A Survey of Methods and Architectures for Joint Resource Allocation, Security, and Efficient Task Scheduling in Cloud Computing Using Hybrid Pyramidal Convolution Split-Attention Networks

Abstract

Cloud computing has revolutionized modern computing by enabling scalable, on-demand access to computational resources, storage, and services. However, the rapid growth of cloud infrastructures has introduced significant challenges in resource allocation, task scheduling, and security management. Efficient resource allocation ensures optimal utilization of computing resources such as CPU, memory, and bandwidth, while task scheduling determines the execution order and placement of tasks across distributed cloud environments. These problems are inherently complex due to dynamic workloads, heterogeneous resources, and multi-objective constraints including latency, cost, and energy efficiency. Traditional approaches based on heuristic and metaheuristic algorithms have been widely used to address these challenges. However, such approaches often lack adaptability and predictive capabilities in highly dynamic cloud environments. Recent research has shifted towards machine learning and deep learning–based frameworks that can analyse historical workload patterns and predict future resource demands. In particular, attention-based neural architectures and pyramidal convolution networks have shown strong potential in capturing multi-scale system behaviours and improving scheduling decisions. This survey provides a comprehensive review of recent methods and architectures for joint resource allocation, security, and efficient task scheduling in cloud computing, with a focus on hybrid models integrating pyramidal convolution and split-attention mechanisms. These architectures enable hierarchical feature extraction and intelligent decision-making, improving system performance and scalability. Additionally, the study highlights the importance of integrating security-aware mechanisms such as anomaly detection and intrusion prevention within scheduling frameworks to ensure safe and reliable cloud operations.