RXs Directions based Codebook Solution for Passive RIS Beamforming

Recently, reconfigurable intelligent surface (RIS) has gained popularity to overcome blockage issues and boost coverage in advanced 6G networks. Passive beamforming (PBF) process redirects incident signals from the transmitter (TX) toward the receiver (RX) by adjusting the amplitudes and phase shifts of RIS elements. Though channel estimation (CE)-based PBF schemes are optimal, they need high overheads and time consumption, especially with a substantial number of RIS elements. In contrast, codebook (CB) based PBF solutions explore limited reflection patterns (RPs) to determine the optimal one using a predefined metric, resulting in sub-optimal performance. However, these still require high power and time due to the size of the used CB. To resolve these issues, we suggest a direction-based PBF (D-PBF) scheme, aiming to map the RX's directions to the codebook RPs. Then, this information is stored in an updated database (DB), and if a matching happens between an incoming RX and a specific RP exists in the DB, the proposed scheme directly selects this RP to configure the RIS elements. If not, the scheme records this codeword for future searches. In the event of a failed match, the D-PBF scheme searches through all the recorded RPs to select the optimal one and updates the DB accordingly. After a period, that depends on the CB size, the DB converges, eliminating the need of searching for the optimal RP. The proposed scheme significantly reduces overhead, power, and time requirements compared to the CE and conventional CB-based solutions, while maintaining satisfactory performance in terms of effective rate. For example, it can obtain at least 80% of optimal effective rate with 10
³
lower computational complexity comparable to lowest complexity CE based scheme.