CN102713703A - Waveguide optically pre-amplified detector with passband wavelength filtering - Google Patents

Abstract

The present invention describes integrated photonic arrangements achievable in multi-guided vertical integration (MGVI) structures composed of III-V semiconductors and grown in one epitaxial growth flow, thereby allowing the formation of multiple vertically integrated wavelength-specific waveguides of MGVI structures The common wavelength specifies the integration of semiconductor optical amplifier (SOA) and PIN photodetector (PIN) structures in the waveguide. Integration includes integrating a wavelength filter between the SOA and the PIN to reduce noise occurring in the PIN of the ASE generated by the SOA. In an exemplary embodiment of the invention, the wavelength filter is integrated into a common wavelength-specifying waveguide or into an MGVI structure in a wavelength-specifying waveguide. Furthermore, in other embodiments, the wavelength filter is provided by a thin film filter abutting the end face of the integrated photonic arrangement, where the optical signal is coupled by an optical waveguide and/or additional optical elements such as a multimode interference device.

Description

Waveguide Optical Preamplified Detector with Passband Wavelength Filtering

technical field

The present invention relates to waveguide photonic devices and photonic integrated circuits, and more particularly to waveguide optical preamplified detectors in III-V semiconductor materials.

Background technique

The requirement for deep penetration of fiber into the access network is driving a parallel mass deployment of optical interface equipment for traffic to and from subscribers. For example, optical transceivers that receive downstream signals on one wavelength and transmit upstream signals on another wavelength (both wavelengths share the same fiber) must be deployed at each Optical Line Terminal (OLT)/Optical Network Unit (ONU). Therefore, cost-efficiency and batch scalability in manufacturing such components are increasingly becoming major issues. It is widely accepted in the telecommunications industry that optical access solutions will not become a commodity service until volume manufacturing of optical transceivers and other mass-deployed optical components reaches consumer product levels of cost efficiency and scalability.

In the framework of the current optical component manufacturing paradigm based primarily on off-the-shelf discrete passive photonic devices and off-the-shelf optical component assembly (OSA) of active photonic devices, the root cause of the problem lies in labor-intensive optical alignment and costly Packed in multiple quantities. Not only do they limit cost efficiency, but they also greatly limit a manufacturer's ability to increase throughput and provide scalability in manufacturing. The solution lies in reducing the optical alignment and packaging content in the OSA, and eventually replacing the optical assembly with photonic integrated circuit (PIC) technology, where all functional elements of the optical circuit are monolithically integrated on the same substrate. Manually performed active optical alignment is then replaced by automated passive alignment defined by lithography and multi-component packaging is completely eliminated, enabling complex optical components based on existing planar and semiconductor wafer fabrication techniques. Automated and batch-scalable mass production.

In the context of the application, the materials of choice for monolithic PICs used in optical transmission systems remain indium phosphide (InP) and its related III-V semiconductors, as they uniquely allow operation in the spectral range of interest for optical telecommunications The active and passive devices in are combined on the same InP substrate. Specifically, InP PICs may be a cost-effective and volume-scalable solution for the most mass-deployed components: optical transceivers operating in the 1.3-μm and 1.5-μm wavelength ranges for access to passive optical networks, see For example V. Tolstikhin ("Integrated Photonics: Enabling Optical Component Technologies for Next Generation Access Networks", Proc. Asia Optical Fiber Communication & Optoelectronic Exposition & Conference, October 2007).

Within each optical transceiver is an optical photodetector that converts a received optical signal into an electrical signal, allowing this received signal to be provided to electrical equipment connected to a telecommunications network, such as with voice over IP ( VOIP) telephone, computer or digital TV set-top box. This class of photodetectors is designed as a PIN diode with low reverse voltage bias, with a lightly doped "near" intrinsic semiconductor region between the p-type and n-type semiconductor regions, or as a high reverse voltage Biased avalanche photodiode (APD). The compatibility of PIN diodes with standard CMOS electronics, typical reverse bias of a few volts instead of tens of volts with APDs, low capacitance, and high bandwidth operation have made PIN diodes the preferred choice in network deployments.

