Robust federated learning for cloud environments using evolutionary optimization and blockchain

The convergence of vehicular technology, artificial intelligence (AI), and distributed computing has catalyzed the emergence of the Internet of Vehicles (IoV) as a cornerstone of next-generation intelligent transportation systems (ITS). By enabling vehicle-to-everything (V2X) communication, IoV supports cooperative perception, real-time decision-making, and autonomous driving. However, the reliance on large-scale, data-driven intelligence in IoV exposes systems to critical challenges, including adversarial poisoning, privacy leakage, identity forgery, and the fragility of centralized learning architectures. Federated Learning (FL) has been proposed as a promising paradigm to alleviate some of these issues by enabling distributed model training without centralizing sensitive vehicular data. Nonetheless, conventional FL remains vulnerable to security and trust limitations, particularly in dynamic vehicular environments. This thesis addresses these challenges by designing secure, privacy-preserving, and scalable FL frameworks that leverage distributed ledger technologies and cutting-edge security mechanisms.The thesis advances knowledge through four interconnected contributions. First, two novel optimization-driven poisoning attack models are introduced: PA-PSOSA and PAPSOGA, which combine particle swarm optimization with simulated annealing and genetic algorithms, respectively. These models demonstrate that even a small poisoning budget can substantially degrade global model utility under black-box and clean-label constraints, highlighting the urgency of robust defenses in vehicular FL. Second, a permissioned blockchain-enabled FL (BCFL) framework is proposed, in which consortium edge nodes running Practical Byzantine Fault Tolerance (PBFT) consensus replace the central aggregator. With blockchain integration and data validation mechanisms, this design ensures identity authentication, verifiable audit trails, and improved resilience against poisoning and Sybil attacks, while maintaining high model accuracy under adversarial conditions. Third, the framework is further enhanced to achieve inference-resistance by integrating secure aggregation (SecAgg) and differential privacy (DP), and lightweight with off-chain commitments. This design significantly reduces ledger storage requirements, increases system throughput, and mitigates inference-based privacy risks. Finally, to overcome the scalability limitations of PBFT-based BCFL, a DAG-enabled FL (DFL) framework is developed. By leveraging parallel validation, utility-score-based tip selection, and reputation-weighted aggregation, this framework significantly improves scalability, reduces communication complexity, and enhances robustness in asynchronous vehicular environments.Together, these contributions articulate a coherent progression from exposing vulnerabilities in vehicular FL to constructing secure, privacy-preserving, and scalable frameworks tailored for IoV ecosystems. The findings demonstrate that interdisciplinary integration of optimization theory, cryptography, differential privacy, and distributed ledger technologies is indispensable for trustworthy vehicular intelligence. Beyond theoretical significance, the proposed frameworks offer practical designs for deployment in safety-critical IoV environments. Future research directions include the integration of zero-knowledge proofs (ZKP) for verifiable privacy, adaptive defenses against evolving adversarial strategies, and experimental validation in real-world vehicular testbeds. Collectively, this thesis establishes a foundation for secure federated intelligence in IoV, contributing to the reliability, efficiency, and trustworthiness of next-generation ITS.

An improved blockchain-based multi-region Federated Learning framework for crop disease diagnosis

Within the federated learning (FL) framework, the client collaboratively trains the model in coordination with a central server, while the training data can be kept locally on the client. Thus, the FL framework mitigates the privacy disclosure and costs related to conventional centralized machine learning. Nevertheless, current surveys indicate that FL still has problems in terms of communication efficiency and privacy risks. In this paper, to solve these problems, we develop an FL framework with communication-efficient and privacy-preserving (FLCP). To realize the FLCP, we design a novel compression algorithm with efficient communication, namely, adaptive weight compression FedAvg (AWC-FedAvg). On the basis of the non-independent and identically distributed (non-IID) and unbalanced data distribution in FL, a specific compression rate is provided for each client, and homomorphic encryption (HE) and differential privacy (DP) are integrated to provide demonstrable privacy protection and maintain the desirability of the model. Therefore, our proposed FLCP smoothly balances communication efficiency and privacy risks, and we prove its security against “honest-but-curious” servers and extreme collusion under the defined threat model. We evaluate the scheme by comparing it with state-of-the-art results on the MNIST and CIFAR-10 datasets. The results show that the FLCP performs better in terms of training efficiency and model accuracy than the baseline method.

