Computer networks in clinical environments are commonly treated as passive data conduits. This narrative review examines them instead as technical infrastructure that conditions how nurses exchange clinical information, communicate across disciplines, and access professional knowledge. Six bibliographic databases and the primary standards corpora of HL7, IHE, IETF, IEEE, ISO/IEC and NIST were consulted for English-language sources published between January 2015 and February 2026, together with regulatory instruments in force in the United States and the European Union. Sources were identified purposively, through topic-based searching, citation chasing, and targeted retrieval of standards and regulatory documents from issuing bodies, rather than through a protocol-driven search; 83 reference entries representing 82 distinct works were retained, comprising 33 empirical clinical studies, 36 conceptual or architectural sources, 12 normative specifications, and 2 regulatory instruments. Retained sources were assigned to four evidence classes, each appraised against class-specific criteria, and every substantive claim is labelled as empirically demonstrated (E), architecturally inferred (A), or proposed (P). Three findings follow, and their evidentiary status differs. First, interoperability maturity, availability, and security architecture are architecturally implicated in the completeness and timeliness of nursing information at the point of care (A); direct empirical evidence isolating network properties from co-implemented application and workflow changes was not identified, and the available clinical evidence concerns composite interventions concentrated in high-income settings (E). Second, reported latency, bandwidth, and jitter figures are application-dependent design targets rather than clinical requirements, and no normative source for the figures most frequently cited was identified (S). Third, capabilities in 5G, edge computing, and distributed ledgers remain predominantly pre-deployment in nursing contexts (P). The review accordingly does not conclude that network architecture determines nursing or patient outcomes; it concludes that the architectural case is substantially stronger than the empirical one. A three-stage analytical model, with candidate measurable indicators, is proposed as a framework requiring empirical validation rather than as a demonstrated causal account.
The unmanned aerial vehicle (UAV) field is rapidly evolving, increasing the need for effective security mechanisms to protect data in transit in resource-constrained environments. Although many solutions have been proposed to achieve the required security objectives, these traditional solutions either suffer from high computational complexity or are vulnerable to quantum computing threats, an open challenge. However, this research presents a lightweight security scheme that uses the Kyber 512 algorithm for a quantum-resistant key-exchange algorithm, followed by a Wi-Fi Protected Access 2 (WPA2)-inspired mechanism for mutual authentication. Furthermore, an Ascon Authenticated Encryption with Associated Data (ASCON-AEAD)-based lightweight algorithm is utilized for data encryption and verification. Additionally, the scheme is implemented in a simulation-based UAV environment to demonstrate its performance in protecting the confidentiality and integrity of MAVLink messages transferred between aircraft and the ground control station (CGS), as well as the communicated party's authenticity. Finally, the randomness of both authentication and encryption keys was evaluated using the National Institute of Standards and Technology (NIST) tests across 50 independent sessions, with a significance level of 0.01. All tests showed average p-values higher than the accepted significance level, with pass rates ranging from 96 to 100%, indicating good statistical randomness for the generated keys and supporting their security quality within the proposed protocol. In contrast, it also demonstrated a successful transfer of encrypted Micro Air Vehicle Link (MAVLink) messages. It ensured their validity only after verification of the identities of both communicating parties (i.e., mutual authentication). Furthermore, the achieved processing time averaged approximately 1.45 ms, ranging from 0.91 to 1.89 ms, which was within acceptable limits for such an environment, making this proposed solution efficient and effective.
