Networking, Computing and Immersive Technologies for Smart Environments

1. Introduction

Smart environments encompass a large set of technologies, including computer networking, wireless communication, computational infrastructures, sensor networks, and algorithm design. Smart environments aim to provide information-related services in diverse environments ranging from cities, industrial facilities, and agricultural and farming units to cultural heritage sites, classrooms, and homes. Services in such environments are often provided on-site and feature user-friendly interfaces, thus making immersive technologies (virtual, augmented, mixed, and cross reality) a suitable option. In addition, immersive applications can be further integrated in a smart environment using ubiquitous user interfaces and wearable equipment.

In this context, emerging high-performance technologies such as cloud computing, 5G, and network functions virtualization co-exist with low-cost computational services and networking approaches including fog computing, wireless networks, and sensor devices. Both sets of technologies and approaches are employed to provide, combined or independently, (i) real-time response and reliable communication in delay-sensitive or mission-critical applications and (ii) dispersed network access, distributed data collection, and distributed computational offloading for varying—conventional or immersive—applications from industrial automation and agriculture to art performances, video games, and home entertainment.

Moreover, smart approaches can extend to business logic, independently of the computational and network approaches. This includes data-driven decision-making, automation, and the integration of artificial intelligence or innovative design strategies, all aimed at developing versatile applications that adapt to smart environments or create them, even in the absence of conventional infrastructure.

This Special Issue is devoted to works identifying challenges in the described field as well as theoretical and experimental approaches to tackle them. Representative works that fall into various positions of the above spectrum have been selected for publication and listed here. Although a Special Issue of such interdisciplinary nature cannot exhaustively cover the full range of smart environments in depth, we hope that the following papers offer a few interesting ideas, innovative concepts, and promising applications pointing the way for future research.

2. Overview of the Published Articles

Table 1 presents a list of the published articles and their subjects. Each article and their research context is briefly presented in the following paragraphs.

Traffic monitoring and management are crucial aspects of smart cities, with approaches ranging from highly decentralized vehicular ad hoc networks (see [1] for a review on VANET research or [2] for a representative example) to more centralized solutions utilizing fixed equipment. The work of Alam et al. [3] published in this Special Issue belongs to the latter case and discusses the development of an intelligent system for vehicle license plate detection and recognition using convolutional neural networks (CNNs), motivated by the growing issue of traffic rule violations and accidents due to the rising number of vehicles on the road. The system operates in two stages: detecting license plates from vehicle images, enhancing the image resolution, and then recognizing specific characteristics using deep learning techniques. The system is particularly designed for license plates in the Bengali language.

Network-dependent environments need to be safeguarded from potential threats and malfunctions; therefore, routing optimization and DDoS protection [4] in the network has become a necessity for mobile network operators to maintain a good QoS and QoE for customers. The article by Dake et al. [5] focuses on a multi-agent reinforcement learning (RL) framework integrated with software-defined networking (SDN) and the IoT to perform transient load detection and prevention. It addresses challenges like traffic bursts and distributed denial-of-service (DDoS) attacks. The Multiagent Deep Deterministic Policy Gradient (MADDPG) algorithm is proposed for efficient network optimization. The system uses RL agents to enhance network security and optimize routing. The results show improved network performance in terms of delay, jitter, packet loss, and bandwidth usage.

The protection, preservation, and dissemination of cultural heritage have an important role in our society. In the context of smart environments, cultural assets may be part of a greater smart environment or cultural heritage sites can be turned into smart environments themselves to enhance cultural understanding or protection and preservation (see [6,7] for representative examples). In both cases, novel approaches merging immersive technologies with computing and network infrastructure towards the realization of cultural-heritage-oriented goals are of great interest. Kontopanagou et al. [8]’s article in this Special Issue focuses on this subject and presents a framework intended to provide monument visitors with interpretative information specifically targeted to paintings. By making use of the capabilities of mobile augmented reality (position tracking, registration, alignment) and image recognition, visitors can use applications can scan the environment, identify works of art, and project information about those works. Moreover, the approach can serve as a tool for crowdsourcing projects for heritage documentation.

The protection of electronic documents during their transfer in information systems, including smart systems [9,10], is an urgent issue. Melman et al. [11] propose a technology for protecting the authorship of electronic documents by embedding digital watermarks into images contained in the documents. The approach is designed to safeguard ownership against unauthorized copying of the entire document, individual images, or text. Their research also explores the robustness of different watermarking algorithms through experiments. This method is particularly relevant for electronic documents used in smart environments, where image-based watermarking can effectively ensure intellectual property protection.

Immersive technologies, by blending multimedia and transmedia with pedagogical approaches, can be used to create smart learning environments [12]. In their paper, Kaimara et al. [13] outline the development of the Waking Up In the Morning (WUIM) project, a smart learning environment designed for students with learning difficulties. It integrates traditional pedagogical theories with cutting-edge technologies like augmented and virtual reality to teach independent living skills. The transmedia program focuses on inclusive education, blending multiple learning tools and media to support students’ diverse needs. The research highlights WUIM’s role in enhancing engagement and adaptability in educational settings.

