Traditional Reconfigurable Intelligent Surfaces (RIS) are often described as "smart mirrors." They consist of an array of passive elements that can shift the phase of incident electromagnetic waves, allowing a base station to steer signals toward a user even without a direct line-of-sight. However, mirrors have a fundamental physical limitation: they are opaque. They only serve users in the "reflection half-space," leaving anyone behind the surface in a communication dead zone.
To achieve truly ubiquitous 6G coverage, researchers have evolved this concept into Simultaneously Transmitting and Reflecting (STAR) RIS, also known as Intelligent Omni-Surfaces (IOS). By allowing signals to both bounce off and pass through the surface, these devices transform the "smart mirror" into a "smart window," providing 360-degree coverage.
Breaking the Half-Space Barrier
The core innovation of STAR-RIS is the ability to manipulate the magnetic and electric currents on a metasurface to split incident energy. As Shuhang Zhang, Boya Di, and Marco Di Renzo demonstrated in Intelligent Omni-Surface: Ubiquitous Wireless Transmission by Reflective-Transmissive Metasurface, this dual-mode operation eliminates the coverage "blind spots" inherent in reflective-only designs.
In a typical deployment, a STAR-RIS might be placed on a building's window. It can simultaneously reflect signals from an outdoor base station to a pedestrian on the street while transmitting (refracting) those same signals to a user inside the building. This "full-space" manipulation effectively doubles the utility of the hardware compared to traditional RIS.
Three Protocols for Signal Splitting
How does a single surface element decide whether to reflect or transmit? Researchers have identified three primary operating protocols to manage this resource allocation:
- Energy Splitting (ES): Every element on the surface splits its received power into two components—one reflected and one transmitted. As noted by Jiakuo Zuo and Yuanwei Liu in Joint Design for Simultaneously Transmitting And Reflecting (STAR) RIS Assisted NOMA Systems, this provides the highest degrees of freedom for beamforming because every element contributes to the signal gain for users on both sides.
- Mode Switching (MS): The surface elements are partitioned into two groups. One set is dedicated entirely to reflection, while the other is dedicated to transmission. Xinwei Yue and Jin Xie show in Simultaneously Transmitting and Reflecting Reconfigurable Intelligent Surface Assisted NOMA Networks that while this is easier to implement in hardware, it reduces the total beamforming gain for each individual user.
- Time Switching (TS): The entire surface toggles between reflection and transmission modes in different time slots.
Synergies with NOMA
STAR-RIS is particularly powerful when paired with Non-Orthogonal Multiple Access (NOMA). In a NOMA system, multiple users share the same frequency and time resources, separated instead by power levels.
Chenyu Wu and Xidong Mu argue in Coverage Characterization of STAR-RIS Networks: NOMA and OMA that STAR-RIS provides the necessary spatial degrees of freedom to make NOMA more effective. By independently controlling the channel conditions for users in the transmission and reflection spaces, the system can "engineer" the channel differences required for NOMA to successfully decode overlapping signals, leading to significantly higher sum-rates and fairer coverage.
Beyond Communication: Sensing and Computing
The 360-degree capability of STAR-RIS is being extended into Integrated Sensing and Communication (ISAC). In STARS Enabled Integrated Sensing and Communications, Zhaolin Wang and Yuanwei Liu describe how a surface can partition the environment into a "sensing zone" and a "communication zone." This allows a single infrastructure to track a moving vehicle (sensing) while simultaneously delivering high-speed data to a user (communication) on the opposite side of the surface.
Furthermore, the field is moving toward Reconfigurable Intelligent Computational Surfaces (RICS). As explored by Bo Yang in Reconfigurable Intelligent Computational Surfaces for MEC-Assisted Autonomous Driving Networks, these surfaces can perform analog signal processing to cancel interference in real-time, which is critical for safety-sensitive applications like autonomous driving.
Hardware Realities: The Coupled Phase-Shift Model
While early theoretical work assumed that reflection and transmission phases could be controlled independently, real-world hardware often imposes constraints. Yuanwei Liu and Robert Schober addressed this in Simultaneously Transmitting and Reflecting (STAR)-RISs: A Coupled Phase-Shift Model, demonstrating that the transmission and reflection coefficients are physically coupled. Even with these constraints, STAR-RIS remains superior to traditional RIS, providing a robust path toward the "programmable wireless environments" envisioned for 6G.