The educational resource management based on image data visualization and deep learning

As the Augmented Intelligence of Things (AIoT) advances within vehicle-road coordination systems, challenges related to road traffic data transmission and processing are being increasingly addressed. However, this progress also brings significant risks of privacy data leakage. Federated learning (FL), a distributed machine learning paradigm, effectively safeguards client data privacy by allowing multiple participants to collaboratively train models while keeping their data localized. Despite its benefits, FL faces challenges, such as model parameter leakage and Byzantine attacks. To tackle these issues, this article introduces an anonymous and efficient FL framework for vehicle-road coordination systems (AEFL), designed to ensure a secure and reliable vehicle data transmission process. This architecture incorporates a novel group pairing onion routing protocol, which leverages pairing cryptography principles for hierarchical data encryption. During the routing process, relay group nodes decrypt the corresponding layer, ensuring both data confidentiality and node anonymity. Additionally, a sampling method is proposed to accurately identify Byzantine vehicle nodes, enhancing the precision of FL aggregation without compromising overall model performance. Experimental results show that AEFL outperforms the classic TOR anonymous routing protocol, achieving a 100% message delivery rate more quickly. Under the same conditions, the anonymity of the source node and the destination node improves by 3.9% and 1.9%, respectively. When half of the nodes are compromised, path anonymity can be increased by 24.8%. Furthermore, our framework excels in FL aggregation efficiency, with a Byzantine adversary detection accuracy of up to 99%.

Trustiness-based hierarchical decentralized federated learning

In the Internet of Things (IoT) field, where interconnected devices generate sensitive data, ensuring privacy is a major challenge. Federated Learning (FL) addresses this by allowing devices to collaboratively train a model without sharing their local data, improving privacy in IoT systems. While the traditional two-layer FL framework is commonly used, adopting a hierarchical client-edge-cloud architecture can significantly accelerate model training, especially in resource-constrained IoT networks. Hierarchical Federated Learning (HFL) offers significant advantages, yet concerns persist regarding potential privacy breaches from analyzing client or edge server data. To address these concerns, we propose PrivHFL, a privacy-preserving solution for HFL that leverages threshold homomorphic encryption. Security and performance analyses demonstrate that the proposed scheme is scalable, supporting larger FL scenarios, including diverse IoT environments, while ensuring data privacy. PrivHFL is resilient to collusion among nearly half of the clients and effectively handles client dropouts. Our approach achieves high accuracy in IID and non-IID scenarios, as demonstrated using the MNIST, CIFAR-10, and CIFAR-100 datasets. Additionally, we show that the added encryption overhead is reasonable, making our solution feasible for real-world IoT applications.

In Industry 4.0, factories encounter significant challenges in making informed decisions to maintain or enhance their industry standing. By utilizing machine learning (ML), they can improve product quality, decrease production downtime, and boost operational efficiency. However, acquiring datasets with sufficient variation and diversity to train a robust neural network centrally is a challenge within the industrial sector. Consequently, federated learning (FL) offers a decentralized approach that safeguards data privacy, enabling smart infrastructures to train collaborative models locally and independently while retaining local data. In this paper, we present FedGA-Meta framework, which combines FL, meta-learning, and domain adaptation to enhance model performance and generalizability, particularly when training across distributed factories with varying network and data conditions. The results obtained demonstrate the effectiveness and efficiency of our FedGA-Meta framework.