As mentioned earlier, PICs are the best hope for realizing the cost-effective and batch-scalable solutions needed for access network transceivers. In a monolithic PIC, the PIN diode is implemented in a waveguide structure, resulting in a waveguide photodetector (WPD), which is compatible with the passive waveguide circuit of the PIC and thus facilitates photodetectors with passive wavelength decomplexing monolithic integration with and routing elements. Accordingly, the requirement for PIC-compatible, high-performance and still inexpensive PIN WPDs has grown further and is essential for such fiber penetration to the subscriber customer base and the resulting PIC penetration to the access communication system.

While drivers for implementing such PIN WPDs are particularly evident in access networks, it should be understood that they are particularly attractive general-purpose devices at high bit rates, where surface-illuminated detectors suffer from carrier transit time absorption efficiency The tradeoffs are limited and limited in PICs, and any non-waveguide devices in them are difficult to integrate.

其中R是总量子效率，e为电子电荷，以及是光子能量。然而，主要取决于装置设计的芯片上PIN WPD中的η值能够达到相当大的70%，参见例如V.Tolstikhin的“One-Step Growth Optical Transceiver PICs in InP”(Proc.ECOC 2009，2009年9月20-24日，Paper 8.6.2)，它仍然始终小于一，并且因此任何PIN检测器的响应率基本上低于

同时，当今光网络发展的一个明显趋势是要求接收器端的越来越高的响应率。例如，在接入PON的情况下，由网络运营商所推动的不变目标是朝向更高的分流比和更长的延伸架构，因为这降低每个订户的中心局设备和操作成本，由此使其能够向最终客户提供更低价格。因此，一些PON标准，例如GPON B+(ITU-T G.984.2)，已经要求检测器响应率对于任何可想像的跨阻抗放大器(TIA)为高于

光电检测器对其通常加载到接收器电路中。显然，这个要求采用任何PIN光电检测器、甚至单独的PIN WPD无法满足，这通常具有更高的插入损失并且因此具有比其表面照明对应体略低的量子效率。A key performance parameter of any photodetector is the responsivity defined as the induced photocurrent relative to the incident light power. It is measured in Amps/Watt (A/W) and can be expressed as

where R is the total quantum efficiency, e is the electron charge, and is the photon energy. However, values of η in on-chip PIN WPDs can reach considerable values of 70%, depending mainly on device design, see e.g. "One-Step Growth Optical Transceiver PICs in InP" by V. Tolstikhin (Proc. ECOC 2009, 9 20-24, Paper 8.6.2), it is still always less than one, and thus the response rate of any PIN detector is substantially lower than

At the same time, an obvious trend in the development of today's optical networks is to require higher and higher response rates at the receiver side. For example, in the case of access PON, constant goals driven by network operators are towards higher split ratios and longer stretch architectures, as this reduces central office equipment and operating costs per subscriber, thereby enabling it to offer lower prices to end customers. Therefore, some PON standards, such as GPON B+ (ITU-T G.984.2), already require detector responsivity for any conceivable transimpedance amplifier (TIA) to be higher than

A photodetector is usually loaded into the receiver circuit. Clearly, this requirement cannot be met with any PIN photodetector, or even a PIN WPD alone, which typically has higher insertion loss and thus a slightly lower quantum efficiency than its surface-illuminated counterpart.

To achieve a total quantum efficiency η > 1, some form of gain must be added between the incoming signal from the fiber and the receiver circuitry. There are three on-chip solutions for this:

a) Electrical gain after detection, for example by using phototransistors, where the signal is amplified while in the electrical domain, which is not exactly a waveguide-based PIC-compatible solution, and actually requires a photonic integrated circuit (integrated onto the same substrate active and passive waveguide-based photonic devices) to optoelectronic integrated circuits (integrated on the same substrate with active and passive waveguide-based photonic devices) at the expense of a substantially more complex and expensive fabrication process electronic device);

b) Electrical gain in the detection process, for example by utilizing an avalanche photodiode (APD), where the signal is amplified while being converted from the optical to the electrical domain, which is fundamental in terms of gain-bandwidth product, especially in its waveguide-based implementation are limited and therefore not well suited for integration into PICs for most networking applications; and

c) optical gain before detection, for example in semiconductor optical amplifiers (SOAs), where the signal is amplified without leaving the optical domain, a waveguide-based solution compatible with the rest of the PIC design and fabrication process; following It is called an optical preamplified detector (OPAD).

