RIS-assisted hybrid RF and FSO communications with index modulation-empowered link selection

Free-space optical (FSO) communication offers advantages such as high bandwidth and high transmission rates, making it suitable for 6G backhaul networks. However, FSO systems have drawbacks like sensitivity to fog, atmospheric turbulence, and pointing errors. Radio frequency (RF) links are incorporated to achieve more reliable and wider-range communication. RF links are susceptible to rain and have limited bandwidth but are not sensitive to fog or atmospheric turbulence. The complementary characteristics of RF and FSO systems make hybrid RF-FSO systems highly effective in mitigating environmental impairments. Reconfigurable intelligent surfaces (RISs) can also further extend the coverage and enhance the system bit error rate (BER) performance of RF-FSO. The state-of-the-art link selection schemes in RIS-assisted hybrid RF-FSO are normally based on channel state information (CSI). However, there is a chance that the CSI is inaccurate or even unavailable to the receiver. This paper investigates a hybrid RF-FSO communication system where CSI is unknown. In our system, the RF link is modeled using the Nakagami-m distribution, and the FSO link accounts for attenuation, Gamma-Gamma distributed turbulence, and pointing errors. The key concept of our system is the adoption of index modulation (IM) for link switching so that CSI is not needed at both transmitter and receiver, and we call it RIS-IM. Exact closed-form expressions for the signal-to-noise ratio (SNR) of the RIS-RF and RIS-FSO links are derived using multivariate Fox’s H-function without using central limit theorem, including the probability density function (PDF) and cumulative distribution function (CDF). We derive the closed-form expressions for the average BER (ABER) for RIS-IM using the SNR statistics. Furthermore, numerical analyses of RIS-IM are conducted. The numerical results of RIS-IM are given comparing with pure-IM, RIS-FSO, and hard switching as benchmarks. Mont-Carlo simulation results demonstrate the superiority of RIS-IM in terms of ABER. In the future, more analysis including outage probability (OP) and ergodic capacity (EC) await further explorations.