<![CDATA[Ultra-dense networks (UDNs) can increase network capacity but also intensify inter-cell interference and degrade communication performance. This study investigates a hybrid radio resource allocation approach that integrates K-means clustering with dynamic resource block (RB) allocation as an interference management method in UDNs. K-means clustering groups Home evolved NodeBs (HeNBs) according to their spatial coordinates, while dynamic RB allocation selects resources based on communication and interference conditions. The proposed approach is evaluated through MATLAB simulations involving three macrocells with HeNB deployment densities ranging from 10 to 210 per macrocell. Its performance is compared with conventional allocation, standalone K-means clustering, and dynamic RB allocation in terms of signal-to-interference-plus-noise ratio (SINR), average user throughput, bit error rate (BER), and network energy consumption. At a deployment density of 210 HeNBs per macrocell, the hybrid approach achieves an average SINR of 30.10 dB, an average user throughput of 1.799 Mbps, a BER of 0.000258732, and energy consumption of 4.44 × 10⁻⁶ J/bit. Compared with conventional allocation, it improves average user throughput by approximately 77.41% and reduces energy consumption by approximately 43.65%. The BER reduction is approximately 99.50%, based on the unrounded BER values. These findings indicate that integrating spatial clustering with dynamic RB allocation can improve interference management and resource utilization under the evaluated simulation conditions.]]>