The goal of production scheduling is to achieve a profitable balance among on-time delivery, short customer lead time, and maximum utilization of resources. However, current practices in precast production scheduling are fairly basic, depending heavily on experience, thereby resulting in inefficient resource utilization and late delivery. Certain computational techniques have been proven effective in scheduling. To enhance precast production scheduling, this research develops a Multi-Objective Precast Production Scheduling Model (MOPPSM). In the model, production resources and buffer size between stations are considered. A multi-objective genetic algorithm is then developed to search for optimum solutions with minimum makespan and tardiness penalties. The performance of the proposed model is validated by using five case studies. The experimental results show that the MOPPSM can successfully search for optimum precast production schedules. Furthermore, considering buffer sizes between stations is crucial for acquiring reasonable and feasible precast production schedules.

Quite recently, the community has adopted the Cloud-native concept, building application workflows based on microservices. The majority of these services are deployed as containers. Unikernels, which package applications within a single-address space library OS, have been proposed as a stronger isolation mechanism. However, due to different storage semantics between them, the state-of-the-art approaches for using unikernels in place of containers result in decreased performance, inefficient resource utilization and limited functionality. In this work we bridge the storage gap by designing a framework, which extends the Docker storage layer to support unikernel requirements. We show that our framework improves boot time while reducing storage space requirements.

As the ongoing evolution in basic education continues to deepen, primary school mathematics classrooms are undergoing a paradigm shift from "knowledge transmission" to "competency development". This paper focuses on the core concept of "dynamic classrooms", identifies the practical challenges of lacking vitality in traditional classrooms through literature analysis, and constructs a three-dimensional implementation framework encompassing subject activation, process reconstruction, and environment optimization based on the zone of proximal development theory. The study reveals that current classrooms suffer from superficial student engagement, monotonous teaching methods, and inefficient resource utilization. By integrating "problem-driven learning, project-based learning, and multi-resource integration" strategies, the depth of classroom interaction and learning effectiveness can be significantly enhanced. The research findings offer theoretical insights and actionable approaches for transforming primary school mathematics classrooms, facilitating the development of a high-quality classroom ecosystem centered on student growth.

Smart emergency hospitals constitute a transformative model that addresses the continual challenges of traditional healthcare systems, together with delayed reaction times, elevated clinical errors, and inefficient resource utilization via the combination of superior technologies, which include Artificial Intelligence (AI), the Internet of Things (IoT), and scientific robotics. Despite the growing body of research on medical and administrative benefits, confined interest has been paid to the architectural and spatial implications of this technology. This examines objectives to discover how clever technologies influence the architectural layout of emergency hospitals, particularly in terms of spatial configuration, operational flow, environmental excellence, and versatility. By using a descriptive-analytical technique and referencing benchmark case studies, which include the Beijing Medical Innovation Hospital and Shanghai Smart Hospital, the studies evaluate performance signs based totally on a comparative framework with traditional hospitals of comparable size and function. Key findings screen that smart hospitals have achieved a seventy 5% discount in common emergency response time (from 20 to 5 mins), a 25% development in diagnostic accuracy via AI-assisted imaging, and a 20% increase in useful resource allocation efficiency due to automatic garage and logistics structures. Moreover, operational costs have been reduced by about 33% as compared to standard structures, normally due to automation and virtual optimization. The research concludes that the integration of smart technologies no longer best improves operational performance and affects person pleasure but additionally necessitates a rethinking of emergency health center design to aid adaptive, information-driven, and technology-responsive areas. This highlights the significance of aligning architectural layout with digital transformation techniques and making sure non-stop team of workers schooling and overall performance assessment for lengthy-time period sustainability.

The proliferation of IoT devices and the adoption of Fog computing architectures have transformed data processing and real-time decision-making across various domains. These advancements enable seamless connectivity and distributed computational power, fostering the development of more intelligent systems. However, managing large-scale IoT and Fog networks presents critical challenges, including high latency, inefficient resource utilization, and scalability limitations, which can undermine system performance. To address these challenges, this research proposes an innovative framework combining Q-Learning and Deep Extreme Learning Machine (DELM). Q-Learning optimizes resource allocation by intelligently learning and adapting to dynamic network conditions, ensuring efficient utilization of resources. It enhances decision-making processes by identifying optimal strategies to manage complex IoT and Fog environments. Meanwhile, DELM provides high-speed and accurate data processing capabilities, enabling it to handle the intensive computational demands of large-scale networks. By leveraging the complementary strengths of these methods, the framework aims to enhance latency, resource utilization, and scalability in large-scale environments. Extensive experimental evaluations validate the framework’s effectiveness, demonstrating significant reductions in latency, improved computational efficiency, and enhanced throughput. Furthermore, the framework efficiently handles complex data processing tasks with minimal overhead, making it suitable for diverse real-time applications across IoT and Fog systems. This study highlights the transformative potential of the proposed approach, offering high performance and real-time efficiency for complex, large-scale IoT and Fog computing environments.

