All-optical signal processing

One of the key technology enablers for the future all-optical networks, particularly ones with ultra-long-haul, burst- or packet-switching capabilities, is the all-optical signal processing including signal regeneration, wavelength conversion, and switching. While many different all-optical implementations of these functions have been proposed over the last two decades, they all face one major challenge: in order to become viable alternatives to their electronic counterparts, the all-optical signal processors need to realize the full advantage of inherent parallelism of optical processing. This means they must be able to handle simultaneously many optical channels carried by different wavelengths (i.e. support wavelength-division multiplexing, or WDM) without dramatic increase in complexity and cost.  Until recently, this challenge has seemed to be insurmountable. Indeed, an all-optical signal processor fundamentally relies on strong nonlinear-optical effects, which in turn lead to debilitating interaction among the WDM channels, e.g. by means of four-wave mixing (FWM) and cross-phase modulation (XPM). So far, attempts to perform nonlinear-optical processing of multiple WDM channels have proved unsuccessful. 

We have recently proposed a way solve this problem and enable simultaneous nonlinear-optical signal processing of many WDM channels (see reference below). The essential feature of our approach is using a novel nonlinear-optical medium with peculiar dispersion properties. These properties are such that different frequency components within each WDM signal's spectrum propagate with nearly the same velocities to support integrity of the pulses and strong intra-channel nonlinearity. At the same time, velocities of any two separate WDM channels are considerably different, ensuring mitigation of inter-channel nonlinear effects through significant FWM phase mismatch and fast XPM bit walk-off between the neighboring channels. The nonlinear material with such dispersion properties does not exist in nature, but can be constructed artificially by alternating sections of highly-dispersive Kerr medium (e.g. optical fiber) and optical filters with periodic group delay response (periodic group-delay devices, or PGDDs).

Currently, we are working on experimental demonstration of multichannel all-optical signal processing based on this principle.

References

M. Vasilyev and T. I. Lakoba, "All-optical multichannel 2R regeneration in a fiber-based device," Opt. Lett. 30, 1458 (2005).