Is the Dyakonov-Shur plasma-wave instability observable in a real device
In plain words
Theory from 1993 predicts that a steady current through a gated two-dimensional electron channel with different conditions at its two ends spontaneously turns into terahertz oscillations. A clear observation of this self-generated oscillation has not been established.
Precise statement
In a gated 2D electron channel of length $L_{\mathrm{ch}}$ with fixed-voltage source and fixed-current drain, a drift velocity $v_{0}$ below the plasma-wave speed s makes plasma waves grow at a rate of order $v_{0}/L_{\mathrm{ch}}$, at fundamental frequency $f \sim s/(4L_{\mathrm{ch}})$, once growth exceeds viscous and momentum-relaxation damping. Does a real device (e.g. high-mobility graphene) show self-sustained emission at this frequency above the predicted threshold current? Answer yes or no.
What would settle it
Detection of narrowband emission near $s/(4 L_{\mathrm{ch}})$ that turns on at a threshold current and tunes with gate voltage as $s \sim n^{1/2}$, in a device whose damping is characterized independently.
Status in the literature
Unverified note
Emission attributed to the instability was reported in III-V high-electron-mobility transistors from 2004 on but not unambiguously separated from thermal or other sources; graphene devices remain theoretical or detector-mode as of 2026 (low confidence).