The edge computing paradigm enables mobile devices with limited memory and processing power to execute delay-sensitive, compute-intensive, and bandwidth-intensive applications on the network by bringing the computational power and storage capacity closer to end users. Edge computing comprises heterogeneous computing platforms with resource constraints that are geographically distributed all over the network. As users are mobile and applications change over time, identifying an optimal task scheduling method is a complex multi-objective optimization problem that is NP-hard, meaning the exhaustive search with a time complexity that grows exponentially can solve the problem. Therefore, various approaches are utilized to discover a good solution for scheduling the tasks within a reasonable time complexity, while achieving the most optimal solution takes exponential time. This study reviews task scheduling algorithms based on centralized and distributed methods in a three-layer computing architecture to identify their strengths and limitations in scheduling tasks to edge service nodes.

Compressors are fundamental components in multi‐operand addition and multiplication hardware. The present study aims to propose several high‐speed hybrid designs of 4:2 compressors which are implemented based on multi‐threshold logic and transmission gate multiplexers, at low supply voltages. In order to implement these circuits, carbon nanotube field‐effect transistors (CNFETs) are utilised. The division of the entire input capacitor network in the third proposed design causes the entire structure operates faster and consumes less power. All designs were simulated with Synopsys HSPICE and CNFET technology in different conditions. Simulation results demonstrate the superiority of the third proposed design with regard to different load and frequency conditions. In addition, it becomes less sensitive to process, voltage and temperature variations compared to other similar capacitive threshold logic circuits. The results indicate the third proposed design outperforms the best reference design in terms of delay and power‐delay product by 34 and 25%, respectively. Furthermore, the performance of each design was evaluated in a 8 × 8‐bit binary multiplier as a uniform test bench, which confirmed the above results on isolated single 4:2 compressors.

Soft real-time applications are becoming increasingly complex, posing significant challenges for scheduling offloaded tasks in edge computing environments while meeting task timing constraints. Moreover, the exponential growth of the search space, presence of multiple objectives and parameters, and highly dynamic nature of edge computing environments further exacerbate the complexity of task scheduling. As a result, schedulers based on heuristic and metaheuristic algorithms frequently encounter difficulties in generating optimal or near-optimal task schedules due to their constrained ability to adapt to the dynamic conditions and complex environmental characteristics of edge computing. Accordingly, reinforcement learning algorithms have been incorporated into schedulers to address the complexity and dynamic conditions inherent in task scheduling in edge computing. However, a significant limitation of reinforcement learning algorithms is the prolonged learning time required to adapt to new environments and to address medium- and large-scale problems. This challenge arises from the extensive global action space and frequent random exploration of irrelevant actions. Therefore, this study proposes Agile Reinforcement learning (aRL), in which the RL-agent performs informed exploration and executes only relevant actions. Consequently, the predictability of the RL-agent is enhanced, leading to rapid adaptation and convergence, which positions aRL as a suitable candidate for scheduling the tasks of soft real-time applications in edge computing. The experiments demonstrate that the combination of informed exploration and action-masking methods enables aRL to achieve a higher hit-ratio and converge faster than the baseline approaches.

Soft real-time applications are becoming increasingly complex, posing significant challenges for scheduling offloaded tasks in edge computing environments while meeting task timing constraints. Moreover, the exponential growth of the search space, presence of multiple objectives and parameters, and highly dynamic nature of edge computing environments further exacerbate the complexity of task scheduling. As a result, schedulers based on heuristic and metaheuristic algorithms frequently encounter difficulties in generating optimal or near-optimal task schedules due to their constrained ability to adapt to the dynamic conditions and complex environmental characteristics of edge computing. Accordingly, reinforcement learning algorithms have been incorporated into schedulers to address the complexity and dynamic conditions inherent in task scheduling in edge computing. However, a significant limitation of reinforcement learning algorithms is the prolonged learning time required to adapt to new environments and to address medium- and large-scale problems. This challenge arises from the extensive global action space and frequent random exploration of irrelevant actions. Therefore, this study proposes Agile Reinforcement learning (aRL), in which the RL-agent performs informed exploration and executes only relevant actions. Consequently, the predictability of the RL-agent is enhanced, leading to rapid adaptation and convergence, which positions aRL as a suitable candidate for scheduling the tasks of soft real-time applications in edge computing. The experiments demonstrate that the combination of informed exploration and action-masking methods enables aRL to achieve a higher hit-ratio and converge faster than the baseline approaches.