Survey of Artificial Intelligence Model Marketplace
Ventinel: Automated Detection of Android Vishing Apps Using Optical Character Recognition
Shared-Custodial Wallet for Multi-Party Crypto-Asset Management
On Edge-Fog-Cloud Collaboration and Reaping Its Benefits: A Heterogeneous Multi-Tier Edge Computing Architecture
A Comparative Survey of Centralised and Decentralised Identity Management Systems: Analysing Scalability, Security, and Feasibility

Journal Description

Future Internet

Future Internet is an international, peer-reviewed, open access journal on internet technologies and the information society, published monthly online by MDPI.

Open Access— free for readers, with article processing charges (APC) paid by authors or their institutions.
High Visibility: indexed within Scopus, ESCI (Web of Science), Ei Compendex, dblp, Inspec, and other databases.
Journal Rank: JCR - Q2 (Computer Science, Information Systems) / CiteScore - Q1 (Computer Networks and Communications)
Rapid Publication: manuscripts are peer-reviewed and a first decision is provided to authors approximately 16.9 days after submission; acceptance to publication is undertaken in 2.7 days (median values for papers published in this journal in the second half of 2024).
Recognition of Reviewers: reviewers who provide timely, thorough peer-review reports receive vouchers entitling them to a discount on the APC of their next publication in any MDPI journal, in appreciation of the work done.

Impact Factor: 2.8 (2023); 5-Year Impact Factor: 3.0 (2023)

Imprint Information Journal Flyer Open Access ISSN: 1999-5903

Latest Articles

20 pages, 2207 KiB

Open AccessArticle

A Novel TLS-Based Fingerprinting Approach That Combines Feature Expansion and Similarity Mapping

by Amanda Thomson, Leandros Maglaras and Naghmeh Moradpoor

Future Internet 2025, 17(3), 120; https://doi.org/10.3390/fi17030120 - 7 Mar 2025

Abstract

Malicious domains are part of the landscape of the internet but are becoming more prevalent and more dangerous both to companies and to individuals. They can be hosted on various technologies and serve an array of content, including malware, command and control and complex phishing sites that are designed to deceive and expose. Tracking, blocking and detecting such domains is complex, and very often it involves complex allowlist or denylist management or SIEM integration with open-source TLS fingerprinting techniques. Many fingerprinting techniques, such as JARM and JA3, are used by threat hunters to determine domain classification, but with the increase in TLS similarity, particularly in CDNs, they are becoming less useful. The aim of this paper was to adapt and evolve open-source TLS fingerprinting techniques with increased features to enhance granularity and to produce a similarity-mapping system that would enable the tracking and detection of previously unknown malicious domains. This was achieved by enriching TLS fingerprints with HTTP header data and producing a fine-grain similarity visualisation that represented high-dimensional data using MinHash and Locality-Sensitive Hashing. Influence was taken from the chemistry domain, where the problem of high-dimensional similarity in chemical fingerprints is often encountered. An enriched fingerprint was produced, which was then visualised across three separate datasets. The results were analysed and evaluated, with 67 previously unknown malicious domains being detected based on their similarity to known malicious domains and nothing else. The similarity-mapping technique produced demonstrates definite promise in the arena of early detection of malware and phishing domains. Full article

(This article belongs to the Special Issue Intrusion Detection and Resiliency in Cyber-Physical Systems and Networks)

► Show Figures

Figure 1

39 pages, 9925 KiB

Open AccessArticle

Dynamic Workload Management System in the Public Sector: A Comparative Analysis

by Konstantinos C. Giotopoulos, Dimitrios Michalopoulos, Gerasimos Vonitsanos, Dimitris Papadopoulos, Ioanna Giannoukou and Spyros Sioutas

Future Internet 2025, 17(3), 119; https://doi.org/10.3390/fi17030119 - 6 Mar 2025

Abstract

Efficient human resource management is critical to public sector performance, particularly in dynamic environments where traditional systems struggle to adapt to fluctuating workloads. The increasing complexity of public sector operations and the need for equitable task allocation highlight the limitations of conventional evaluation methods, which often fail to account for variations in employee performance and workload demands. This study addresses these challenges by optimizing load distribution through predicting employee capability using data-driven approaches, ensuring efficient resource utilization and enhanced productivity. Using a dataset encompassing public/private sector experience, educational history, and age, we evaluate the effectiveness of seven machine learning algorithms: Linear Regression, Artificial Neural Networks (ANNs), Adaptive Neuro-Fuzzy Inference System (ANFIS), Support Vector Machine (SVM), Gradient Boosting Machine (GBM), Bagged Decision Trees, and XGBoost in predicting employee capability and optimizing task allocation. Performance is assessed through ten evaluation metrics, including Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE), ensuring a comprehensive assessment of accuracy, robustness, and bias. The results demonstrate ANFIS as the superior model, consistently outperforming other algorithms across all metrics. By synergizing fuzzy logic’s capacity to model uncertainty with neural networks’ adaptive learning, ANFIS effectively captures non-linear relationships and variations in employee performance, enabling precise capability predictions in dynamic environments. This research highlights the transformative potential of machine learning in public sector workforce management, underscoring the role of data-driven decision-making in improving task allocation, operational efficiency, and resource utilization. Full article

