Tolerancing of board-level-free-space optical interconnects

For optical interconnects to become a mature technology they must be amenable to electronic packaging technology. Two main obstacles to including free-space optical interconnects are alignment and heat-dissipation issues. Here we study the issues of alignment tolerancing that are due to assembly and manufacturing variations (passive-element tolerancing) over long board-level distances (>10 cm) for free-space optical interconnects. We also combine these variations with active optoelectronic device variations (active-element tolerancing). We demonstrate a computer-aided analysis procedure that permits one to determine both active- and passive-element tolerances needed to achieve some system-level specification, such as yield or cost. The procedure that we employ relies on developing a detailed design of the system to be studied in a standard optical design program, such as code v. Using information from this model, we can determine the integrated power falling on the detector, which we term optical throughput, by performing Gaussian propagation or general Fresnel propagation (if significant vignetting occurs). This optical throughput can be used to determine system-level performance criteria, such as bit-error rate. With this computer-aided analysis technique, a sensitivity analysis of all the variations under study is made on a system with realistic board-level interconnect distances to find each perturbation's relative effects (with other perturbations set to 0) on the power falling on the detector. This information is used to set initial tolerances for subsequent tolerancing analysis and design runs. A tolerancing analysis by Monte Carlo techniques is applied to determine if the yield or cost (yield is denned as the percentage of systems that have acceptable system performance) is acceptable. With a technique called parametric sampling, a subsequent tolerancing design run can be applied to optimize this yield or cost with little increase in computation. We study a design example and show that most of the tolerances can be achieved with current technology.     

Plastic modules for free-space optical interconnects

A combined optoelectronic and optomechanical packaging technique for the construction of snap-together free-space optical interconnect systems is described. The modules integrate relaying and routing functions by use of transparent optical molded plastic, which can achieve sufficient alignment precision that further adjustment is not required during system assembly. Methods to integrate the optoelectronic chips, such as vertical-cavity surface-emitting laser and receiver arrays with these plastic optical modules are described. Other chips can also be integrated to form optoelectronic multichip modules. These modules can also be designed to accommodate coupling to or from optical fiber arrays. A test-bed system to demonstrate the concept was assembled to a lower precision by use of conventional machining techniques.     

Diffractive optics applied to free-space optical interconnects

The optimum design of free-space optical interconnection systems utilizing diffractive optics is determined from a practical engineering standpoint for systems ranging from space invariant to fully space variant. System volume is calculated in terms of parameters such as the f-number of the diffractive lens, the wavelength of light, and also the total number, size, and separation of the optical sources and detectors. Performance issues such as interconnection complexity, diffraction efficiency, and signal-tonoise ratio are discussed. Diffractive optics fabricated by electron-beam direct-write techniques are used to provide experimental results for both shuffle-exchange and twin-butterfly free-space optical interconnects.     

Tolerance stackup effects in free-space optical interconnects

A numerical analysis indicates that tolerance stackup effects in free-space optical interconnects are significant even for short systems containing few components. Results prove that worst-case or root-sum-square analyses are not adequate to predict probable performance accurately. A Monte Carlo analysis must be performed.     

Misalignment correction for optical interconnects using vertical cavity semiconductor optical amplifiers

In board-to-board optical interconnects, the misalignment between the board and the backplane connections can cause both optical loss and interchannel cross talk. A vertical cavity semiconductor optical amplifier (VCSOA) is proposed to correct optical misalignment in an optical connector between the board and the backplane. Angular or lateral misalignment can be corrected with the designed module. The correction ability is determined by the acceptance angle of the VCSOA, which was characterized to be 9.4 degrees full angle at a 3 dB gain drop for a 30 microW optical signal at 1 GHz. The lateral misalignment correction ability is 0.16f, where f is the focal length of the mini lens to converge the input light onto the VCSOA.     

Practical realization of massively parallel fiber-free-space optical interconnects

We propose a novel approach to realizing massively parallel optical interconnects based on commercially available multifiber ribbons with MT-type connectors and custom-designed planar-integrated free-space components. It combines the advantages of fiber optics, that is, a long range and convenient and flexible installation, with those of (planar-integrated) free-space optics, that is, a wide range of implementable functions and a high potential for integration and parallelization. For the interface between fibers and free-space optical systems a low-cost practical solution is presented. It consists of using a metal connector plate that was manufactured on a computer-controlled milling machine. Channel densities are of the order of 100/mm(2) between optoelectronic VLSI chips and the free-space optical systems and 1/mm(2) between the free-space optical systems and MT-type fiber connectors. Experiments in combination with specially designed planar-integrated test systems prove that multiple one-to-one and one-to-many interconnects can be established with not more than 10% uniformity error.     

