Silicon-Photonics Optical Switch with 3D Waveguide Routing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current silicon photonic switches face challenges with high optical loss, polarization sensitivity, and limited scalability due to their design, which restricts their use in high-port-count datacenter networks, where low loss and polarization independence are crucial.

Innovation Solution

The development of a silicon-photonics optical cross-connect system with 2x2 switching cells featuring bus waveguides in different planes and a movable shunt waveguide, eliminating polarization-dependent elements and using adiabatic couplers for efficient switching, thereby reducing polarization sensitivity and increasing port count.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If silicon photonic switches use conventional waveguide designs, then manufacturing is simplified, but polarization sensitivity increases causing high polarization-dependent loss

Engineering Contradiction:
Improvewaveguide design simplicityVSAvoidpolarization-dependent loss
Core Design Contradiction:
Ease of manufactureVSReliability

Solution Approach 1:

The patent transitions from planar 2D waveguide layouts to a 3D configuration where waveguides are positioned at different vertical levels (z-dimension). Specifically, the through waveguide and drop waveguide are separated in the vertical dimension, allowing them to cross without interaction. This dimensional separation eliminates polarization-dependent coupling while maintaining manufacturing simplicity through standard photonic integration techniques.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Productivity

If silicon photonic switches increase port count for high radix switching, then network scalability improves, but optical loss increases exceeding loss budget

Engineering Contradiction:
Improveswitch radixVSAvoidoptical loss
Core Design Contradiction:
ProductivityVSLoss of energy

Solution Approach 1:

The patent divides the optical switching function into distinct segmented waveguide paths: input waveguides, through waveguides, drop waveguides, and output waveguides. Each segment is optimized for its specific function with minimal coupling points. The 1x2 switching cell architecture segments the routing logic, allowing complex high-radix switching to be built from low-loss modular units, thereby maintaining acceptable optical loss budgets even as port count increases.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent addresses the inherent optical loss in silicon photonic materials by designing waveguides with extended interaction lengths that convert the propagating light into evanescent fields for coupling purposes. The adiabatic coupling regions are designed to maximize mode overlap and coupling efficiency, transforming what would be propagation loss into useful switching functionality with minimal insertion loss.

Inventive Principle:
Principle #22Blessing in disguise (Convert harm into benefit)

3Device complexity

If silicon photonic switches use planar waveguide crossings, then device complexity is reduced, but crosstalk increases due to polarization-dependent coupling

Engineering Contradiction:
Improvewaveguide arrangementVSAvoidcrosstalk
Core Design Contradiction:
Device complexityVSObject-generated harmful factors

Solution Approach 1:

The patent resolves crosstalk issues by moving the waveguide crossing from the 2D planar domain to the 3D space. The through waveguide is positioned at a different vertical level than the drop waveguide, allowing them to cross without electromagnetic coupling. This spatial separation in the z-dimension eliminates polarization-dependent crosstalk that would occur in planar crossings, while the overall device complexity remains manageable through systematic waveguide routing.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

This solution enables low-loss, polarization-independent optical switching with scalable high port counts, suitable for datacenter networks, reducing the need for external polarization splitters and minimizing crosstalk, thus overcoming the limitations of prior art.

Implementation Method 1

using adiabatic couplers for efficient switching

Methodology Applied
Scientific EffectAdiabatic coupling:

Data Source

PatentEP3510441B1Silicon-photonics-based optical switch with low polarization sensitivity
Publication Date: 2021.03.31 RGT UNIV OF CALIFORNIA
  • EP3510441B1 patent drawingFigure 1
  • EP3510441B1 patent drawingFigure 2A
  • EP3510441B1 patent drawingFigure 2B

AI summary

A polarization-independent optical switching system capable of rerouting light signals is disclosed. The system includes a plurality of switching cells, each including a pair of bus waveguides that are formed in different planes above a substrate. Each bus waveguide supports low-loss propagation of both the TE- and TM-polarization modes and are optically decoupled when the switch is in an unswitched state. In its switched state, a shunt waveguide that also supports low-loss propagation of both polarization modes is moved into proximity with both bus waveguides to form a pair of adiabatic directional couplers that enable the light signal to evanescently couple between each bus waveguide and the shunt waveguide. As a result, the path of a light signal through the switching cell is reconfigured.