► Show Figures

Figure 1

41 pages, 603 KiB

Open AccessReview

Edge and Cloud Computing in Smart Cities

by Maria Trigka and Elias Dritsas

Future Internet 2025, 17(3), 118; https://doi.org/10.3390/fi17030118 - 6 Mar 2025

Abstract

The evolution of smart cities is intrinsically linked to advancements in computing paradigms that support real-time data processing, intelligent decision-making, and efficient resource utilization. Edge and cloud computing have emerged as fundamental pillars that enable scalable, distributed, and latency-aware services in urban environments. Cloud computing provides extensive computational capabilities and centralized data storage, whereas edge computing ensures localized processing to mitigate network congestion and latency. This survey presents an in-depth analysis of the integration of edge and cloud computing in smart cities, highlighting architectural frameworks, enabling technologies, application domains, and key research challenges. The study examines resource allocation strategies, real-time analytics, and security considerations, emphasizing the synergies and trade-offs between cloud and edge computing paradigms. The present survey also notes future directions that address critical challenges, paving the way for sustainable and intelligent urban development. Full article

(This article belongs to the Special Issue IoT, Edge, and Cloud Computing in Smart Cities)

► Show Figures

Figure 1

25 pages, 818 KiB

Open AccessArticle

Emergency Messaging System for Urban Vehicular Networks Inspired by Social Insects’ Stigmergic Communication

by Ojilvie Avila-Cortés, Saúl E. Pomares Hernández, Julio César Pérez-Sansalvador and Lil María Xibai Rodríguez-Henríquez

Future Internet 2025, 17(3), 117; https://doi.org/10.3390/fi17030117 - 6 Mar 2025

Abstract

For occupant safety in vehicular networks, emergency messages derived from vehicular incidents should be exchanged only during their validity period and in zones containing involved entities. Problems arise for mobile entities in vehicular networks that change their location over time, where data may be further communicated in out-of-context space and time. Current solutions extend from the naive assumption of notifying every entity in the network about emergencies with data flooding and clusters and by means of specific communication only in the affected zones—geo-routing—of incidents’ relative data. However, delivering useless data to uninvolved entities results in wasted resources and more overheads in the former cases and the work of obtaining knowledge and secondary site services from neighbors in the latter. In this paper, we propose that the common task of disseminating emergency messages for occupant safety among entities should only be communicated only where and when useful, namely, if spatio-temporal constraints apply regarding those incidents. Our solution is inspired by the communication of working social insects that exchange data through pheromones regardless of closeness or knowledge among colony members for food retrieval. The results show that communication based on space–time constraints makes better use of resources than other solutions. Full article

(This article belongs to the Special Issue Intelligent Telecommunications Mobile Networks)

► Show Figures

Figure 1

19 pages, 1230 KiB

Open AccessArticle

A Neural-Symbolic Approach to Extract Trust Patterns in IoT Scenarios

by Fabrizio Messina, Domenico Rosaci and Giuseppe M. L. Sarnè

Future Internet 2025, 17(3), 116; https://doi.org/10.3390/fi17030116 - 6 Mar 2025

Abstract

Trust and reputation relationships among objects represent key aspects of smart IoT object communities with social characteristics. In this context, several trustworthiness models have been presented in the literature that could be applied to IoT scenarios; however, most of these approaches use scalar measures to represent different dimensions of trust, which are then integrated into a single global trustworthiness value. Nevertheless, this scalar approach within the IoT context holds a few limitations that emphasize the need for models that can capture complex trust relationships beyond vector-based representations. To overcome these limitations, we already proposed a novel trust model where the trust perceived by one object with respect to another is represented by a directed, weighted graph. In this model, called T-pattern, the vertices represent individual trust dimensions, and the arcs capture the relationships between these dimensions. This model allows the IoT community to represent scenarios where an object may lack direct knowledge of a particular trust dimension, such as reliability, but can infer it from another dimension, like honesty. The proposed model can represent trust structures of the type described, where multiple trust dimensions are interdependent. This work represents a further contribution by presenting the first real implementation of the T-pattern model, where a neural-symbolic approach has been adopted as inference engine. We performed experiments that demonstrate the capability in inferring trust of both the T-pattern and this specific implementation. Full article