Path-reversed substrate- guided-wave optical interconnects for wavelength-division demultiplexing

A path-reversed substrate-guided-wave holographic interconnection scheme is investigated for a wavelength-division demultiplexing application. Using a beveled edge of a waveguiding plate allows optical signals to be coupled into the waveguiding plate and then to be coupled out of the plate by a waveguide hologram. Theoretical analyses are given for dispersion, bandwidth, and recording parameters of various guided-wave holographic gratings. A device is fabricated with a 45 degrees incident angle and a 45 degrees diffraction angle by use of a 20-microm photopolymer film. The 3-dB bandwidth of the device is measured to be 20 nm. Four-channel wavelength demultiplexing is demonstrated at 796, 798, 800, and 802 nm with no cross talk observed. A one-to-five cascaded four-channel wavelength-division demultiplexer with +/-5% energy uniformity under s polarization is also demonstrated to increase the user-sharing capacity. Twenty fan-out channels (5 x 4) are achieved experimentally.     

Multichip module with planar-integrated free-space optical vector-matrix-type interconnects

Even in the semiconductor industry, free-space optical technology is nowadays seen as a prime option for solving the continually aggravating problem with VLSI chips, namely, that the interconnect technology has failed to keep pace with the increase in communication volume. To make free-space optics compatible with established lithography-based design and fabrication techniques the concept of planar integration was proposed approximately a decade ago. Here its evolution into a photonic microsystems engineering concept is described. For demonstration, a multichip module with planar-integrated freespace optical vector-matrix-type interconnects was designed and built. It contains flip-chip-bonded vertical-cavity surface emitting laser arrays and a hybrid chip with an array of multiple-quantum-well p-i-n diodes on top of a standard complementary metal-oxide semiconductor circuit as key optoelectronic hardware components. The optical system is integrated into a handy fused-silica substrate and fabricated with surface-relief diffractive phase elements. It has been optimized for the given geometrical and technological constraints and provides a good interconnection performance, as was verified in computer simulations on the basis of ray tracing and in practical experiments.     

Performance-based adaptive power optimization for digital optical interconnects

Optical links are traditionally set to transmit maximum power for worst-case loss and consequently to dissipate more power than is required. We describe a technique to minimize power consumption based on the measured bit-error rate (BER) of the link. This technique uses a novel power-negotiation algorithm that optimizes the link power setting to achieve minimum power dissipation for a target BER. A 0.5 microm complementary metal-oxide semiconductor optical transceiver chip was fabricated, and a free-space optical interconnect system was built for validation. The results showed that the algorithm was able to find the optimum power settings for the VCSELs for a target BER and to account for dynamic changes such as variation in the optical loss in the system.     

Multicasting optical interconnects using liquid crystal over silicon devices

This work presents the characteristics and expected capabilities of an optical interconnect that uses a diffractive liquid crystal over silicon (LCOS) device as a routing element. Such an interconnect may be used in a neighborhood's optical network to distribute high definition television, thus avoiding an electronic or optical transmitter for each user. The optimal characteristics of the LCOS device are calculated in terms of pixel number and silicon area and found to be feasible with today's technology. Finally, its performance in terms of optical efficiency and number of output ports is evaluated and found suitable for a neighborhood with hundreds of households.     

Achromatic fan-out diffractive system for white-light free-space optical interconnects

Abstract

A simple and versatile white-light fan-out diffractive system based on the achromatization of the fractional Talbot effect is proposed. This achromatic configuration is able to interconnect a single polychromatic point source with a 2-D array of optoelectronic microdevices with low residual chromatic aberration even for white light. The whole broadband beamsplitter system is formed by two simple diffractive optical elements, a periodic diffractive lenslet array and a diffractive lens, that are made with a direct laser writing technique giving high light efficiency. The focal amplitude distribution corresponding to the lenslet array produces, by free-space propagation, self-replicas with different density of light points. These patterns, in conjunction with the achromatization process carried out by the additional diffractive lens, are, in short, the key to achieving a set of undistorted white-light spots at the output plane with high uniformity and variable separation between them. Experimental results are also shown.     