光纤耦合响应性的适当PIC解决方案，被定义为PIC传递到接收器电路中的电流相对于光信号传递到PIC的光功率。这种解决方案没有具体速度限制(除非SOA处于饱和情况下，并且其光增益受到经放大的光信号影响)，并且能够提供数十的端对端增益，由此实现优良增益带宽积。由于这些原因，高功能PIC兼容OPAD装置的设计近年来引起了极大关注。Due to its compatibility with waveguide-based PIC architectures and manufacturing processes, the OPAD appears to be superior to

The proper PIC solution for fiber-coupled responsiveness is defined as the current delivered by the PIC into the receiver circuit relative to the optical power delivered by the optical signal to the PIC. This solution has no specific speed limit (unless the SOA is saturated and its optical gain is affected by the amplified optical signal) and can provide tens of tens of end-to-end gains, thereby achieving a good gain-bandwidth product. For these reasons, the design of highly functional PIC-compatible OPAD devices has attracted considerable attention in recent years.

Any integrated OPAD is generally a waveguide based device that combines a gain waveguide section (where optical amplification takes place) and a detection waveguide section (where optical conversion to the electrical domain takes place) that passes through to/from the two elements of the OPAD Passive waveguide optical connection of optical signals. Monolithic integration of multiple waveguide devices with different waveguide core regions made of different semiconductor materials, such as optical amplifiers (OA) and photodetectors (PD) required for OPADs, can be basically done in the following three ways: One of the implementations:

1. Direct butt coupling; this utilizes the ability to perform multiple steps of epitaxial growth, including selective area etch and regrowth, in order to provide multiple semiconductor materials that are spatially differentiated horizontally from a common vertical plane across the PIC die, and different semiconductor materials are grown horizontally adjacent such that waveguides are formed in each direct abutment with each other to form a transition from one material to another;

2. Modified docking coupling; this exploits selective area post-growth modification of semiconductor material, for example by quantum well hybrid technology, rather than etch and regrowth, to form yet spatially differentiated in a vertically oriented common plane across the PIC die area of desired semiconducting material; and

3. Evanescent Field Coupling; where vertically separated but still optically coupled waveguides featuring different semiconductor materials in the core region are used to provide the required material differentiation without additional growth steps such that it is now in the common vertical of the PIC die Distinguished in the stack.

Examples of prior art can be found in each of these three categories. For example, by Haleman et al. in U.S. Patent 5029297 "Optical-Amplifier-Photodetector Device", W. Rideout et al. in U.S. Patent 5299057 "Monolithically Integrated Optical Amplifier and Photodetector Tap" and J.Walker et al. in U.S. Patent 6909536 "Optical Receiver Indcluding a Linear Semiconductor Optical Amplifier" reported an integrated OPAD device using direct butt-coupling. An example of modified butt coupling is presented by M. Aoki et al. in US Patent 5574289 "Semiconductor Optical Integrated Device and Light Receiver Using Said Device". Finally, an integrated OPAD based on evanescent field coupling in a vertical twin waveguide structure was reported by S. Forrest et al. in US Patent 7343061 entitled "Integrated Photonic Amplifier and Detector".

Each of these design solutions has its benefits and drawbacks. Consider direct butt-coupling, which allows planar integration with minimal vertical topology, which is an advantage from a planar technology point of view since no or minimal planarization is required in handling PICs during fabrication. However, direct butt-coupling requires multiple epitaxial steps to provide multiple semiconductor materials, which not only creates difficulties in managing optical reflections from these material interfaces, but also significantly impacts manufacturing yield and thereby significantly increases the cost of the final PIC device. Modified butt-coupling could potentially remove an extra epitaxy step and improve fabrication yield in this way, but its capabilities are limited in relation to possible semiconductor material modification: typically, only the bandgap of the quantum well layer can be shifted up to 100 nm blue, while other layers, such as heavily doped contact layers, which are required in the active waveguide section but are highly undesirable in the passive waveguide section because of the propagation losses they generate, remain unchanged. In contrast, evanescent field coupling does not have the disadvantages of the butt-coupling approach described above, however, since it is based on vertical rather than planar integration, it does so by requiring multiple etch steps at different vertical levels and creating an increased vertical topology, instead A slightly more complex manufacturing process based on planar technology.

