This thesis proposes an integrated structure to improve the computational complexity of MVDR beamforming, mutual interference between adjacent radars, and angular resolution limitations caused by a restricted number of antenna elements in multi-FMCW/M...
This thesis proposes an integrated structure to improve the computational complexity of MVDR beamforming, mutual interference between adjacent radars, and angular resolution limitations caused by a restricted number of antenna elements in multi-FMCW/MIMO radar environments. The proposed structure applies Stochastic Weighted Sub-block Update (SWSU) to the covariance matrix estimation stage, reducing processing time by approximately 66.3% compared with conventional MVDR while maintaining similar beamforming performance; in FPGA implementation, LUT and FF usage are reduced by approximately 5.8% and 17.4%, power consumption by approximately 6.7%, and URAM usage is eliminated. In addition, orthogonal-code-based encoding and decoding separate multiple radar transmission signals and mitigate mutual interference between adjacent radars, while the experiment using the non-overlapping regions of two arrays with different beamwidths shows a BER increase of approximately 25.7% compared with the average BER of individual radars, confirming a trade-off between angular resolution and detection reliability.