Substrate-mode holograms used in optical interconnects: design issues

We discuss a number of design issues that affect the performance tolerances of substrate-mode holograms used for optical interconnect systems. We examine the effects of emulsion uniformity, thickness variation, and index variation on the ability to determine the Bragg angle and the diffraction angle within the substrate accurately. The environmental stability with respect to temperature, laser irradiance, and humidity are considered. Experimental results are presented for substrate-mode holograms fabricated in spin-coated dichromated-gelatin emulsions. The coupling properties for a 1 × 2 multiplexed substrate-mode hologram with two superimposed gratings are also described.     

Angular sensitivities of volume gratings for substrate-mode optical interconnects

The angular sensitivities of slanted volume gratings (VGs) illuminated by three-dimensional (3-D) converging-diverging spherical Gaussian beams for substrate-mode optical interconnects in microelectronics are analyzed by application of 3-D finite-beam rigorous coupled-wave analysis. Angular misalignments about the z, y, and x axes that correspond to yaw, pitch, and roll misalignments resulting from manufacturing tolerances of chips are investigated. Two cases of linear polarization of the central beam of the Gaussian are considered: E perpendicular K and H perpendicular K, where K is the grating vector. From worst-case manufacturing tolerances, the ranges of yaw, pitch, and roll misalignment angles are alpha = +/-1.17 degrees, beta= +/-3.04 degrees, and gamma = +/-3.04 degrees, respectively. Based on these ranges of misalignment angles, the decreases of diffraction efficiencies for slanted VGs that are due to both the yaw and the roll misalignments are relatively small. However, the efficiency of substrate-mode optical interconnects achieved by slanted VGs could be reduced by 61.04% for E perpendicular K polarization and by 58.63% for H perpendicular K polarization because of the pitch misalignment. Thus the performance of a VG optical interconnect is most sensitive to pitch misalignment.     

Relay lens-free birefringence-customized stackable beam-steering modules for optical interconnects

A previously suggested birefringence-customized modular optical interconnect technique is extended for lens-free relay operation. Various lens-free relay imaging models are developed. We claim that the lens-free relay system is important in simplifying an optical interconnect system whenever the imaging conditions permit. To verify the validity of various proposed concepts, we experimentally implemented some 8 x 8 optical permutation modules. High-power efficiency and low channel cross talk were experimentally observed. In general, the larger the channel spacing, the less the cross talk. A quantitative cross-talk measurement of the lens-free relay system shows that, for a fixed channel width of 0.5 mm and channel spacings of 0.5, 1, and 2 mm, a less than -20-dB cross-talk performance can be guaranteed for lens-free relay distances of 40, 280, and 430 mm, respectively.     

Misalignment tolerance analysis of free-space optical interconnects via statistical methods

System-level packaging is one of the critical issues that need to be addressed for free space optical interconnections (FSOI) to become useful in desktop systems. The performance of FSOI, e.g., in terms of system bit-error rate, is greatly affected by misalignments in the optical system. Therefore tolerancing, i.e., the ability to analyze and predict the effects of misalignments in the system, is of prime importance to system designers. We introduce an approach in which we study the effects of optical misalignments and other tolerance factors using statistical methods. We use Monte Carlo simulations, design of the experiments, and regression techniques to fit a polynomial equation that expresses the relationship between the system performance and the tolerance factors. This prediction model can be used for design, cost optimization, and quality control purposes. In addition, we perform a sensitivity analysis to determine those tolerance variables that have the greatest effect on system performance.     

Fan-out routing and optical splitting techniques for compact optical interconnects using single-mode polymer waveguides

Polymer waveguide (WG) S-bends are necessary for fan-out routing techniques and optical splitting in high-density optical interconnects. Designing and manufacturing of optimal S-bends are critical for minimizing optical link loss while maintaining overall size and layout constraints. Complete structural loss analysis is demonstrated theoretically and shown experimentally utilizing both radial and transitional loss in single-mode (SM) polymer WG radial arc, cosine, and raised-sine S-bend profiles. SM polymer WG straights were first fabricated to measure standard propagation loss. SM WG S-bends were fabricated incorporating straight lead-in and lead-out sections to incorporate transitional loss present in workable designs. S-bend designs were measured at different dimensions and matched to theoretical losses. Compact cosine and radial arc S-bends exhibited the lowest structure loss for low and high NA WGs, respectively. High-speed performance of SM WG straights and S-bends was measured at 10 Gbit/s demonstrating low error rate. Optical splitters designed with S-bends and tapers were also evaluated and fabricated. Trade-off between optimal loss and minimal device size is discussed.     

