Experimental Validation Of Direction Of Arrival Estimation For High Chirp-Rate Linear Frequency Modulated Radar Signals

Direction of arrival (DOA) estimation for signals with large time-bandwidth products is very important in modern communication and radar systems. Traditional narrowband DOA estimation algorithms are not accurate for such wideband signals. Further complexities arise with non-stationary signals, for example, high chirp-rate linear frequency modulated (LFM). These complexities imply that achieving real-time performance with good DOA estimation accuracy is a challenge. This article uses experimentally acquired data on high chirp-rate LFM signals to validate a channelised version of MUltiple SIgnal Classification (MUSIC) for DOA estimation. This technique eliminates pre-processing and incoherently combines spatial pseudospectra from individual frequency channels to obtain a single accurate estimate and has much less computational complexity compared to coherent techniques. LFM signals of up to 500 MHz bandwidth and chirp rates of up to 50 MHz/mu s have been used in the investigations and results show accurate DOA estimation at low SNR (0 dB). This technique is suitable for even wider operational bandwidths and low latency implementations. To the best of our knowledge, this is the first time DOA estimation for such high chirp-rate LFM signals has been validated using real experimental data which is also corroborated with validation using synthetic data.