(This article belongs to the Special Issue Joint Design and Integration in Smart IoT Systems)

► Show Figures

Figure 1

Figure 1
An example of a simple TPN with two smart IoT objects (i.e., <math display="inline"><semantics> <msub> <mi>o</mi> <mn>1</mn> </msub> </semantics></math> and <math display="inline"><semantics> <msub> <mi>o</mi> <mn>2</mn> </msub> </semantics></math>) considering expertise (X), honesty (H), reliability (R), and security (S) trust issues. Full article ">Figure 2
The three-layer TPA architecture. Full article ">Figure 3
Percentage error <math display="inline"><semantics> <mrow> <mi>e</mi> <mo>(</mo> <mi>H</mi> <mi>O</mi> <mi>N</mi> <mo>)</mo> </mrow> </semantics></math> for different values of number of smart IoT objects n and number of groups k. Full article ">Figure 4
Percentage error <math display="inline"><semantics> <mrow> <mi>e</mi> <mo>(</mo> <mi>S</mi> <mi>E</mi> <mi>C</mi> <mo>)</mo> </mrow> </semantics></math> for different values of number of smart IoT objects n and number of groups k. Full article ">Figure 5
Percentage error <math display="inline"><semantics> <mrow> <mi>e</mi> <mo>(</mo> <mi>E</mi> <mi>X</mi> <mi>P</mi> <mo>)</mo> </mrow> </semantics></math> for different values of number of smart IoT objects n and number of groups k. Full article ">

36 pages, 21621 KiB

Open AccessArticle

CityBuildAR: Enhancing Community Engagement in Placemaking Through Mobile Augmented Reality

by Daneesha Ranasinghe, Nayomi Kankanamge, Chathura De Silva, Nuwani Kangana, Rifat Mahamood and Tan Yigitcanlar

Future Internet 2025, 17(3), 115; https://doi.org/10.3390/fi17030115 - 6 Mar 2025

Abstract

Mostly, public places are planned and designed by professionals rather engaging the community in the design process. Even if the community engaged, the engagement process was limited to hand drawings, manual mappings, or public discussions, which limited the general public to visualize and well-communicate their aspirations with the professionals. Against this backdrop, this study intends to develop a mobile application called “CityBuildAR”, which uses Augmented Reality technology that allows the end user to visualize their public spaces in a way they want. CityBuildAR was developed by the authors using the Unity Real-Time Development Platform, and the app was developed for an Android Operating System. The app was used to assess community interests in designing open spaces by categorizing participants into three groups: those with limited, average, and professional knowledge of space design. The open cafeteria of the University of Moratuwa, Sri Lanka served as the testbed for this study. The study findings revealed that: (a) Mobile Augmented Reality is an effective way to engage people with limited knowledge in space design to express their design thinking, (b) Compared to professionals, the general public wanted to have more green elements in the public space; (c) Compared to the professionals, the general public who were not conversant with the designing skills found the app more useful to express their ideas. The study guides urban authorities in their placemaking efforts by introducing a novel approach to effectively capture community ideas for creating inclusive public spaces. Full article

(This article belongs to the Special Issue Virtual Reality and Metaverse: Impact on the Digital Transformation of Society II)

► Show Figures

Figure 1

20 pages, 1552 KiB

Open AccessArticle

SwiftSession: A Novel Incremental and Adaptive Approach to Rapid Traffic Classification by Leveraging Local Features

by Tieqi Xi, Qiuhua Zheng, Chuanhui Cheng, Ting Wu, Guojie Xie, Xuebiao Qian, Haochen Ye and Zhenyu Sun

Future Internet 2025, 17(3), 114; https://doi.org/10.3390/fi17030114 - 3 Mar 2025

Abstract

Network traffic classification is crucial for effective security management. However, the increasing prevalence of encrypted traffic and the confidentiality of protocol details have made this task more challenging. To address this issue, we propose a progressive, adaptive traffic classification method called SwiftSession, designed to achieve real-time and accurate classification. SwiftSession extracts statistical and sequential features from the first K packets of traffic. Statistical features capture overall characteristics, while sequential features reflect communication patterns. An initial classification is conducted based on the first K packets during the classification process. If the prediction meets the predefined probability threshold, processing stops; otherwise, additional packets are received. This progressive approach dynamically adjusts the required packets, enhancing classification efficiency. Experimental results show that traffic can be effectively classified by using only the initial K packets. Moreover, on most datasets, the classification time is reduced by more than 70%. Unlike existing methods, SwiftSession enhances the classification speed while ensuring classification accuracy. Full article

