Unmanned aerial vehicles (UAVs) play an important role in mobile edge computing (MEC) systems because of their high mobility and flexibility. However, existing task offloading strategies suffer from considerable resource allocation imbalances in multi-agent decision-making, leading to UAVs overload, increased task delays, and higher operational costs. To address these issues, this paper presents a Three-Layer Multi-Agent Strategic Decision-Making Model (3L-MSADM) that integrates Markov Decision Processes (MDP), Stackelberg game theory, and auction mechanisms to optimize task offloading, mitigate resource imbalances, and enhance computational efficiency. Additionally, a task offloading ratio optimization mechanism is proposed to dynamically adjust task distribution according to system load, thereby minimizing task latency and improving overall efficiency. Furthermore, we introduce the Stackelberg-guided multi-agent policy gradient (SG-MAPG) algorithm, utilizing a centralized training and decentralized execution (CTDE) paradigm to improve decision-making efficiency and service fairness. Simulation results demonstrate that our approach improves service fairness by 12.3% and reduces system costs by 22.7% compared to benchmark algorithms, significantly enhancing the task processing capabilities of UAVs-assisted MEC systems. This study provides an innovative solution for multi-agent decision-making and resource management in wireless networks, offering substantial theoretical and practical contributions.

Kubernetes provides powerful container orchestration features, like auto-scaling, which dynamically adjusts the scale of containerized applications. However, the default auto scaling mechanism in Kubernetes typically responds to workload changes only after they have occurred, which can result in resource mismatches. While there has been significant research into proactive auto-scaling approaches, most existing solutions struggle to adapt to the constantly evolving characteristics of prediction targets. This paper presents PAHPA, an intelligent, prediction-assisted approach aimed at enhancing Kubernetes' autoscaling capabilities, which integrates an prediction model, SLMD-LightGBM proposed in this paper, featuring a self updating mechanism for continuous prediction optimization, queueing theory analysis to improve resource allocation decisions, and a correction mechanism that refines prediction metrics using real-time data. A central insight of this paper is that while predictions are crucial, they should not be the sole basis for decisions and must be complemented by real-time monitoring data to ensure robust and adaptive autoscaling. Experimental results demonstrate that PAHPA significantly enhances system stability and reduces service latency, achieving a 9.5% lower Violation Rate (16.3%) than HPA (25.8%) and a 63% reduction in 99th percentile latency (880ms vs. 2400ms). This research highlights how combining predictive analytics with real-time monitoring can lead to more effective autoscaling strategies in cloud-native environments.

Cyber-physical systems (CPSs) realize the automatic control of entities through computing systems and networks. Stability is an important factor in CPS for system upgrading and troubleshooting. Traditional analysis methods focus on simulation and formal analysis, which have two major limitations: first, the current state information of CPS is difficult to obtain; second, most CPS face the state space explosion problem. These problems can be avoided and a good analysis can be provided based on empirical data. The main work of this article is summarized as follows: first, a phase space reconstruction method is designed to divide the dataset into several subsequences with the same shape; second, we propose a stability analysis method based on correlation dimensions. Results indicate that the proposed approach can obtain a stable correlation dimension. CPS perform better if the correlation dimension is maintained within a certain range; otherwise, a destabilizing factor exists. The proposed stability analysis has less complexity and running time.

In this paper, the joint optimization problem with energy efficiency and effective resource utilization is investigated for heterogeneous and distributed multi-core embedded systems. The system model is considered to be fully a heterogeneous model, that is, all nodes have different maximum speeds and power consumption levels from the perspective of hardware while they can employ different scheduling strategies from the perspective of applications. Since the concerned problem by nature is a multi-constrained and multi-variable optimization problem in which a closed-form solution cannot be obtained, our aim is to propose a power allocation and load balancing strategy based on Lagrange theory. Furthermore, when the problem cannot be fully solved by Lagrange approach, a data fitting method is employed to obtain core speed first, and then load balancing schedule is solved by Lagrange method. Several numerical examples are given to show the effectiveness of the proposed method and to demonstrate the impact of each factor to the present optimization system. Finally, simulation and practical evaluations show that the theoretical results are consistent with the practical results. To the best of our knowledge, this is the first work that combines load balancing, energy efficiency, hardware heterogeneity and application heterogeneity in heterogeneous and distributed embedded systems.