Moreover, interactivity in immersive environments is of crucial importance. When users are faced with interactions and decisions, entropy can be deemed an important factor [14]. Understanding how entropy—variations in unpredictability and complexity—affects user immersion can help design smarter, more intuitive systems that adapt to user needs, enhancing their experience in various smart environments. Papavlasopoulos et al. [15] explore the relationship between entropy and video game immersion. They focus on how low-entropy scenes in interactive drama games (such as Heavy Rain, Until Dawn, and Man of Medan) contribute to narrative, physical, and emotional immersion. Using Interpretative Phenomenological Analysis (IPA), the research suggests that entropy is a key factor in creating engaging and immersive experiences, and it explores how these findings can be applied to other immersive technologies.

Creating smart collaborative environments can boost productivity, efficiency, creativity, and innovation. Leonidis et al. [16] present a sophisticated framework, named CognitOS Board, designed to enhance presentation-related activities. The framework integrates principles from the fields of human–computer interaction, ambient intelligence, and software engineering, providing a unified working environment for large interactive boards. CognitOS Board relies on multimodal and well-established interaction metaphors to promote natural user interactions. From a technological perspective, CognitOS Board comprises projectors, touch sensors, speakers, a PC to drive the equipment, and a framework that follows a microservice architecture. The article includes an evaluative study showing that CognitOS Board users did not encounter major issues and rated it with an “A” score on the SUS [17] questionnaire.

Table 1. Overview of published articles, their subjects, and employed technologies and methods.

Authors-Ref.	Title	Domain/Purpose	Technologies/Methods
Alam et al. [3]	Intelligent System for Vehicles Number Plate Detection and Recognition Using Convolutional Neural Networks	Smart city, traffic monitoring, license plate detection	Computer vision, convolutional neural networks (CNNs)
Dake et al. [5]	Multi-Agent Reinforcement Learning Framework in SDN-IoT for Transient Load Detection and Prevention	Software-defined networks–IoT, network optimization, DDoS protection	Deep reinforcement learning, multi-agent reinforcement learning
Kontopanagou et al. [8]	A Framework for Exploring Churches/Monuments/Museums of Byzantine Cultural Influence Exploiting Immersive Technologies in Real-Time Networked Environments	Cultural heritage, dissemination, digitization	Immersive technologies (AR/MR), cloud/fog computing, image recognition (KNN, SVM, BPNN, CNN)
Melman et al. [11]	An Authorship Protection Technology for Electronic Documents Based on Image Watermarking	Authorship protection	Image watermarking
Kaimara et al. [13]	Waking Up In the Morning (WUIM): A Smart Learning Environment for Students with Learning Difficulties	Smart learning environments, inclusive education, special educational needs	Multimedia, transmedia, VR, AR
Papavlasopoulos et al. [15]	Entropy as a Transitional In-Game Variable	Video games, immersive environments	Interpretative phenomenological analysis (IPA)
Leonidis et al. [16]	CognitOS Board: A Wall-Sized Board to Support Presentations in Intelligent Environments	Education, interactive board	Ambient intelligence, interactive displays

Table 1. Overview of published articles, their subjects, and employed technologies and methods.

Authors-Ref.	Title	Domain/Purpose	Technologies/Methods
Alam et al. [3]	Intelligent System for Vehicles Number Plate Detection and Recognition Using Convolutional Neural Networks	Smart city, traffic monitoring, license plate detection	Computer vision, convolutional neural networks (CNNs)
Dake et al. [5]	Multi-Agent Reinforcement Learning Framework in SDN-IoT for Transient Load Detection and Prevention	Software-defined networks–IoT, network optimization, DDoS protection	Deep reinforcement learning, multi-agent reinforcement learning
Kontopanagou et al. [8]	A Framework for Exploring Churches/Monuments/Museums of Byzantine Cultural Influence Exploiting Immersive Technologies in Real-Time Networked Environments	Cultural heritage, dissemination, digitization	Immersive technologies (AR/MR), cloud/fog computing, image recognition (KNN, SVM, BPNN, CNN)
Melman et al. [11]	An Authorship Protection Technology for Electronic Documents Based on Image Watermarking	Authorship protection	Image watermarking
Kaimara et al. [13]	Waking Up In the Morning (WUIM): A Smart Learning Environment for Students with Learning Difficulties	Smart learning environments, inclusive education, special educational needs	Multimedia, transmedia, VR, AR
Papavlasopoulos et al. [15]	Entropy as a Transitional In-Game Variable	Video games, immersive environments	Interpretative phenomenological analysis (IPA)
Leonidis et al. [16]	CognitOS Board: A Wall-Sized Board to Support Presentations in Intelligent Environments	Education, interactive board	Ambient intelligence, interactive displays

3. Conclusions

Though varied, the works collected in this Special Issue highlight potential aspects of smart environments and showcase the application of various approaches, both technological and methodological, in the field. Some of these works [3,5,8] demonstrate the importance of machine learning in the context of smart environments. Some studies focus on smart city applications [3], while others focus on cultural heritage environments [8] or the network infrastructure that supports smart environments [5]. Information security is another major issue in smart environments, especially in e-government applications, as discussed by [11]. Immersive technologies are valuable tools for promoting user engagement with smart environments and effectively communicating information, as shown by [8,13]. Moreover, as shown in [13], smart environments can be adapted for educational purposes and to promote inclusivity. Interaction is a vital component in the use of applications within smart environments, and new methodological approaches can be tested to provide improved user experiences, as demonstrated by [15]. Finally, an interesting approach to smart collaborative environments, applicable in both educational and business contexts, is presented by [16]. In closing, these works illustrate the broad impact and versatility of smart environments, demonstrating how varied approaches can address unique challenges, provide valuable services, and enrich user experiences across multiple domains.