► Show Figures

Figure 1

28 pages, 368 KiB

Open AccessArticle

A CIA Triad-Based Taxonomy of Prompt Attacks on Large Language Models

by Nicholas Jones, Md Whaiduzzaman, Tony Jan, Amr Adel, Ammar Alazab and Afnan Alkreisat

Future Internet 2025, 17(3), 113; https://doi.org/10.3390/fi17030113 - 3 Mar 2025

Abstract

The rapid proliferation of Large Language Models (LLMs) across industries such as healthcare, finance, and legal services has revolutionized modern applications. However, their increasing adoption exposes critical vulnerabilities, particularly through adversarial prompt attacks that compromise LLM security. These prompt-based attacks exploit weaknesses in LLMs to manipulate outputs, leading to breaches of confidentiality, corruption of integrity, and disruption of availability. Despite their significance, existing research lacks a comprehensive framework to systematically understand and mitigate these threats. This paper addresses this gap by introducing a taxonomy of prompt attacks based on the Confidentiality, Integrity, and Availability (CIA) triad, an important cornerstone of cybersecurity. This structured taxonomy lays the foundation for a unique framework of prompt security engineering, which is essential for identifying risks, understanding their mechanisms, and devising targeted security protocols. By bridging this critical knowledge gap, the present study provides actionable insights that can enhance the resilience of LLM to ensure their secure deployment in high-stakes and real-world environments. Full article

(This article belongs to the Special Issue Generative Artificial Intelligence (AI) for Cybersecurity)

► Show Figures

Figure 1

28 pages, 8659 KiB

Open AccessArticle

A Regional Multi-Agent Air Monitoring Platform

by Stanimir Stoyanov, Emil Doychev, Asya Stoyanova-Doycheva, Veneta Tabakova-Komsalova, Ivan Stoyanov and Iliya Nedelchev

Future Internet 2025, 17(3), 112; https://doi.org/10.3390/fi17030112 - 3 Mar 2025

Abstract

Plovdiv faces significant air pollution challenges due to geographic, climatic, and industrial factors, making accurate air quality assessment critical. This study presents a hybrid multi-agent platform that integrates symbolic and sub-symbolic artificial intelligence to improve the reliability of air quality monitoring. The platform features a BDI agent, developed using JaCaMo, for processing real-time sensor measurements and a ReAct agent, implemented with LangChain, to incorporate external data sources and perform advanced analytics. By combining these AI approaches, the platform enhances data integration, detects anomalies, and resolves discrepancies between conflicting air quality reports. Furthermore, its scalable and adaptable architecture lays the foundation for future advancements in environmental monitoring. This research represents the first stage in developing an AI-powered system that supports more objective and data-driven decision-making for air quality management in Plovdiv. Full article

(This article belongs to the Special Issue Intelligent Agents and Their Application)

► Show Figures

Figure 1

18 pages, 474 KiB

Open AccessArticle

Frame Aggregation with Simple Block Acknowledgement Mechanism to Provide Strict Quality of Service Guarantee to Emergency Traffic in Wireless Networks

by Shuaib K. Memon, Md Akbar Hossain and Nurul I. Sarkar

Future Internet 2025, 17(3), 111; https://doi.org/10.3390/fi17030111 - 3 Mar 2025

Abstract

This paper proposes a frame aggregation with a simple block acknowledgement (FASBA) mechanism to provide a strict QoS guarantee to life-saving emergency traffic in wireless local area networks. This work builds on our previous work on a multi-preemptive enhanced distributed channel access protocol called MP-EDCA. The main difference between FASBA and MP-EDCA is that MP-EDCA does not provide a strict QoS guarantee to life-saving emergency traffic (e.g., ambulance calls), especially in high-load conditions. Our proposed FASBA protocol solves the problems of achieving a strict QoS guarantee to life-saving emergency traffic. The strict QoS guarantee is achieved by aggregating multiple frames with a two-bit block acknowledgement for transmissions. FASBA assures guaranteed network services by reducing MAC overheads; consequently, it offers higher throughput, lower packet delays, and accommodates a larger number of life-saving emergency nodes during emergencies. The performance of the proposed FASBA is validated by Riverbed Modeler and MATLAB 2024a-based simulation. Results obtained show that the proposed FASBA offers about 30% lower delays, 17% higher throughput, and 60% lower retransmission attempts than MP-EDCA under high-traffic loads. Full article

► Show Figures

Figure 1

24 pages, 9612 KiB

Open AccessArticle

Developing an Urban Digital Twin for Environmental and Risk Assessment: A Case Study on Public Lighting and Hydrogeological Risk

by Vincenzo Barrile, Emanuela Genovese, Clemente Maesano, Sonia Calluso and Maurizio Pasquale Manti

