Radar and communication systems may share spectrum without sharing timing, phase, or channel information. That becomes especially difficult for a communication user inside a radar exclusion zone: the radar remains active, its interference phase is hard to track, and reconfigurable intelligent surface (RIS) phase errors can undermine coherent processing. Prof. Paschalis C. Sofotasios and coauthors study detection rules designed for precisely this incomplete-information setting.
A receiver cannot assume perfect coordination
The public record models an RIS-assisted communication link exposed to uncoordinated radar interference. Two uncertainties matter at once. The receiver may lack instantaneous channel state information, while the RIS itself may apply phases imperfectly. A detector that quietly assumes either uncertainty away can look attractive analytically but become unusable in the scenario the model is meant to represent.
The paper therefore treats the radar phase as a nuisance quantity rather than something the receiver must estimate on every decision. This shifts the design question from recovering every hidden variable to extracting a reliable decision statistic from the information that is actually available.
Two detectors match two information budgets
The first design is non-coherent: it avoids instantaneous channel state information and incorporates RIS phase uncertainty directly. The authors derive an exact likelihood expression and, for low-to-moderate signal-to-interference-plus-noise ratio, a closed-form detector. The second design is a mismatched coherent detector for a receiver that has channel information, but only imperfectly. Its analysis leads to a pairwise-error-probability approximation evaluated through Gauss-Chebyshev quadrature.
These are not merely two algorithms competing under identical assumptions. They represent two receiver operating points. One spends less on channel knowledge; the other uses imperfect channel estimates and accepts the resulting model mismatch.
What the public evidence establishes
According to the public abstract, numerical and analytical results place the proposed non-coherent rule close to the optimal maximum-likelihood detector in the evaluated regime. The result is encouraging because the practical rule avoids tracking the radar interference phase. The evidence available here does not establish performance outside the modeled channel, RIS-error, and SINR assumptions, so hardware behavior and broader deployment conditions remain open questions.
The main design lesson is broader than a single detector: in coexistence problems, robustness can come from changing what the receiver tries to estimate, not only from improving an estimator for an inaccessible quantity.
Research notes
RIS-Assisted Radar-Communication Coexistence: Detection Analysis with Channel Uncertainties
Authors: Rawan Derbas, Shimaa Naser, Hamad Yahya, Sanjeev Gurugopinath, Paschalis C. Sofotasios, and Sami Muhaidat.
Status: Public preprint record dated 24 August 2026.
What the public evidence establishes: The work derives non-coherent and mismatched coherent detectors for an RIS-assisted communication user facing uncoordinated radar interference, RIS phase uncertainty, and imperfect or unavailable instantaneous channel information.
Limits: The public abstract reports analytical and numerical evaluation; it does not establish hardware validation or performance beyond the stated model and operating regime.