Modeling Adaptive Interference Management in Ultra-Dense Networks Assisted by Reconfigurable Intelligent Surfaces Using Dynamic Soft-FFRR and Clustering

Abstract

Ultra-Dense Networks (UDN) and Reconfigurable Intelligent Surfaces (RIS) have emerged as two pivotal technologies for meeting the growing demand for network capacity and service quality. UDN boosts capacity by densely deploying femtocells, whereas RIS broadens coverage by relaying obstructed signals toward the receiver. However, the dense femtocell deployment introduces co-tier and cross-tier interference, and the reflective nature of RIS aggravates this problem, since interfering signals are likewise steered toward the receiver. To overcome this, the present study proposes an adaptive interference management scheme that combines dynamic clustering and Soft Fractional Frequency Reuse (Soft-FFR) within a RIS-assisted UDN. Clustering suppresses co-tier interference, while re-clustering and channel allocation are triggered by the network state, namely the estimated SINR and service zone of each femtocell. Four system models baseline, Dynamic Clustering, dynamic Soft-FFR, and the proposed integration were simulated on a three-macrocell network with up to 330 femtocells. The proposed method achieves the best overall performance among the evaluated schemes. Compared with the baseline, the SINR CDF at 20 dB falls from 64% to 43% and the throughput CDF at 70 Mbps decreases from 84% to 62%, indicating that a substantially larger proportion of users attain high SINR and throughput. The spectral efficiency at 330 HeNBs increases from 5.9 to 6.9 bit/s/Hz, and the bit error rate is reduced from about 0.13 to 0.05. These results confirm that adaptively integrating Dynamic Clustering and Soft-FFR effectively suppresses both co-tier and cross-tier interference, thereby improving the reliability and spectral efficiency of dense RIS-assisted UDNs.