Future Internet 2025, 17(3), 110; https://doi.org/10.3390/fi17030110 - 1 Mar 2025

Abstract

Improvements in immersive technology are opening up new opportunities for land management and urban planning, enabling the creation of detailed virtual models for examining and simulating real-world short-, medium-, and long-term scenarios. The goal of this research is to present the creation of an urban digital twin based on a virtual reality city replica, that models and visualizes the urban environment in three dimensions using advanced geomatics techniques and IoT technologies. The methodology focuses on two case studies that utilize environmental analysis and virtual simulation: assessing hydrogeological risk and evaluating public light pollution. The Cesium platform was employed to build high-precision 3D models based on topographic, meteorological, and infrastructure data. The proposed methodology calculated a correlation between light pollution and CO₂ equal to 0.51 and a correlation between precipitation, slope, and risk area higher than 0.80. The most critical and high-risk classes are as follows: Dense Discontinuous Urban Fabric, Roads and Associated Lands, Pastures, and Forests. Results show how an urban digital twin can be a powerful tool for monitoring and territorial planning, with concrete applications in the public and risk management fields. This study also highlights the importance of geomatics technologies in the creation of realistic and functional virtual environments for the assessment and sustainable management of urban resources. Full article

(This article belongs to the Special Issue Advances in Smart Environments and Digital Twin Technologies)

► Show Figures

Figure 1

20 pages, 833 KiB

Open AccessArticle

Mobility Prediction and Resource-Aware Client Selection for Federated Learning in IoT

by Rana Albelaihi

Future Internet 2025, 17(3), 109; https://doi.org/10.3390/fi17030109 - 1 Mar 2025

Abstract

This paper presents the Mobility-Aware Client Selection (MACS) strategy, developed to address the challenges associated with client mobility in Federated Learning (FL). FL enables decentralized machine learning by allowing collaborative model training without sharing raw data, preserving privacy. However, client mobility and limited resources in IoT environments pose significant challenges to the efficiency and reliability of FL. MACS is designed to maximize client participation while ensuring timely updates under computational and communication constraints. The proposed approach incorporates a Mobility Prediction Model to forecast client connectivity and resource availability and a Resource-Aware Client Evaluation mechanism to assess eligibility based on predicted latencies. MACS optimizes client selection, improves convergence rates, and enhances overall system performance by employing these predictive capabilities and a dynamic resource allocation strategy. The evaluation includes comparisons with advanced baselines such as Reinforcement Learning-based FL (RL-based) and Deep Learning-based FL (DL-based), in addition to Static and Random selection methods. For the CIFAR dataset, MACS achieved a final accuracy of 95%, outperforming Static selection (85%), Random selection (80%), RL-based FL (90%), and DL-based FL (93%). Similarly, for the MNIST dataset, MACS reached 98% accuracy, surpassing Static selection (92%), Random selection (88%), RL-based FL (94%), and DL-based FL (96%). Additionally, MACS consistently required fewer iterations to achieve target accuracy levels, demonstrating its efficiency in dynamic IoT environments. This strategy provides a scalable and adaptable solution for sustainable federated learning across diverse IoT applications, including smart cities, healthcare, and industrial automation. Full article

(This article belongs to the Special Issue Edge-Cloud Computing and Federated-Split Learning in Internet of Things—Second Edition)

► Show Figures

Figure 1

19 pages, 2746 KiB

Open AccessArticle

Decentralized and Secure Blockchain Solution for Tamper-Proof Logging Events

by J. D. Morillo Reina and T. J. Mateo Sanguino

Future Internet 2025, 17(3), 108; https://doi.org/10.3390/fi17030108 - 1 Mar 2025

Abstract

Log files are essential assets for IT engineers engaged in the security of server and computer systems. They provide crucial information for identifying malicious events, conducting cybersecurity incident analyses, performing audits, system maintenance, and ensuring compliance with security regulations. Nevertheless, there is still the possibility of deliberate data manipulation by own personnel, especially with regard to system access and configuration changes, where error tracking or debugging traces are vital. To address tampering of log files, this work proposes a solution to ensure data integrity, immutability, and non-repudiation through different blockchain-based public registry systems. This approach offers an additional layer of security through a decentralized, tamper-resistant ledger. To this end, this manuscript aims to provide a solid guideline for creating secure log storage systems. For this purpose, methodologies and experiments using two different blockchains are presented to demonstrate their effectiveness in various contexts, such as transactions with and without metadata. The findings suggest that Solana’s response times make it well suited for environments with moderately critical records requiring certification. In contrast, Cardano shows higher response times, thus making it suitable for less frequent events with metadata that requires legitimacy. Full article