要大10倍的光纤耦合响应度，其中对于50nm波长带宽的极化灵敏度小于0.4dB，在室温下工作在1490nm附近的OPAD中注入电流大约为150mA。Therefore, evanescent field coupling is the only practical way that can be realized in one-step epitaxial growth without any post-growth modification of semiconductor material, and thus offers the possibility of highest fabrication yield in combination with cost-effective fabrication process, and accordingly Lowest cost possible for PIC devices. It also provides a direct solution to the design of integrated OPADs based on twin waveguide structures, where the lower waveguide of the two vertically coupled waveguides is a passive waveguide with a core layer bandgap much higher than the photon energy of the optical signal expected to be sent to the OPAD , allowing low-loss propagation, while the upper waveguide of the two vertically coupled waveguides is a PIN structure with an intrinsic material bandgap close to the spectral range of the optical signal to be processed by the OPAD. This upper waveguide is an active waveguide with optical amplification (under forward electrical bias) or detection (under reverse electrical bias) capability for the spectral range of interest. Optical coupling between the two waveguides can be achieved with optional lateral tapers to facilitate smooth and controllable vertical transitions of the guided light signal. In this way, with proper waveguide and transverse tapered design, the optical signal can be adiabatically transferred from the amplification waveguide section to the detection waveguide section via the passive waveguide section in between, in this case, there is no intrinsic active layer and The passive waveguide section with the upper contact layer but present with the lower contact layer also serves as electrical isolation between the forward (amplification) and reverse (detection) bias sections of the waveguide PIN. For example by K.-T.Shiu et al. in "A Simple Monolithically Integrated Optical Receiver Consisting of an Optical Preamplifier and pin Photodiode" (Photon. Technol. Lett., Vol. 18, pp. 956-958, April 2006) and This approach was reported by V. Tolstikhin et al. in "Optically Pre-Amplified Detectors for Multi-GuideVertical Integration in InP" (Proc. Indium Phosphide and Related Materials, 2009, pp. 155-158, Newport Beach, 2009). Reported by Tolstikhin et al.

To have a fiber-coupled responsivity 10 times larger, where the polarization sensitivity for a 50nm wavelength bandwidth is less than 0.4dB, the injection current in an OPAD operating near 1490nm at room temperature is about 150mA.

Twin-guided integration of active and passive waveguides is the simplest and most common example of vertical integration based on evanescent fields, and can be realized in various forms, e.g. based on conventional directional couplers (DC) (see e.g. Y. "Integrated Twin-Guide AlGaAs Laser with Multiheterostructure" by Suematsu et al. (IEEE J. Quantum Electron., Vol. 11, pp. 457-460, July 1975)) or enhanced by an impedance matching layer between coupled optical waveguides (see e.g. "Impedance Matching for Enhanced Waveguide/Photodetector Integration" by R.J.Deri et al. (App. Phys. Lett., Vol. 55, pp. 2712-2714, Dec. 1989) - or with DC for lateral taper assisted coupling see e.g. "Efficient Coupling in Integrated Twin-Waveguide Lasers using Waveguide Tapers" by P.V. Studenkov et al. (IEEE Photon. Technol. Lett., Vol. 11, pp. 1096-1098, Nov. 1999) ) in phase matching.

Multi-guided vertical integration (MGVI) is an extension of this approach towards multifunctional PICs, in which optical waveguides with different functions are stacked vertically in ascending order of mutually coupled evanescent fields and waveguide bandgap wavelengths, and thermal insulation between vertically arranged waveguides The transition is effected by lateral tapers defined at the necessary vertical guide levels and acting in relation to each other, see e.g. "Laterally-Coupled DFB Lasers for One-Step Growth Photonic Integration in InP" by V. Tolstikhin et al. (IEEE Photon. Technol .Lett., Vol.21, pp. 621-623, May 2009), V. Tolstikhin et al. "Optically Pre-Amplified Detectors for Multi-Guide Vertical Integration in InP" (Proc.Indium Phosphide and Related Materials 2009 Conference, pp. 155-158, Newport Beach, 2009) and also V. Tolstikhin et al. entitled "Integrated-Optics Arrangement for Wavelength (De) Multiplexing in a Multi-Guide Vertical Stack".

