Comments on “transform-limited pulse” and optical intensity autocorrelation

The cleanness of fast optical pulses, especially those transmitted through optical communication networks, is overwhelmingly referenced to and judged by the so-called “transform-limited pulse”, i.e., pulses with full-width at half-maximum (FWHM)-defined time-bandwidth product around 0.5. In this paper, we will demonstrate the fundamental flaw of this notion using our simulation results. Moreover, the pulse width is usually obtained by optical intensity autocorrelation plus curve-fitting, which may be misleading as well. Consequently, these standard methods of pulse characterization should be thoroughly re-considered.