Wavelength response of waveguide volume grating couplers for optical interconnects

The wavelength response of a waveguide volume grating coupler (WVGC) is analyzed for coupling light from a slab waveguide into the superstrate. A leaky-mode approach is used in conjunction with rigorous coupled-wave analysis. A quantitative theoretical study of the effect of index modulation, waveguide index, and grating thickness on the wavelength bandpass of a WVGC is also presented. The FWHM wavelength bandpasses found for high-efficiency couplers range from 173 to 525 nm. The various Bragg conditions that can be used in designing a WVGC are also presented and compared. The use of the propagation constant of the mode being outcoupled as the incident wave vector in the Bragg condition is shown to produce the highest coupling efficiency.     

Error correction for free-space optical interconnects: space-time resource optimization

We study the joint optimization of time and space resources withinfree-space optical interconnect (FSOI) systems. Both analyticaland simulation results are presented to support this optimization studyfor two different models of FSOI cross-talk noise: diffraction froma rectangular aperture and Gaussian propagation. Under realisticpower and signal-to-noise ratio constraints, optimum designs based onthe Gaussian propagation model achieve a capacity of 2.91 x10(15) bits s(-1) m(-2), while therectangular model offers a smaller capacity of 1.91 x10(13) bits s(-1) m(-2). We alsostudy the use of error-correction codes (ECC) within FSOIsystems. We present optimal Reed-Solomon codes of various length, and their use is shown to facilitate an increase in both spatialdensity and data rate, resulting in FSOI capacity gains in excess of8.2 for the rectangular model and 3.7 for the Gaussian case. Atolerancing study of FSOI systems shows that ECC can provide toleranceto implementational error sources. We find that optimally codedFSOI systems can fail when system errors become large, and we present acompromise solution that results in a balanced design in time, space, and error-correction resources.     

Parallel optical interconnects may reduce the communication bottleneck in symmetric multiprocessors

We start with a detailed analysis of the communication issues in today's symmetric multiprocessor (SMP) architectures to study the benefits of implementing optical interconnects (OI) in these machines. We show that the transmission of block addresses is the most critical communication bottleneck of future large SMPs owing to the need to preserve the coherence of data duplicated in caches. An address transmission bandwidth as high as 200-300 Gb/s may be necessary in ten years from now; this requirement will represent a difficult challenge for shared electric buses. In this context we suggest the introduction of simple point-to-point OIs for a SMP cache-coherent switch, i.e., for a VLSI switch that would emulate the shared-bus function. The operation might require as much as 10,000 input-outputs (IOs) to connect 100 processors, particularly if one maintains the present parallelism of transmissions to preserve a large bandwidth and a short memory access latency. The interest for OIs comes from the potential increase of the transmission frequency and from the possible integration of such a high density of IOs on top of electronic chips to overcome packaging issues. Then we consider the implementation of an optical bus that is a multipoint optical line involving more optical technology. This solution allows multiple simultaneous accesses to the bus, but the preservation of the coherence of caches can no longer be maintained with the usual fast snooping protocols.     

Modeling diffraction in free-space optical interconnects by the mode expansion method

Free-space optical interconnects (FSOIs), made up of dense arrays of vertical-cavity surface-emitting lasers, photodetectors and microlenses can be used for implementing high-speed and high-density communication links, and hence replace the inferior electrical interconnects. A major concern in the design of FSOIs is minimization of the optical channel cross talk arising from laser beam diffraction. In this article we introduce modifications to the mode expansion method of Tanaka et al. [IEEE Trans. Microwave Theory Tech. MTT-20, 749 (1972)] to make it an efficient tool for modelling and design of FSOIs in the presence of diffraction. We demonstrate that our modified mode expansion method has accuracy similar to the exact solution of the Huygens-Kirchhoff diffraction integral in cases of both weak and strong beam clipping, and that it is much more accurate than the existing approximations. The strength of the method is twofold: first, it is applicable in the region of pronounced diffraction (strong beam clipping) where all other approximations fail and, second, unlike the exact-solution method, it can be efficiently used for modelling diffraction on multiple apertures. These features make the mode expansion method useful for design and optimization of free-space architectures containing multiple optical elements inclusive of optical interconnects and optical clock distribution systems.