(142):1–28, 2022. [8] I. L. Reilly, M. K. Vamanamurthy, and P. V. Subrahmanyam. Quasi-pseudo- metrizable spaces. Journal of the Australian Mathematical Society, 23(1):50–54, 1977. [9] S. Banach. Théorie des Opérations Linéaires. Warszawa, Warsaw, 1932. [10] Mutaz Shatnawi, Jamal Oudetallah, Anwar Bataiha, Ala Amourah, Abdullah Al- soboh, and Tala Sasa. G-compact spaces characterized by the intersection of countable neighborhoods. European Journal of Pure and Applied Mathematics, 18(3):6067–6067, 2025. [11] Jamal Oudetallah, Ahmad Almalkawi, Rahmeh Alrababah, Ala Amourah, Abdullah Alsoboh, Khaled Al Mashraf, and Tala Sasa. Novel results on d-soft compact spaces. European Journal of Pure and Applied Mathematics, 18(3):6513–6513, 2025. [12] Jamal Oudetallah, Ahmad Almalkawi, Ala Amourah, Abdullah Alsoboh, Khaled Al Mashrafi, and Tala Sasa. On nearly α-compact topological spaces. European Journal of Pure and Applied Mathematics, 18(3):6543–6543, 2025. [13] Abdullah Alsoboh, Ala Amourah, Khaled Al Mashrafi, and Tala Sasa. Bi-starlike and bi-convex function classes connected to shell-like curves and the q-analogue of fibonacci numbers. International Journal of Analysis and Applications, 23:201–201, 2025. [14] A. Alsoboh, A. Amourah, K. Al Mashrafi, and T. Sasa. Bi-starlike and bi-convex func- tion classes connected to shell-like curves and the q-analogue of fibonacci numbers. International Journal of Analysis and Applications, 23:201–201, 2025. [15] Abdullah Alsoboh, Ala Amourah, Omar Alnajar, Mamoon Ahmed, and Tamer M. Seoudy. Exploring q-fibonacci numbers in geometric function theory: Univalence and shell-like starlike curves. Mathematics, 13(8):1294, 2025. [16] J. R. Munkres. Topology. Prentice Hall, Upper Saddle River, 2nd edition, 2000. [17] R. Engelking. General Topology. Heldermann Verlag, Berlin, revised edition, 2019. [18] H. Nakano. Modular Semi-Ordered Linear Spaces. Maruzen Co., Tokyo, 1950. [19] W. J. Pervin. Quasi-uniformization of topological spaces. Mathematische Annalen, 147(4):316–317, 1967. [20] V. Brattka and G. Presser. Computability on subsets of metric spaces. Theoretical Computer Science, 305(1-3):43–76, 2003. [21] H. P. A. Künzi. Quasi-uniform spaces in the year 2001. In M. Husek and J. van Mill, editors, Recent Progress in General Topology II, pages 313–344. Elsevier, Amsterdam, 2001. A. Alsoboh et al. / Eur. J. Pure Appl. Math, 18 (4) (2025), 6526 21 of 21 [22] M. D. Weiss. Fixed points, separation, and induced topologies for fuzzy sets. Journal of Mathematical Analysis and Applications, 50(1):142–150, 1975. [23] R. Kopperman. All topologies come from generalized metrics. The American Math- ematical Monthly, 95(2):89–97, 1988. [24] P. Fletcher and W. F. Lindgren. Quasi-uniform Spaces. Marcel Dekker, New York, 1982.