Neural Network on a Photonic Chip
In the PELM project the basic element is a SCISSOR (side-coupled integrated spaced sequences of optical resonators), that’s designed in simple linear geometries and in complex topologies.
The first step was analyzing different single ring resonators (with different gaps and different bending of the waveguides and extracting the characteristic parameters for both (such: quality factor, group index, coupling coefficient…). The analysis of the single micro-resonator will enable us to study in a better way the SCISSOR. By adding ring resonators to a single bus waveguide on a fixed interval, one can create a SCISSOR, in neural network point of view this SCISSOR works as a reservoir with a complex perceptron in the output. In our case, we are measuring a new SCISSOR with 4 ring resonators and trying to understand its behavior, for that the thermal tuning is the most appropriate technique and this is achieved by using micro-heaters which are placed on the top of the device as a heat source with sufficient isolation to avoid optical loss.
PhD workshop day 01-12-2020