Robust brain-tumor segmentation in MRI must withstand domain shifts from heterogeneous scanners and protocols. Unsupervised domain adaptation (UDA) can exploit plentiful unlabeled data, yet many approaches erode boundaries, suppress tumor cues, or amplify errors from noisy pseudo-labels. Generalpurpose models like SAM also struggle on MRI due to the domain gap and a lack of morphology-aware consistency. We propose SAM-RefiSeR, a two-phase UDA framework that integrates SAM for reliable, annotation-efficient segmentation. Phase I reduces source-target discrepancy via Fourier-based frequency adaptation and adversarial feature alignment, aligning appearance and representations while preserving anatomy. Phase II adopts a student-teacher scheme in which SAM first refines pseudo-labels, then gates them with confidence- and morphology-aware criteria to suppress unreliable masks and curb error propagation. Across diverse cross-modality settings, SAM-RefiSeR consistently surpasses strong UDA baselines, particularly under severe shifts. By improving boundary fidelity and robustness without additional labels, SAM-RefiSeR brings brain-tumor segmentation closer to practical, generalizable clinical deployment.

Glaucoma is one of the leading causes of blindness worldwide, where early and precise diagnosis is critical for effective treatment. We explore the potential of deep learning to generate a Visual Field Grayscale Map (GM) directly from a Conventional Fundus Image (CFI) at the same time point, enabling an intuitive representation of visual field defects for the diagnosis of glaucoma. The Humphrey Visual Field (HVF) test is a standard method for assessing visual field defects, but its dependency on active participation and prolonged testing time limits its reliability, particularly in elderly patients. To address these challenges, we propose CFI2GM-GAN, a novel deep learning framework that directly generates GM from CFI, offering a faster and cost-effective alternative while complementing the HVF test. Our method incorporates dataset partitioning based on defect severity to mitigate class imbalance, attention-guided fusion techniques for improved prediction reliability, and a refined loss function customized to fundus image characteristics.

Conditional Handover (CHO) by the 3rd Generation Partnership Project (3GPP) enables efficient user mobility between Base Stations (BSs) by preselecting and preparing Target BSs (T-BSs). However, CHO relies on signal strength for T-BS selection, leading to resource blocking on multiple T-BSs due to signal fluctuations. Existing state-of-the-art methods use deep learning to narrow the list of T-BSs but still lack an effective method for resource reservation timing. This paper presents in-time CHO (iCHO) which exploits historical mobility data to estimate user dwell time at the current BS to reduce resource reservation duration. The proposed iCHO employs a Multivariate Multi-output Single-step Prediction (MMSP) model that leverages a multi-task learning approach to simultaneously predict the minimal list of required T-BSs together with the user dwell time. The model demonstrates remarkable performance across two mobility datasets of different scales, achieving T-BS prediction accuracies of 98% and 95%. It also ensures a 100% handover success rate with a minimum of three and four predicted T-BSs for both datasets, respectively, significantly limiting the list of T-BSs. Moreover, the MMSP model achieves a Mean Absolute Error (MAE) of 19 s and 45 s when predicting the user’s dwell time at the current BS. By utilizing these predictions, iCHO reserves resources at the minimum number of T-BSs immediately before handover. Thus, iCHO can save up to 99% of resources from blockage as compared to the CHO, enabling operators to increase revenue by serving up to eighteen more users with the saved resources.

An automated analysis of respiratory sounds using Deep Learning (DL) plays a pivotal role in the early detection of lung diseases. However, current DL methods often examine the spatial and temporal characteristics of respiratory sounds in isolation, which inherently limit their potential. This study proposes a novel DL framework that captures spatial features through convolution operations and exploits the spatiotemporal correlations of these features using temporal convolution networks. The proposed framework incorporates Multi-Level Temporal Convolutional Networks (ML-TCN) to considerably enhance the model accuracy in detecting anomaly breathing cycles and respiratory recordings from lung sound audio. Moreover, a transfer learning technique is also employed to extract semantic features efficiently from limited and imbalanced data in this domain. Thorough experiments on the well-known ICBHI 2017 challenge dataset show that the proposed framework outperforms state-of-the-art methods in both binary and multi-class classification tasks for respiratory anomaly and disease detection. In particular, improvements of up to 2.29% and 2.27% in terms of the Score metric, average sensitivity and specificity, are demonstrated in binary and multi-class anomaly breathing cycle detection tasks, respectively. In respiratory recording classification tasks, the classification accuracy is improved by 2.69% for healthy-unhealthy binary classification and 1.47% for healthy, chronic, and non-chronic diagnosis. These results highlight the marked advantage of the ML-TCN over existing techniques, showcasing its potential to drive future innovations in respiratory healthcare technology.

