Robust channel estimation for reconfigurable intelligent surfaces in presence of impulsive noise

Abstract

Reconfigurable intelligent surfaces (RIS) are poised to revolutionize 6G communication systems by manipulating wave propagationforenhancedcapacityandcoverage.However,optimalRISoperationhingesonaccuratechannelstateinformation (CSI), achallengeunderreal-worldimpulsivenoiseconditionswhereconventionalmethodsfalter.Thispaperproposesanovel correntropy-based stochastic gradient ascent (CSGA) learning algorithm for robust CSI estimation in RIS systems plagued by impulsive noise. Our CSGA method demonstrably outperforms existing techniques, leading to a significant improvement in the communication system’s average bit error rate (BER). This paves the way for reliable RIS operation in future 6G networks.