An experimental study of a self-organizing acoustic communication channel

We present the results of an experimental study of an adaptive information transmission system with phase diversity. The adaptation is carried out by machine-learning methods and consists in regulating the emitter phases. An acoustic waveguide with a tunable metasurface was used as a propagation medium. A transmission rate of 1 kbit/s was achieved at a multipath propagation and significant intersymbol interference. The adaptation time of the studied system to abrupt changes in channel parameters does not exceed ten training cycles even in the case of significant fading.