An Adaptive Software Defined Network-on-Chip (SD-NoC) For Varying Network Resources in Offering Optimal Service Quality: Survey

In the evolving landscape of multiprocessor systems-on-chip (MPSoCs), the Network-on-Chip (NoC) paradigm has emerged as a pivotal solution for on-chip communication challenges. However, static NoC architectures often struggle to adapt to the dynamic nature of application workloads and resource availability, leading to suboptimal performance and inefficiencies. This paper introduces an adaptive Software Defined Network-on-Chip (SD-NoC) framework that dynamically adjusts network resources to optimize service quality. The proposed SD-NoC leverages a centralized controller to monitor network conditions and manage data traffic, ensuring efficient resource utilization and minimal latency. By employing adaptive routing algorithms and reconfigurable network topologies, the SD-NoC can respond to varying workloads and network conditions in real-time. Our approach integrates machine learning techniques to predict network traffic patterns and preemptively adjust the network configuration, further enhancing performance and reducing energy consumption. Extensive simulations and experimental evaluations demonstrate that the adaptive SD-NoC significantly outperforms traditional NoC architectures in terms of throughput, latency, and energy efficiency. The results highlight the potential of SD-NoC to meet the demanding requirements of future MPSoCs, providing a scalable and flexible solution for high-performance computing applications.