Optics Group -

Physics and Mathematical Physics, Univ. of Adelaide

The effect of noise on the nonlinear dynamics of lasers

The work with nonlinear optical processes and noise has a limitation in that one cannot follow the dynamics directly - the electric field and atomic polarisation change on a timescale of 10^-15s, and we don't have detectors and oscilloscopes that can directly see anything that fast. So to get around that limitation we are looking at slower nonlinear systems with a view to adding noise and measuring its effects on the behaviour. This type of experiment is usually done with analogue circuits, or on a computer (if you stretch the meaning of "experiment" somewhat), so to be different we have chosen a real physical system, the nonlinear relaxation oscillation of a laser. This typically has a timescale between 10^-6s and 10^-10s, slow enough to follow directly. The benefit of staying with an optical system is that external influences are relatively easily controlled.

We are looking at two lasers, diode lasers and Nd lasers. In semiconductor diode lasers, the refractive index of the lasing medium depends on the the gain in a way that has a fairly strong effect on the laser's operation. There are also many approximations made in getting from fundamental solid state physics to the usual equations used to describe the laser's behaviour. The nonlinear dynamics is strongly influenced by these details and we are currently sorting this out.

Our Nd laser used in these dynamical studies operates on several longitudinal modes which are coupled through having to share the same gain medium. Thus we have several coupled oscillators and the interaction of these shows some interesting effects, whereby they can sometimes operate in an antiphased regime. This means that individual modes fluctuate so as to cancel each other out, leaving a relatively simple total behaviour. This is borne out by the application of phase space reconstruction techniques to data where the laser is operating chaotically, telling us that the overall behaviour is governed by about three variables, but we know by "dissecting" the system that there are at least nine or ten. A further interesting question is what happens to this antiphasing in the presence of external noise? We're working on it!

Recent Publications and Reports

Optics Home Page

queries and comments to mwh@physics.adelaide.edu.au