Time-shifted pseudo-noise multibeam lidar array using acousto-optic deflectors

We present a multibeam lidar system that simultaneously transmits an array of K low-crosstalk codes and detects the backscattered returns from all directions on a single large-area high-speed detector via direct detection, without a local oscillator beam. We generate a series of long, time-shifted, self-synchronizing maximum-length pseudo-noise (PN) codes in the form of an array of K intensity-modulated beams from a single aperture, using only two traveling-wave acousto-optic deflectors (AODs). The cascaded AODs, driven by counter-propagating acoustic waves, are imaged onto each other, and the diffracted fields are imaged onto the far field target, where the interferometric product between two binary phase-shift keying (BPSK)-encoded maximum-length (ML) sequences generates time-permuted versions of the PN code as an array of K beams highly identifiable as well-separated peaks in the cross-correlation. The time delay of the various beams is optimized to be well separated across the multibeam array and is delayed by many milliseconds, even though the acoustic propagation time is only a few microseconds. This relies on the shift-and-add property of ML codes to produce time-shifted versions of a PN code from the bitwise mod-2 addition (XOR) of two relatively delayed copies of the code, where XOR is instead implemented as the interferometric product between BPSK bipolar PN codes. This PN-encoded interferometry enables an unconventional multibeam lidar imaging system that does not require scanning mirrors at the transmitter or multi-pixel wideband detector arrays at the focal plane of the receiver. A preliminary experimental demonstration of a 32-beam array parallel ranging, using N=2¹⁵-1 maximal-length codes, shows how a single cross-correlator computationally identifies the different time-shifted beams and their time-of-flight range delays.