Mobile traffic data in urban regions shows differentiated patterns during different hours of the day. The exploitation of these patterns enables highly accurate mobile traffic prediction for proactive network management. However, recent Deep Learning (DL) driven studies have only exploited spatiotemporal features and have ignored the geographical correlations, causing high complexity and erroneous mobile traffic predictions. This paper addresses these limitations by proposing an enhanced mobile traffic prediction scheme that combines the clustering strategy of daily mobile traffic peak time and novel multi Temporal Convolutional Network with a Long Short Term Memory (multi TCN-LSTM) model. The mobile network cells that exhibit peak traffic during the same hour of the day are clustered together. Our experiments on large-scale real-world mobile traffic data show up to 28% performance improvement compared to state-of-the-art studies, which confirms the efficacy and viability of the proposed approach.

Data aggregation is an important approach in IoT sensor networks since it reduces data transfer while also preserving energy and bandwidth. This research investigates the challenge of time-efficient data aggregation in wireless sensor networks, which is critical in military, civilian, and industrial applications. Effective data aggregation algorithm design and optimization are required for quick and interference-free data collection. Machine learning has received attention for outperforming classical heuristic techniques. The research presents the first Graph Neural Network (GNN) model for data aggregation in IoT sensor networks, which incorporates Graph Attention Networks (GATs) and fully connected layers. The GNN-based model learns network topology and node attributes, creating node embeddings and correcting sensor node transmitting time slots. With a centralized training procedure and adapted execution for network size change, the proposed approach achieves satisfactory performance compared to the heuristic algorithm.

Fast data aggregation is crucial for facilitating critical Internet of Things services as it enables the collection of sensory data within strict volume and time constraints. Over the past decades, the data aggregation scheduling problem for minimum latency has garnered significant research attention. Existing approaches to this problem typically schedule all data transmissions based on an aggregation tree, which is constructed without secondary interference. However, such interference can introduce delays when scheduling a transmission from a node to its parent in the tree. To this end, this study proposes an approach called Active Neighbor EXploitation (ANEX) that enables sensor nodes to switch their parents by identifying active neighbors for potential connectivity, irrespective of the receivers established in the tree. Additionally, the scheme prioritizes scheduling nodes with the fewest unscheduled active neighbors, thereby allowing for more concurrent transmissions. ANEX is evaluated through theoretical analysis and extensive simulations under various scenarios. The results demonstrate that ANEX achieves up to 86% faster aggregation compared to the state-of-the-art approach while maintaining an equivalent time complexity.

Service mobility in Multi-access Edge Computing (MEC) paradigm is necessary to provide ultra-Reliable Low Latency Communications for the erratically roaming MEC users. It involves relocation of containerized application services to a strategically selected optimal edge host. During relocation, service containers are unavailable (downtime), resulting in the interruption of ongoing user sessions and increased operational expenses for the network operator. Prolonged service downtime degrades perceived quality of experience for users, and this study handles this problem by proposing a downtime-aware Policy Learning based Capped Downtime (PLCD) service mobility strategy. It exploits Deep Actor-Critic prowess for effectively deciding when and where to relocate a containerized application service while taking user mobility and MEC server resource fluctuations into account. Efficacy of the proposed PLCD strategy is confirmed through simulation experiments, and results indicate over 90% average reduction in service downtime comparing to a baseline scheme.

Over the past few years, the use of wireless sensor networks in a range of Internet of Things (IoT) scenarios has grown in popularity. Since IoT sensor devices have restricted battery power, a proper IoT data aggregation approach is crucial to prolong the network lifetime. To this end, current approaches typically form a virtual aggregation backbone based on a connected dominating set or maximal independent set to utilize independent transmissions of dominators. However, they usually have a fairly long aggregation delay because the dominators become bottlenecks for receiving data from all dominatees. The problem of time-efficient data aggregation in multichannel duty-cycled IoT sensor networks is analyzed in this paper. We propose a novel aggregation approach, named LInk-delay-aware REinforcement (LIRE), leveraging active slots of sensors to explore a routing structure with pipeline links, then scheduling all transmissions in a bottom-up manner. The reinforcement schedule accelerates the aggregation by exploiting unused channels and time slots left off at every scheduling round. LIRE is evaluated in a variety of simulation scenarios through theoretical analysis and performance comparisons with a state-of-the-art scheme. The simulation results show that LIRE reduces more than 80% aggregation delay compared to the existing scheme.