Global agrifood systems face three interconnected challenges: ensuring food security, promoting environmental sustainability, and restoring soil health in the face of climate change. Conventional practices have prioritized productivity over ecological resilience, leading to soil degradation, increased greenhouse gas (GHG) emissions, and inefficient resource utilization. Here, we introduce a “triple-goal” agrifood framework that enhances food production, soil health, and GHG mitigation simultaneously through integrated innovations. Using a second-order meta-analysis of 104 meta-analyses that cover 39,162 studies and 300,139 global field comparisons, we identified key interventions, including optimized fertigation, diversified cropping systems, organic amendments, and precision N management, that increased productivity by 14%–28% while reducing environmental impacts. Diversified systems boosted yields by 19.6% and reduced land use by 19%. Integrating legumes and cover crops lowered N2O emissions by 18%–65%, while organic amendments increased soil organic carbon stocks by 7%–13%. Structural equation modeling identified nitrogen use efficiency and microbial activity as central to the food-soil-emissions nexus. However, tradeoffs remain; yield-focused strategies can elevate emissions if not tailored to local conditions. By integrating agronomic, biological, and technological interventions such as conservation tillage, biofertilization, and digital agriculture, this triple-goal framework supports a 15%–30% reduction in anthropogenic CO2-equivalent emissions. These findings underscore the need for policy reform and multi-stakeholder collaboration to scale up the adaptation of integrated strategies in alignment with the UN’s Sustainable Development Goals and the “One Health” initiative. The triple-goal framework provides a transformative pathway to climate-smart, equitable, and resilient agrifood systems that strike a balance between productivity and planetary health.

Automated guided vehicle (AGV) is widely used in transportation and distribution of materials. In the development of multi-AGV systems, how to handle deadlocks is a core issue. Most existing researches focus on open systems. Thus those results are not applicable in real-world systems. A few works have paid attention to closed systems recently, but the results are not flexible and generic enough. Therefore, an improved deadlock avoidance algorithm in closed guide-path system is proposed in this paper. The proposed algorithm is divided into an offline stage and an online stage. At the offline stage, the guide-path graph is processed to obtain useful information for the online stage. At the online stage, the algorithm assesses the safety of each resource allocation to ensure a deadlock-free operation. The computational complexity of the algorithm is shown to be <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$O(|E|)$ </tex-math></inline-formula> , where <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$E$ </tex-math></inline-formula> is the set of the edges in the guide-path graph. Based on the algorithm, an overall control strategy is also developed, and then implemented and tested in real systems. It is also compared with Banker’s algorithm and its variants, the results of which show that the proposed algorithm has a better performance. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —Deadlock is a core issue in guide-path based multi-AGV systems. Once deadlock occurs, the whole system will collapse. This problem is more serious in closed systems, which is paid less attention to in literature. Current deadlock solutions are either too restrictive, resulting in inefficient resource utilization and poor flexibility, or computationally intensive and thus are unable to scale to large systems. To solve this problem, an improved deadlock avoidance algorithm with polynomial computational complexity is proposed in this paper, and it is more flexible than the Banker’s algorithm and its variants. Furthermore, an overall control framework (i.e., including task assignment, path planning and real-time scheduling) is developed based on the deadlock avoidance algorithm, which keeps persistent operation of the considered system while still guaranteeing deadlock-free. However, in this overall control framework, the choice of path is not optimized with respect to travel distance and makespan, which is the direction of future research.