A key feature that distinguishes the MGVI approach from the previously described prior art sequential twin-guided integration in the same multiguided vertical stack is the multifunctional PIC with more than two optical waveguides overlapping the stack and evanescent field coupling The ability for optical signals in the waveguides to be transferred adiabatically between these waveguides by means of transverse tapers defined in at least a portion of the vertically guided stages and acting in relation to each other in use. This may be suitable for parallel adiabatic transfer, as opposed to serial adiabatic transfer, where no more than two vertically stacked guides are simultaneously evanescently field coupled, and if the PIC structure has more than two functions and thus more than two directed vertical stages, The transition of the optical signal between them is then achieved by a continuous transition between two adjacent waveguides, excluding all other guiding layers in the process. Examples of such parallel and serial approaches to evanescent field-based integration in multi-guided vertical stacks are respectively V. Tolstikhin et al. in U.S. Patent 7,532,784 entitled "Integrated Vertical Wavelength (De) multiplexer" and S. Forrest et al. are given in US Patent 6,795,622 entitled "Photonic Integrated Circuits".

Regardless of the specific active-passive waveguide integration technique (i.e. planar butt coupling or vertical evanescent field coupling) or its specific implementation (e.g. parallel or serial approach for vertical integration based on evanescent field coupling), any OPAD device basically Gain-enhanced responsivity should be provided without significant degradation of the signal-to-noise ratio. Among other things, as a component to be used in a receiver for signal transfer from the optical to the electrical domain, an OPAD should ideally combine high gain with low noise. The main noise source specific to OPADs that adds to other, but not general, noise sources such as thermal and shot noise in receiver circuits is amplified spontaneous emission (ASE) generated in the amplifying section of the OPAD. Inherent ASE independent of monolithic design in optical amplifiers such as OPADs, hybrid or fiber-based such as Erbium-Doped Fiber Amplifiers (EDFAs). In case the ASE related noise becomes the main contributor to the receiver noise, the optical signal amplification provided by the OPAD won't help much as it makes the signal to noise ratio worse and ultimately the receiver sensitivity worse despite increasing its response rate. This aspect of OPAD performance is critical to device applications, especially in extended extension/increased split ratio PONs, but has not been adequately addressed in prior art OPAD designs.

In order to better understand the impact that ASE can have on receiver sensitivity and ways to reduce it, it is helpful to consider the current fluctuations in the receiver circuit produced by ASE. Neglecting all noise sources except thermal noise (usually determined by the equivalent input noise of the transimpedance amplifier on which the detector is loaded), an estimate of the current mean square value of the induced photocurrent is given by equation (1) written as:

是相对于检测段前面的光功率的响应率，E_ASE是在检测段输入端的ASE功率的谱密度，B_e是接收器电路带宽，B_o≈(cΔλ_PBF)/λ²是相当于从放大到检测波导段的转变中的光波长通带Δλ_PBF的频率带宽，G是波导折算（waveguide-referred）合计增益，以及P是信号的时间平均波导耦合光功率。where _iD is the RMS noise current in the receiver circuit produced by a device with a similar PIN detector but without an optical amplifier, and the second term on the right accounts for the excess ASE-related noise produced by the optical preamplifier, which arises from Combinations of spontaneous-spontaneous and spontaneous-signal beat frequencies represented by the first and second terms in brackets on the right side of this equation, respectively (eg NAOllson, J. Lightwave Technol., Vol. 7, pp. 1071-1082, July 1991). it's here,

is the responsivity relative to the optical power in front of the detection section, E _ASE is the spectral density of the ASE power at the input of the detection section, Be is the receiver circuit bandwidth, and _{B o ≈} (cΔλ _PBF )/λ ² is equivalent to amplifying from where G is the waveguide-referred _aggregate gain, and P is the time-averaged waveguide-coupled optical power of the signal.

