photonic integrated circuitry - utdallas.edudlm/duncan phot ics jan 2004.pdf · photonic integrated...
TRANSCRIPT
The Erik Jonsson School of Engineering and Computer Science
Photonic Integrated Circuitry
Duncan MacFarlaneThe University of Texas at Dallas
Gary EvansSouthern Methodist University
Jack Kilby’s revolutionary integrated circuit
The Erik Jonsson School of Engineering and Computer Science
Introduction
g Early electronic filters used only L, C and R’s –“Passive Filters”
g Gain elements, first vacuum tubes, later transistors, allowed “Active Filters”
g Current optical filters are purely passiveg It is time to develop optical filters with gain that are
higher performance, and adaptive
We are developing a manufacturable, programmable, Photonic Integrated Circuit!
The Erik Jonsson School of Engineering and Computer Science
Active Lattice Filters
An active optical filter has a gain element that allows the weights associated with the different poles and zeroes to be tuned.
g Lattice Filters good for:– Linear Prediction– Frequency Discrimination– Robust wrt realization
limitations
g Gain allows:– Improved quality factors– Adaptive filters
+
+
+
+
Z-1/2
Z-1/2
G
G
Tk(z)
Rk(z) Rk+1(z)
Tk+1(z)tk
tktk-1
tk-1
-rkrkrk-1 -rk-1
Gain Block Delay Block
The Erik Jonsson School of Engineering and Computer Science
Surface Grating Photonic IC
We are using the GSE technology for integrated active optical filters
AlGaInAs/InP Material System w. MQW waveguide/active gain region
electrode electrodeelectrode electrodeInput
Reflected Output
Transmitted Output
SurfaceGrating
Gain AndDelay Stage
Grating
Gain Region
The Erik Jonsson School of Engineering and Computer Science
Four Directional Couplers
g Grating can couple in multiple directions– Input/output– Mesh structure
g FIB nanostructureg Holographic lithography
– Two step process– Crossed Gratings
The Erik Jonsson School of Engineering and Computer Science
2D GSE Lattices: “Thick Linear”
0 0.2 0.4 0.6 0.8 1-1000
-800
-600
-400
-200
0
Normalized Frequency (×π rad/sample)
Phas
e (d
egre
es)
0 0.2 0.4 0.6 0.8 10
0.2
0.4
0.6
0.8
1
Normalized Frequency (×π rad/sample)
Mag
nitu
de
0 0.2 0.4 0.6 0.8 1-100
-50
0
50
100
Normalized Frequency (×π rad/sample)
Phas
e (d
egre
es)
0 0.2 0.4 0.6 0.8 10
0.2
0.4
0.6
0.8
1
Normalized Frequency (×π rad/sample)M
agni
tude
The Erik Jonsson School of Engineering and Computer Science
Two Dimensional GSE Lattices
Completely new filter structure solved by layer peeling with 2nx2n matrices … Research is benefiting fundamental DSP engineering
~60 um
The Erik Jonsson School of Engineering and Computer Science
Two Dimensional GSE Lattices
g Multiple Input, Multiple Output Applications– Add/drop applications, dwdm, o-cdma on same chip– Cross correlators, decouplers, cross-talk cancelation– multi-trajectory tracking, joint process estimation
g Fully Tunable … programmableg Extends the palette of filter realizations for
optimal implementations
The Erik Jonsson School of Engineering and Computer Science
Frequency Selection
0 0.2 0.4 0.6 0.8 1-200
-100
0
100
200
Normalized Frequency (×π rad/sample)
Phas
e (d
egre
es)
0 0.2 0.4 0.6 0.8 10
50
100
150
200
Normalized Frequency (×π rad/sample)
Mag
nitu
de
Stable operation withQ >75,000
The Erik Jonsson School of Engineering and Computer Science
Bandpass Frequency Tuning via Gain
0.40.5
0.60.7
0.80.9
0.4
0.6
0.8
10.26
0.27
0.28
0.29
0.3
Gx00
Peak Postion vs.Gx00&Gy00
Gy00
Peak
Pos
ition
The Erik Jonsson School of Engineering and Computer Science
Economically Viable Device
g Structure is highly manufacturable– Standard Epitaxy– Photolithographic gratings
• Last step in process• No regrowth• Chip scale process (parallel)
– Being commercialized for lasersg Basic structure can be standardized for fabrication
purposes– Individual users may then program them for a particular
application– Just like an FPGA or a DSP
The Erik Jonsson School of Engineering and Computer Science
Outcome of the Research Program
g Standardized Photonics– Two dimensional lattice has wealth of transfer functions
supporting widespread applications– Programming determines application
g Basic advances in DSP– Two dimensional structure is new– Highly appropriate for MIMO applications
100 GHz clock rates and GHz tuning rates will enable high volume information engineering applications