Ricochet: Scalable Passive Beamforming for mmWave Networks Using Reflectarrays

Millimeter-wave (mmWave) technologies have seen explosive growth in the past few years with bleeding edge solutions offering unprecedented communication capacity. Unfortunately, they also introduce new challenges such as limitations in range, coverage blind spots and severe disruptions due to blockages from obstructions. Reconfigurable intelligent surfaces (RIS) have been used to alleviate some of these problems. However, RIS solutions create additional problems such as limited NLoS range and complicated beam management protocols making it incompatible with 5G. In this paper, we propose Ricochet, a fully passive, PCB-fabricated reflective surface which uses reflect-arrays for far-field communications. Ricochet can re-steer and reshape incidental mmWave signals towards anomalous directions to divert around obstructions, and generate wider/narrower or multi-beam patterns to fill coverage holes. To overcome the reflect-array's lack of reconfigurability, we introduce post-fabrication techniques that enable incremental deployment of Ricochet to adapt to environment/specification changes, enhancing the SNR or broadening the coverage of an existing surface. We further design a multi-user algorithm that enables multiple transmitters to share the same Ricochet surface while avoiding interference. Our experiments with 5G towers and WiGig access points, demonstrate that Ricochet can effectively improve coverage without any modification to existing network infrastructure or mobile devices.