Beamforming tests

First test is to send a CW signal and record the response on the 16 channels. The result is beamformed and visualized as a polar-image, where angle-range is plotted. Range is in meters, angels is in deg.

Total range: 30m

Swath is pre-set to the open-angle, as decribed above. Right is the color-scale. Resulting image is built up of 5-pings. Images are stacked. Stacking process removes some of the specke-noise or background noice and amplifies the objects. 

The brighter lines are actual objects that reflect the ping. Distance is measured by counting samples and convert to meters. Looking closely at the image, we can see that the objects fade above 15m out. The room I am testing in, is much smaller so everything above 1.5m-2.5m out is multipath or sound reflected off walls and furniture. 

Similar distance measurement, like the SR04 is designed to detect objects within the swath and closer than 5m. Above image shows objects more than 15m away, plus where they are.

The CW pulse is short, 100us long, which require more power to transmit in order to get further away. 

One solution is to use a different pulse, FM (frequency modulated) where frequency sweeps. Using this in combination with a matchfilter, we can use longer pulses, 0.5ms-3ms, so less power is needed. Result should be that we can detect objects further way that the 15m.

Total range is 50m. 

Notice the object just above 35m and -7deg. This is not visible using the CW pulse.

On purpose, I have not interpolated the beams in the above images. This is to highlight how wide the beams actually are and how large the objects are.  The 16 channels have been oversampled into 128 beams, which explains why the objects witch seem larger than the beam-width from the beamforming. 

The oversampling trick is especially interesting when we apply movement to the sonar, similar to what mobile phones are doing with super-resolution. The use multiple images and align them to get an even higher resolution image.

Back to phased array project...