ADAPTIVE DIMENSIONAL SEARCH BASED IMPROVISED ENSEMBLE METHOD FOR COGNITIVE RADIO IN SOFTWARE-DEFINED NETWORKS
B. Lalitha, S. Poongothai · ICTACT Journal on Communication Technology · 2025
Software-Defined Networks (SDNs) coupled with Cognitive Radio (CR) systems offer dynamic spectrum access capabilities, enabling efficient spectrum utilization. However, the high dimensionality of network parameters and unpredictable spectrum availability pose critical challenges in achieving real-time adaptation and optimal throughput. Existing adaptive search and decision-making algorithms often fail to scale effectively in high-dimensional state spaces, leading to reduced convergence rates and suboptimal spectrum allocation. Traditional ensemble techniques lack dynamic interaction between learning agents and real-time feedback mechanisms. This work introduces an Improvised Ensemble Method built upon an Empowered Adaptive Dimensional Search (EADS) algorithm. The proposed system yields a 17% increase in throughput, 22% lower latency, and 19% improvement in spectral efficiency.