REVIEW OF EXISTING CONGESTION CONTROL TECHNIQUES ACROSS WIRELESS AND CLOUD COMPUTING ENVIRONMENTS

The growing appreciation and usage of cloud computing has heightened the need for a sound network-management practice, particularly under a dynamic and resource intensive workload. The current systematic literature review (SLR), in its turn, methodically explores the available approaches to congestion-control in wireless, and cloud computing network contexts, with a dual goal of identifying the gaps that exist in the current congestion-control approaches and developing a more dynamic, intelligent-based approach to the same congestion-control, which suits the needs of cloud-based systems. To this extent, this review compiles the results of available researches of the recent past across the subject of AI-driven protocols, decision-based algorithms, and classical congestion-control operations. Interestingly,despite the efficacy demonstrated by the Distributed Congestion Control Protocol (DCCP) and the Markov Decision Processes, these tasks commonly fall short when it comes to addressing cloud-specific environments typified by momentary surges, resource contention, and network congestion due to high concurrent request rates. The systematic analysis thus points out a small extent of scalability, flexibility, and context-awareness of the existing solutions in real-time cloud deployments of congestion control. Collectively, these results highlight the absolute need to develop a powerful and smart congestion-control architecture that has the adaptive capacity to dynamically react to chaotic, variable conditions in the cloud network. The proposed direction contributes to improving quality of service (QoS), enhancing network efficiency, and advancing the design of next-generation congestion control mechanisms in cloud computing environments.