Active optical lattice filters

Highly integrated, novel architectures for optical filtering can leverage structure, gain and variable delays to provide a multi-use photonic platform. Hardware and software results are presented.


Abstract
Highly integrated, novel architectures for optical filtering can leverage structure, gain and variable delays to provide a multi-use photonic platform. Hardware and software results will be presented.
Summary The filtering of signals is a very broad concept. A wide variety of physical effects and practical calculations may be analyzed and performed using the mathematics of filtering, signal processing and control theory. The authors of this paper see the application of filtering to photonics as having the potential to unlock a very powerful versatility to optical devices and systems. Can a flexible pIatfonn be developed that allows an end user to program functionality in a manner similar to a DSP or an FPGA?
Toward this end, the authors are developing photonic filters with new architectures, and with novel extensions including gain, for active filters, and variable phase, for multirate signal processing. Our experimenta1 work is in Inp and represents a photonic circuit with a highly integrated architecture. The InP epitaxy provides semiconductor optical amplifier regions between the nanophotonic couplers. These semiconductor optical amplifiers provide the delay and the broad gain bandwidth for optical signal processing. The speed of these amplifiers provides tremendous agility to the photonic integrated circuit. A one dimensional lattice filter is well known, and examples such as thin film filters, and multiple quantum wells are plentiful. The two dimensional lattice however is a non-trivial extension to a simple, traditional, Iayered structure, and yields a very rich array of behavior directly usehl to optical filtering. A block diagram of the two dimensional active lattice filter is shown in Fig. 1. The two dimensional mesh is enabled by a four directional coupler. The signal flow diagram for a four directional coupler is shown in Fig. 2. In operation, the fourdirectional coupler accepts an input signal from any of its four ports, and couples a fraction of that input to the outputs of each of the four ports. In most filter theory to date, the coupling coefficients between the delays are the only parameters that adjust the filter properties. An innovation of the active filter is the movement of the design parameter from the coupling coefficients to the gain coefficients between the couplers.
The researchers have fabricated an InP device, and scanning electron micrographs (SEMs) of this first generation wafer is shown in Fig. 3. In Fig. 3 are two views of a field of semiconductor optical amplifiers configured in the arrangement depicted in Fig. 4 above. On the left is a wide view of the field and on the right is a close up with higher magnification. The triangular regions in Fig. 3 are bond pads that allow current to be injected into the semiconductor optical amplifier ridge waveguides. In this device the ridges are approximately 2 microns wide and 120 microns long. As per Fig. 1 there are ridges aligned in the x direction and the y direction. At the junction of four semiconductor optical amplifier ridges is a 30 micron by 30 micron region in which a nanophotonic coupler will be fabricated to couple light among the ridges. Near the center of the wide view are ridge amplifiers that are three times Ionger than standard ridges. These features were included on the mask to provide an on-chip laser source. The layout of the photonic integrated circuit shown in Fig. 3 is versatile. The specific filter structure is determined by the choice of nanophotonic coupler at the intersection of the semiconductor optical amplifier ridges. It is economically significant that this step in the process OCCLUS after all the photolithographic steps. One processed wafer may thus support multiple filter applications. Different examples of four directional couplers are shown in Fig. 4.
This architecture offers a sophisticated, and complicated, approach to process optical information, The transfer functions connecting multiple inputs to multiple ouputs may be adjusted in nanoseconds. The large number of poles may be viewed as a large number of degrees of fieedom, The presence of gain allows higher quality factors, and system scalability. The architecture is an efficient use of material; programmability suggests the architecture is economically sound,