On the Use of CVRP to Diagnose Faulty Elements in Antenna Arrays

TL;DR

This work shows that Constrained-View Radiated Power (CVRP), a FoV-area-normalized, amplitude-only metric, can diagnose the number of faulty elements in a phased-array without requiring full phase information or complete pattern acquisition. Using a 2×8 cosine-element array and simulated on-off faults under beamsteering, the study demonstrates that CVRP can distinguish between different fault counts when the angular resolution is sufficiently fine and measurement errors are controlled; feasibility is shown across direct, indirect far-field, and near-field setups. The findings indicate practical, time-efficient diagnostic potential for QA and field testing, with limitations: CVRP does not identify the identities of the failed elements, and discriminability depends on $\theta_{FoV}$, $\text{RES}$, and error levels. Future work will broaden error models, test other topologies, and explore amplitude/phase-excitation faults to extend CVRP-based diagnostics.

Abstract

This paper investigates the application of Constrained-View Radiated Power (CVRP) for diagnosing phased array element failures, specifically focusing on on-off element failure. CVRP, similar to Partial Radiated Power (PRP), considers a specific Field-of-View (FoV) but normalizes it by the FoV area. The study explores CVRP's effectiveness in detecting failures in a 2x8 cosine element array under beam-steering conditions, accounting for random and depointing errors, angular resolution, and pattern rotation. Results indicate that CVRP can detect on-off failures based on angular resolution and error severity, under the assumption of reduced Total Radiated Power (TRP) with element failures. Additionally, CVRP is effective with partial far-field patterns, making it suitable for near-field, indirect far-field, and far-field measurement systems without requiring phase acquisition in the latter two.

Figures (8)

• Figure 1: Definition of $\varphi$ and $\theta$. Source: MATLAB_PHITHETA.
• Figure 2: Simulated array geometry and directivity pattern. On the left, the array geometry plot includes the indexes of each element, as well as their normals in red. Generated using the Sensor Array Analyzer app from MATLAB.
• Figure 3: Simulated array geometry and directivity pattern. On the left, the pattern of the array with $45^\circ$ beamsteering without rotation. In the middle, the pattern of the array with $45^\circ$ beamsteering and a $45^\circ$ clockwise rotation around the y-axis. On the right, the geometry of the array with the $45^\circ$ clockwise rotation around the y-axis. Note that the rotation is applied to both the positions of the elements and their normals. Also, note that the geometry of the array without the rotation is the same as the one depicted in Fig. \ref{['F2']}.
• Figure 4: Post-processing rotation CVRP minus physical rotation CVRP for $45^\circ$ beamsteering case. The same angular resolution ("RES") is used for computing the CVRP values to be subtracted, e.g., the case "FE: 15, RES: 5º, DEP: 0º" is the result of the case "FE: 15, RES: 5º, DEP: 0º" with post-processing rotation minus the case "FE: 15, RES: 5º, DEP: 0º" with physical rotation. No depointing nor random errors are considered.
• Figure 5: CVRP difference between the considered cases in the legend and the no faulty element case, with the corresponding angular resolution and without depointing, which we denote as reference. Upper bound of CVRP CI [dBm] minus reference CVRP [dBm] in dashed line. Lower bound of CVRP CI [dBm] minus reference CVRP in dotted line. Average of CVRP with random errors [dBm] minus reference [dBm] in solid line. The considered $\sigma_{err,dB}$ is $1$ dB. E.g, the case "FE: 15, RES: 5º, DEP: 3º" is the result of the average of the $1000$CVRP values considering a $\sigma_{err,dB}$ of $1$ dB minus the case "FE: 0, RES: 5º, DEP: 0º" without any random error.