Analysis and prediction of UAV-assisted mobile edge computing systems

Abstract

<abstract><p>As the demand for the internet of things (IoT) continues to grow, there is an increasing need for low-latency networks. Mobile edge computing (MEC) provides a solution to reduce latency by offloading computational tasks to edge servers. However, this study primarily focuses on the integration of back propagation (BP) neural networks into the realm of MEC, aiming to address intricate network challenges. Our innovation lies in the fusion of BP neural networks with MEC, particularly for optimizing task scheduling and processing. Firstly, we introduce a drone-assisted MEC model that categorizes computation offloading into synchronous and asynchronous modes based on task scheduling. Secondly, we employ Markov chains and probability-generation functions to accurately compute parameters such as average queue length, cycle time, throughput, and average delay in the synchronous mode. We also derive the first and second-order derivatives of the probability-generation function to support these computations. Finally, we establish a BP neural network to solve for the average queue length and latency in the asynchronous mode. Our results from the BP neural network closely align with the theoretical values obtained through the probability-generation function, demonstrating the effectiveness of our approach. Additionally, our proposed UAV-assisted MEC model outperforms the synchronous mode. Overall, our MEC scheduling approach significantly reduces latency, enhances speed, and improves throughput, with our model reducing latency by approximately 11.72$ \% $ and queue length by around 9.45$ \% $.</p></abstract>

Multi-Strategy Improved Dung Beetle Optimization Algorithm and Its Applications

“…However, MDBO still faces challenges in obtaining the theoretical optimum when solving some complex problems in a short time. In future work, on the one hand, some other novel algorithms can be combined to improve the efficiency and optimization ability of the algorithm; on the other hand, the optimized algorithm can be used to solve more complex optimization problems in reality, such as the UAV path planning, polling system [45,46], and the NP-hard problem.…”

A Sinh–Cosh-Enhanced DBO Algorithm Applied to Global Optimization Problems

“…It is determined by the following Equation (28): applications. SCDBO is not only applicable to these engineering problems, but also to solving optimization problems in edge computing and resource scheduling [24,25].…”