Practical Limitations of the Known Waveform DoA Estimation Algorithm

Known-waveform (KW) direction-of-arrival estimators exploit a known reference signal to improve spatial selectivity, but their practical performance depends on operating conditions that have not been systematically quantified. This paper provides a rigorous simulation-backed characterization of three such conditions: interferer modulation format, synchronization error between the received signal and the local reference, and available training-sequence length. We adopt a packet-based observation model in which the known preamble occupies a finite portion of the received buffer so that synchronization errors shift the correlation window into pre-arrival noise or unknown payload, and we evaluate performance via Monte Carlo trials on a compact uniform circular array. Results show that the KW approach retains a clear advantage over classical blind methods under unstructured interference; it remains largely robust across common structured interferer modulations but exhibits measurable residual sensitivity dependent on symbol alphabet and pulse shaping; its performance degrades sharply outside a narrow synchronization tolerance around correct alignment; and a regime-dependent minimum training length is required before KW gains appear. Together, these results delineate the operating envelope for KW spatial processing and yield concrete guidelines: prefer KW estimators when synchronization can be maintained within the identified tolerance and sufficient preamble length is available; otherwise classical covariance-based approaches may be preferable.