(This article belongs to the Special Issue Future Directions in Blockchain Technologies)

► Show Figures

Figure 1

48 pages, 1061 KiB

Open AccessReview

Navigating Challenges and Harnessing Opportunities: Deep Learning Applications in Internet of Medical Things

by John Mulo, Hengshuo Liang, Mian Qian, Milon Biswas, Bharat Rawal, Yifan Guo and Wei Yu

Future Internet 2025, 17(3), 107; https://doi.org/10.3390/fi17030107 - 1 Mar 2025

Abstract

Integrating deep learning (DL) with the Internet of Medical Things (IoMT) is a paradigm shift in modern healthcare, offering enormous opportunities for patient care, diagnostics, and treatment. Implementing DL with IoMT has the potential to deliver better diagnosis, treatment, and patient management. However, the practical implementation has challenges, including data quality, privacy, interoperability, and limited computational resources. This survey article provides a conceptual IoMT framework for healthcare, synthesizes and identifies the state-of-the-art solutions that tackle the challenges of the current applications of DL, and analyzes existing limitations and potential future developments. Through an analysis of case studies and real-world implementations, this work provides insights into best practices and lessons learned, including the importance of robust data preprocessing, integration with legacy systems, and human-centric design. Finally, we outline future research directions, emphasizing the development of transparent, scalable, and privacy-preserving DL models to realize the full potential of IoMT in healthcare. This survey aims to serve as a foundational reference for researchers and practitioners seeking to navigate the challenges and harness the opportunities in this rapidly evolving field. Full article

(This article belongs to the Special Issue The Future Internet of Medical Things, 3rd Edition)

► Show Figures

Figure 1

32 pages, 2442 KiB

Open AccessArticle

Federated Learning System for Dynamic Radio/MEC Resource Allocation and Slicing Control in Open Radio Access Network

by Mario Martínez-Morfa, Carlos Ruiz de Mendoza, Cristina Cervelló-Pastor and Sebastia Sallent-Ribes

Future Internet 2025, 17(3), 106; https://doi.org/10.3390/fi17030106 - 26 Feb 2025

Abstract

The evolution of cellular networks from fifth-generation (5G) architectures to beyond 5G (B5G) and sixth-generation (6G) systems necessitates innovative solutions to overcome the limitations of traditional Radio Access Network (RAN) infrastructures. Existing monolithic and proprietary RAN components restrict adaptability, interoperability, and optimal resource utilization, posing challenges in meeting the stringent requirements of next-generation applications. The Open Radio Access Network (O-RAN) and Multi-Access Edge Computing (MEC) have emerged as transformative paradigms, enabling disaggregation, virtualization, and real-time adaptability—which are key to achieving ultra-low latency, enhanced bandwidth efficiency, and intelligent resource management in future cellular systems. This paper presents a Federated Deep Reinforcement Learning (FDRL) framework for dynamic radio and edge computing resource allocation and slicing management in O-RAN environments. An Integer Linear Programming (ILP) model has also been developed, resulting in the proposed FDRL solution drastically reducing the system response time. On the other hand, unlike centralized Reinforcement Learning (RL) approaches, the proposed FDRL solution leverages Federated Learning (FL) to optimize performance while preserving data privacy and reducing communication overhead. Comparative evaluations against centralized models demonstrate that the federated approach improves learning efficiency and reduces bandwidth consumption. The system has been rigorously tested across multiple scenarios, including multi-client O-RAN environments and loss-of-synchronization conditions, confirming its resilience in distributed deployments. Additionally, a case study simulating realistic traffic profiles validates the proposed framework’s ability to dynamically manage radio and computational resources, ensuring efficient and adaptive O-RAN slicing for diverse and high-mobility scenarios. Full article

(This article belongs to the Special Issue AI and Security in 5G Cooperative Cognitive Radio Networks)

► Show Figures

Figure 1

20 pages, 553 KiB

Open AccessArticle

User-Generated Content and Its Impact on Purchase Intent for Tourism Products: A Comparative Analysis of Millennials and Centennials on TikTok

by Eva Correia Ramos and Célia M. Q. Ramos

Future Internet 2025, 17(3), 105; https://doi.org/10.3390/fi17030105 - 25 Feb 2025

Abstract

In an increasingly technological society, online social networks are essential to support consumer purchasing decisions, primarily through User Generated Content (UGC). In this research, we look at the influence of UGC on purchase intent applied to the tourism product on the TikTok social network. In this sense, a survey was applied to TikTok users aged between 18 and 42 to compare their behaviour with that of the two generations: Millennials and Centennials. The results indicate a relationship of influence between credibility and the usefulness of information and between usefulness and social influence on the intention to buy tourism products and services. In addition, a comparison was made between the results of the sample of individuals belonging to the Millennial Generation and the sample of individuals belonging to the Centennial Generation, with the main discrepancy in the results being the relationship between the need for information and the usefulness of information. These insights pave the way for further research aimed at establishing more robust conclusions in this area. Full article