The Smart City contexts and the adoption of Industry 4.0 are being upgraded with the latest technologies throughout all systems. New data are continuously collected in huge amounts; therefore, an efficient data aggregation scheduling scheme is highly demanded. This paper addresses the minimum time aggregation scheduling problem in duty-cycled sensor networks. Existing solutions schedule the sensory data through predefined routing structures, which limit the utilization of diverse active time slots of sensors. We propose a period-driven pipeline scheduling approach, namely PDA, that simultaneously grows the aggregation tree and assigns a transmission schedule for each node being added to the tree. Particularly, this process is performed in a top-down manner. In each iteration, corresponding to a time slot, PDA uses a multi-level ranking strategy to schedule several$\langle sender, receiver \rangle$pairs, so that in a working period ahead, the possibility to pipeline as many transmissions as possible is high. Intensive simulation results show that the proposed scheme notably works better than the best known recent algorithms by having up to 35 percent shorter aggregation time, as well as having a significantly improved network throughput and time utilization.

Anomaly detection is essential to guarantee the correctness of sensory data collected from Internet of Things. The latest detection approaches only operate under one of the two general forms of short-term or long-term anomaly. In this article, we propose monotone split and conquer (MSC) scheme to tackle both anomaly forms. The proposed scheme exploits the spatial–temporal correlation between neighboring sensors to detect abnormal sensory data. MSC splits the collected data into monotonic subtrends in the training phase to establish the trend-based normal profiles for the online detection phase. To eradicate the overfitting phenomenon, we further develop a general formulation to estimate the square prediction error (SPE) control limit. Both monotone split and general formulation contribute to the advancement of MSC in terms of accuracy (ACC) and false positive rate (FPR) in the online detection phase. To evaluate the performance of MSC, we design a general anomaly model to generate artificial short-term and long-term anomalies. Numerical experiments with the Intel Berkeley Research Lab (IBRL) data set demonstrate that MSC obtains about 8% higher ACC and 5% lower FPR on average when compared to existing schemes. Remarkably, MSC requires only a few observations for anomaly detection as it is applicable to real-time systems.

Internet-of-Things (IoT) systems and their applications rely on sensory data for operating context-aware functions. To ensure reliability, these data must arrive at the base station in a timely manner. Because IoT sensors are battery powered and expected to last for years, energy consumption has always been a dominant concern. The duty cycling mechanism efficiently extends the lifespan of sensor nodes; however, it increases data aggregation time. This article studies the minimum time aggregation scheduling problem in duty-cycled sensor networks. The existing solutions route the sensory data through a predefined structure, but such structures are constructed without full awareness of collisions in the wireless medium. Therefore, in this study, the collision-resistant dynamic (CORD) scheduling approach has been proposed with an aim to provide fresh data for emerging IoT applications. The proposed approach is applicable to any initial routing structure and it dynamically changes the receiver of a transmission whenever this change can reduce aggregation time. The receiver-changing decision is made on the fly, resisting all collisions between the determined transmissions. We evaluate CORD in various simulation scenarios, and the results demonstrate that the proposed approach yields a notable improvement of up to 62% from the viewpoint of aggregation time while having comparable time complexity compared to state-of-the-art methods.

Minimum latency scheduling has arisen as one of the most crucial problems for broadcasting in duty‐cycled Wireless Sensor Networks (WSNs). Typical solutions for the broadcast scheduling iteratively search for nodes able to transmit a message simultaneously. Other nodes are prevented from transmissions to ensure that there is no collision in a network. Such collision‐preventions result in extra delays for a broadcast and may increase overall latency if the delays occur along critical paths of the network. To facilitate the broadcast latency minimization, we propose a novel approach, critical‐path aware scheduling (CAS), which schedules transmissions with a preference of nodes in critical paths of a duty‐cycled WSN. This paper presents two schemes employing CAS which produce collision‐free and collision‐tolerant broadcast schedules, respectively. The collision‐free CAS scheme guarantees an approximation ratio of (Δ − 1)T in terms of latency, where Δ denotes the maximum node degree in a network. By allowing collision at noncritical nodes, the collision‐tolerant CAS scheme reduces up to 10.2 percent broadcast latency compared with the collision‐free ones while requiring additional transmissions for the noncritical nodes experiencing collisions. Simulation results show that broadcast latencies of the two proposed schemes are significantly shorter than those of the existing methods.