Parkinson's Disease (PD) significantly affects motor functions, with gait disturbances being a signal symptom. This study leverages biomedical information fusion to enhance PD diagnosis through gait analysis. Using a 6-axis inertial measurement unit (IMU) mounted on insoles, we collected a comprehensive dataset comprising over 2,000 samples and 164 gait features from PD patients and healthy controls. A novel sequential feature selection strategy was developed, reducing redundancy and identifying 41 critical gait metrics, enabling efficient and interpretable analysis. An adaptive voting ensemble model was proposed, effectively addressing patient heterogeneity and achieving superior predictive performance with an AUC of 0.9055. This method has the potential to optimize wearable systems to facilitate real-time monitoring and early interventions for PD.
The Internet of Things (IoT) has become a fundamental component of modern cyber-physical systems; however, securing IoT deployments remains challenging due to strict constraints on computation, memory, and energy resources. Conventional cryptographic mechanisms provide strong security guarantees but often introduce excessive overhead for low-power embedded devices, whereas lightweight cryptographic solutions may reduce security robustness under dynamic operating conditions. This paper presents an adaptive hybrid cryptographic framework for secure and energy-efficient IoT systems. The proposed framework combines lightweight cryptographic primitives, hybrid key establishment, and machine-learning-assisted context-aware security selection within a unified adaptive architecture. Cryptographic configurations are dynamically selected according to runtime resource availability, data sensitivity, and inferred threat conditions while preserving bounded adaptation overhead and lightweight deployment characteristics. Experimental evaluation across representative 8-bit, 16-bit, and 32-bit IoT platforms demonstrates that the proposed framework reduces energy consumption and execution latency by approximately 15-25% compared with static cryptographic configurations while maintaining strong security guarantees and improved operational scalability. Additional adversarial validation further demonstrates resistance against downgrade manipulation, replayed context signals, oscillation-triggering attacks, and unauthorized reconfiguration attempts. The results indicate that adaptive hybrid cryptographic selection provides a practical balance between security robustness, runtime efficiency, and deployment scalability for heterogeneous IoT environments.
Container based microservice architecture is becoming the priority of service development in edge computing. Its flexibility and scalability could provide stable service response. However, due to the limited computing resources of each edge node in edge computing, how to use computing resources efficiently has become an important issue in microservice deployment. This study formulates the multi-objective microservice deployment problem (MMDP) in Kubernetes, balancing service latency and resource utilization. We propose a deep reinforcement learning method with reward shaping and heuristic scaling to derive optimal deployment policies. Experiments show that our approach reduces response time by 25-30% and improves load balancing by 20% compared with Kubernetes default scheduling and DQL. These results demonstrate the applicability of reinforcement learning for distributed computing and automatic control in microservice systems.
The rapid growth of large-scale medical imaging datasets has exposed critical limitations of single-node deep learning workflows for clinical decision support in lung disease diagnosis. We present a unified, Spark-based distributed framework for scalable multi-class lung disease classification from both chest X-ray (CXR) and computed tomography (CT) images. Our approach integrates Apache Spark with Elephas to enable synchronous and asynchronous data-parallel training of Keras models via a parameter-server architecture on a GPU-enabled cluster, supporting binary and multi-class classification across multiple pulmonary pathologies including COVID-19 pneumonia, viral pneumonia, lung opacity, and normal cases. We implement transfer learning with DenseNet169, ResNet50, and MobileNetV2 and evaluate three classification tasks: binary lung disease detection and three-class diagnosis on CXR images, and three-class diagnosis on CT scans. Experiments on large public datasets demonstrate that the proposed system achieves high diagnostic performance, reaching 97.63% accuracy for three-class CXR and 95.49% for binary CT classification, while providing substantial runtime gains over single-node baselines. On a multi-node Spark cluster, synchronous training attains near-linear to super-linear scaling with speedups up to 6.26 & times; and parallel efficiency exceeding 156%,attributed to improved cache utilization and reduced I/O contention across distributed workers. Comparative analysis consistently demonstrates that synchronous parameter updates outperform asynchronous training in both convergence stability and final diagnostic accuracy across all tasks and modalities. This work delivers an end-to-end, cross-modality, distributed deep learning pipeline with explicit scalability characterization. The proposed framework is generalizable beyond pulmonary imaging and demonstrates how cluster computing infrastructures can be leveraged to build scalable, high-throughput experimental pipelines for multi-class lung disease classification in research settings.
This paper presents the design and implementation of a robust adaptive control strategy for a Dual Active Bridge (DAB) converter used in More Electric Aircraft (MEA) applications. In MEA electrical architectures, a 270 V DC main bus must reliably supply regulated 28 V DC power for onboard systems such as avionics and battery subsystems. In response to this demand, a new controller based on a Fractional-Order Adaptive Super-Twisting Algorithm (FO-ASTA) integrated with a Model Reference Adaptive Control (MRAC) framework is proposed. The controller enhanced the robustness and chattering-free properties of super-twisting sliding mode control, the dynamic memory effects of fractional-order systems, and the real-time adaptability of MRAC.The proposed FO-ASTA controller integrated with the MRAC scheme is evaluated under steady-state conditions, load variations, reference changes, and parametric uncertainties. Moreover, its performance is compared with the Fractional-Order Sliding Mode Controller (FOSMC) and Fractional-Order Super-Twisting Algorithm (FOSTA) controllers. Simulation results on the circuit level confirm that the proposed controller outperforms the alternatives in terms of tracking accuracy, convergence speed, disturbance rejection, and control smoothness. This makes it a strong candidate for high-performance DC-DC conversion in next-generation MEA power systems.