(This article belongs to the Special Issue Advances in Smart Environments and Digital Twin Technologies)

► Show Figures

Figure 1

15 pages, 1933 KiB

Open AccessArticle

Assessing Browser Security: A Detailed Study Based on CVE Metrics

by Oleksii Chalyi, Kęstutis Driaunys and Vytautas Rudžionis

Future Internet 2025, 17(3), 104; https://doi.org/10.3390/fi17030104 - 25 Feb 2025

Abstract

This study systematically evaluates the vulnerabilities of modern web browsers using developed indices derived from the CVE database, including I_CVE, I_CVSS, I_R and I_T. These indices incorporate metrics such as vulnerability severity and risks, along with browser popularity, to enable a balanced comparison of browser security. The results highlight significant differences in browser security: while Google Chrome and Samsung Internet exhibited lower threat indices, Mozilla Firefox demonstrated consistently higher scores, indicating greater exposure to risks. These observations a slightly contradict widespread opinion. The findings emphasize the importance of timely software updates in mitigating vulnerabilities, as many incidents were linked to outdated browser versions. This study also introduces a robust methodology for assessing browser threats, providing a framework for future research. Potential applications include developing browser-based penetration testing systems to simulate phishing and data extraction scenarios, offering insights into user-specific risks and broader organizational impacts. By combining theoretical analysis with practical implications, this work contributes to advancing browser security and lays the foundation for future applied research in cybersecurity. Full article

(This article belongs to the Special Issue Information and Future Internet Security, Trust and Privacy—Volume III)

► Show Figures

Figure 1

15 pages, 1925 KiB

Open AccessArticle

Gamitest: A Game-like Online Student Assessment System

by Jakub Swacha and Artur Kulpa

Future Internet 2025, 17(3), 103; https://doi.org/10.3390/fi17030103 - 24 Feb 2025

Abstract

The widespread availability of mobile devices has led to the emergence of multiple gamified web and mobile applications for online assessment of students during classes. Their common weak side is that they focus mostly on positive reinforcement, without exploiting the pedagogical potential of loss and failure experience; thus, they are far from an actual game-like experience. In this paper, we present Gamitest, a course-subject-agnostic web application for student assessment that features an original game-like scheme to improve students’ perception of engagement and fun, as well as to reduce their examination stress. The results of the survey-based evaluation of the tool indicate that its design goals have been met and allow us to recommend it for consideration in various forms of student assessment, as well as provide grounds for future work on analyzing the tool’s effects on learning. Full article

(This article belongs to the Topic Advances in Online and Distance Learning)

► Show Figures

Figure 1

19 pages, 5346 KiB

Open AccessArticle

Metastable Substructure Embedding and Robust Classification of Multichannel EEG Data Using Spectral Graph Kernels

by Rashmi N. Muralinath, Vishwambhar Pathak and Prabhat K. Mahanti

Future Internet 2025, 17(3), 102; https://doi.org/10.3390/fi17030102 - 23 Feb 2025

Abstract

Classification of neurocognitive states from Electroencephalography (EEG) data is complex due to inherent challenges such as noise, non-stationarity, non-linearity, and the high-dimensional and sparse nature of connectivity patterns. Graph-theoretical approaches provide a powerful framework for analysing the latent state dynamics using connectivity measures across spatio-temporal-spectral dimensions. This study applies the graph Koopman embedding kernels (GKKE) method to extract latent neuro-markers of seizures from epileptiform EEG activity. EEG-derived graphs were constructed using correlation and mean phase locking value (mPLV), with adjacency matrices generated via threshold-binarised connectivity. Graph kernels, including Random Walk, Weisfeiler–Lehman (WL), and spectral-decomposition (SD) kernels, were evaluated for latent space feature extraction by approximating Koopman spectral decomposition. The potential of graph Koopman embeddings in identifying latent metastable connectivity structures has been demonstrated with empirical analyses. The robustness of these features was evaluated using classifiers such as Decision Trees, Support Vector Machine (SVM), and Random Forest, on Epilepsy-EEG from the Children’s Hospital Boston’s (CHB)-MIT dataset and cognitive-load-EEG datasets from online repositories. The classification workflow combining mPLV connectivity measure, WL graph Koopman kernel, and Decision Tree (DT) outperformed the alternative combinations, particularly considering the accuracy (91.7%) and F1-score (88.9%), The comparative investigation presented in results section convinces that employing cost-sensitive learning improved the F1-score for the mPLV-WL-DT workflow to 91% compared to 88.9% without cost-sensitive learning. This work advances EEG-based neuro-marker estimation, facilitating reliable assistive tools for prognosis and cognitive training protocols. Full article