Image captioning models are vulnerable to adversarial examples that can induce descriptions inconsistent with visual evidence, commonsense knowledge, or physical rules. Existing detection methods mainly rely on confidence anomalies, feature-space deviations, or reconstruction errors, making them less suitable for cross-modal generative tasks. To address this limitation, we propose SPC-Verify, a Semantic-Phenomenon Consistency Verification framework for adversarial example detection in image captioning. The framework combines Visual-Semantic Consistency Verification (VSCV), which assesses whether token-level visual evidence supports the corresponding generated concepts, with Phenomenon Plausibility Verification (PPV), which evaluates the consistency of complete captions with commonsense knowledge and physical constraints. To improve the reliability of local visual evidence, we introduce SCAR-SAT, a post-hoc heatmap refinement method for Show-Attend-and-Tell-based models that integrates semantic consistency, lightweight counterfactual support, spatial continuity, and temporal stability. We further propose Zero-Shot Cross-Model Threshold Transfer (ZS-CMTT), which learns shared decision thresholds from the joint score distributions of SAT and BLIP and transfers them to unseen captioning models without requiring target-model samples. Experiments on MSCOCO and Flickr30K across SAT and BLIP, under multiple attack settings, demonstrate the effectiveness of the proposed framework. Ablation studies show that the Grounded Category Vocabulary improves token-level verification, while PPV provides complementary sentence-level plausibility assessment. Further evaluations indicate that SCAR-SAT improves token-level visual grounding and that ZS-CMTT facilitates threshold transfer across captioning models. Overall, SPC-Verify provides an interpretable and transferable approach to adversarial example detection by jointly evaluating local visual-semantic consistency and global caption plausibility.
IntroductionThe compute continuum, together with the high-speed and reliable connectivity provided by advanced communication technologies such as 5G or beyond 5G (B5G) networks, enables the development of new types of applications and services. However, deciding when application components should be re-instantiated and selecting the most appropriate destination node according to dynamic execution conditions remains a challenge. Although artificial intelligence (AI), among other techniques, has emerged as a promising approach to address these challenges, the integration and evaluation of such orchestrators remain complex due to the lack of modular and standardized architectures.MethodsIn this work, we present FABRIC, an orchestration platform that aims to evolve current 5G networks by enabling the re-instantiation of computational tasks whenever required through continuous computing and advanced task orchestration across distributed resources, all this whilst being aligned with current 3GPP and ETSI standards. This architecture bridges the gap between distributed computing paradigms and modern mobile network architectures, anticipating the computational and operational demands of next-generation mobile networks thanks to its standards-based nature. This modular design facilitates the implementation of sophisticated re-instantiation strategies while promoting interoperability and extensibility, fostering the adoption of more advanced orchestration mechanisms for future compute continuum environments supported by 5G/B5G networks.ResultsFABRIC is assessed with two use case scenarios, one with a CPU-intensive task and another one re-instantiating an application with user traffic.DiscussionThe results demonstrate that FABRIC supports interoperable service re-instantiation across the compute continuum, while facilitating service orchestration adapted to the requirements of each use case.
IntroductionLightweight cryptographic schemes are needed for resource-constrained environments such as Internet of Things systems, wireless sensor networks, and embedded platforms. This paper presents a symmetric encryption framework based on the coloring of complete n-partite graphs, in which the keystream is derived from partition sizes and multi-character color labels assigned to graph partitions.MethodsA cyclic-padding convention extends all color labels to a common length m = max(ℓ1, ℓ2, …, ℓn) before summation, making the keystream period exactly m by construction. Plaintext characters are encoded using ASCII values and encrypted by symbol-wise modular addition over ℤ127, giving a repeated-key additive cipher with time complexity O(L+n·m) and space complexity O(L+n·m), where L is the message length. A formal threat model distinguishes four adversary categories (ciphertext-only, known-plaintext, chosen-plaintext, and active network adversaries) and sets security thresholds in terms of m.ResultsEncryption and decryption time scale linearly with message length. Empirical ciphertext entropy approaches 6.88 bits per symbol at n = 8 partitions. The scheme provides meaningful confidentiality against ciphertext-only adversaries when L ≤ m.DiscussionThe scheme is vulnerable to known-plaintext attacks, a fundamental limitation shared by all repeated-key additive ciphers. It is therefore positioned as a graph-structured, session-bound, repeated-key additive cipher framework, not a replacement for standardized cryptographic primitives. Its low computational cost supports use in lightweight and resource-constrained settings.