(This article belongs to the Special Issue eHealth and mHealth)

► Show Figures

Figure 1

Figure 1
The proposed workflow. Full article ">Figure 2
Plots show the optimal clusters of eigenvectors for a synthetic dataset: (a) synthetic dataset with 5 clusters, (b) optimal clustering structure for top-2 eigenvectors, (c) silhouette scores, (d) SSE distance scores. Full article ">Figure 3
Plots show the existence of connectivity substructures (clusters) for sample MIT CHB Epileptiform EEG data. (a) Heatmap of connectivity matrix (>0.5) for preictal-state sample; (b) inherent clustering structure for the preictal-state sample; (c) heatmap of connectivity matrix (>0.5) for ictal-state sample; (d) inherent clustering structure for the ictal-state sample. Full article ">Figure 3 Cont.
Plots show the existence of connectivity substructures (clusters) for sample MIT CHB Epileptiform EEG data. (a) Heatmap of connectivity matrix (>0.5) for preictal-state sample; (b) inherent clustering structure for the preictal-state sample; (c) heatmap of connectivity matrix (>0.5) for ictal-state sample; (d) inherent clustering structure for the ictal-state sample. Full article ">Figure 4
Plots show the existence of connectivity substructures (clusters) for sample cognitive load EEG-ERPs. (a) Heatmap of the correlation matrix (>0.5) for idle-state sample; (b) inherent clustering structure for the idle-state sample; (c) heatmap of the correlation matrix (>0.5) for d1B-state sample; (d) inherent clustering structure for the d1B-state sample. Full article ">Figure 5
Plots (a–c) show the heatmap of the Gram matrices produced by the WL kernel, the optimal clusters of top-2 eigenvectors, and the corresponding silhouette scores for 30 subsequent preictal-state samples. Plots (d–f) show the plots for 30 subsequent ictal-state samples. Full article ">Figure 6
Performance comparison regarding accuracy, AUC, precision, recall, and F1-score with and without cost-sensitive learning for different workflows comprising different combinations of connectivity metrics, graph kernels, and classifiers: Plot (a) shows the combinations with mPLV; Plot. (b) shows combinations with correlation coefficient. Full article ">

5 pages, 135 KiB

Open AccessEditorial

Edge Intelligence: Edge Computing for 5G and the Internet of Things

by Yuezhi Zhou and Xu Chen

Future Internet 2025, 17(3), 101; https://doi.org/10.3390/fi17030101 - 23 Feb 2025

Abstract

In recent years, the field of artificial intelligence has made a revolutionary breakthrough [...] Full article

(This article belongs to the Special Issue Edge Intelligence: Edge Computing for 5G and the Internet of Things)

Journal Menu

Journal Browser

► Journal Browser

Highly Accessed Articles

Latest Books

More Books and Reprints...

E-Mail Alert

News

27 February 2025
More than 300 MDPI Journals Indexed in Web of Science

20 February 2025
Future Internet | Selected Editor’s Choice Articles Published in the First Quarter of 2024

19 February 2025
Meet Us at the 2nd International Conference on AI Sensors and Transducers (AIS 2025), 29 July–3 August 2025, Kuala Lumpur, Malaysia

More News & Announcements...

Topics

Propose a Topic

Topic in Applied Sciences, Electronics, Future Internet, Sensors, Smart Cities

Cloud and Edge Computing for Smart Devices Topic Editors: Mehdi Sookhak, Francesco Moscato
Deadline: 20 May 2025

Topic in Applied Sciences, Electronics, Future Internet, Machines, Systems, Technologies, Biomimetics

Theoretical and Applied Problems in Human-Computer Intelligent Systems Topic Editors: Jiahui Yu, Charlie Yang, Zhenyu Wen, Dalin Zhou, Dongxu Gao, Changting Lin
Deadline: 30 June 2025

Topic in Algorithms, Future Internet, Information, Mathematics, Symmetry

Research on Data Mining of Electronic Health Records Using Deep Learning Methods Topic Editors: Dawei Yang, Yu Zhu, Hongyi Xin
Deadline: 31 August 2025

Topic in Algorithms, Applied Sciences, Future Internet, Information, Mathematics

Soft Computing and Machine Learning Topic Editors: Rui Araújo, António Pedro Aguiar, Nuno Lau, Rodrigo Ventura, João Fabro
Deadline: 30 September 2025