If the receiver noise is dominated by sources other than ASE, i.e., the first term of Equation 1 dominates, then the waveguide coupling sensitivity is estimated as shown in Equation (2):

在其它情况下，当等式(1)的第二项主导时，即，在ASE噪声限制情况中，波导耦合接收器灵敏度能够如下式(3)所示来近似计算，以便近似计算最小光功率

where Q is the Q factor under the following assumptions: the noise is Gaussian; the receiver decision circuit threshold is set to give an equal probability of error for either a 1 or a 0 bit of the data signal (see "Fiber-Optic Communication Systems" by G. Agrawal, p. Second revised edition, Wiley, 1997), and the average power P ₁ in 1 bits is much higher than P ₀ in 0 bits, that is,

In other cases, when the second term of equation (1) dominates, i.e., in the ASE noise-limited case, the waveguide-coupled receiver sensitivity can be approximated as shown in equation (3) below, in order to approximate the minimum optical power

$\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}}<span\; class="notranslate">\; \&ap;</span>\mathrm{<span\; class="notranslate">\; h\&nu;</span>}<span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; g</span>}}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; Q</span>}<span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Q</span>}<span\; class="notranslate">\; +</span>\sqrt{\frac{{\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; o</span>}}}{{\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}}}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$

where hν is the photon energy and _Fg is the noise factor of the OPAD amplification section (see "Sensitivity of Optically Preamplified Receivers with Optical Filtering" by RC Steele et al. (IEEE Photon. Technol. Lett., Vol. 3, pp. 545-547) , June 1991).

如从等式(2)能够看到，以及由此改进接收器性能。其次，在ASE噪声受限情况中，即，当合计增益变高时，增益的进一步增加没有有益效果，因为这引起

的饱和，如从等式(3)能够看到。第三，能够通过将波长滤波器插入OPAD段的放大与检测段之间来抑制与自发-自发拍频关联的ASE噪声的至少一部分，使得滤波器的通带足够宽到以允许通过所有信号波长，但是同时比自发-自发拍频带宽要窄。Equations (1) to (3) provide a useful insight into the limits of OPAD performance and optimization. First, as long as the receiver noise is determined by something other than ASE, i.e., when the aggregate gain is relatively low, an increase in gain decreases

As can be seen from equation (2), and thereby improves receiver performance. Second, in the ASE noise-limited case, i.e., when the aggregate gain becomes high, further increases in gain have no beneficial effect, since this causes

Saturation of , as can be seen from equation (3). Third, at least a portion of the ASE noise associated with spontaneous-spontaneous beats can be suppressed by inserting a wavelength filter between the amplification and detection sections of the OPAD section such that the passband of the filter is wide enough to allow passage of all signal wavelengths , but at the same time narrower than the spontaneous-spontaneous beat bandwidth.

In this way, optical signals in a predetermined narrow wavelength range pass through and are detected in the photodetector section, while ASE noise does not. It can be rerouted away from the detection section of the OPAD, or absorbed in an intermediate PIC circuit before the detection section, or both, so that the ASE-related OPAD noise is limited to the expected wavelength range of the received signal.

Accordingly, the present invention provides an improvement in OPADs by providing an MGVI compliant design solution featuring a passband filtering between amplification and detection of the received optical signal. In this way, performance improvements are combined with the capabilities and advantages of one-step epitaxially grown MGVI technology, thereby providing, for example, a high-function and low-cost PIC solution for OPAD-based receivers for mass deployment in extended extension/increased split ratio PONs

purpose of invention

The object of the present invention is an integrated OPAD design compatible with the MGVI platform by providing on-chip ASE filtering outside the signal wavelength range and thus reducing the impact of ASE-related noise on the sensitivity of OPAD-based receivers while providing greater than receiver response rate to enhance device performance. The ASE filtering OPAD is formed in the MGVI platform so that in use, the amplification and detection waveguide sections are formed in the same waveguide designated active waveguide layer, while the passive waveguide section and elements of the waveguide circuit are defined in the passive waveguide layer, without The source waveguide layer is positioned below the active waveguide layer in the multi-guide vertical stack, which may also include other waveguide-assigned active and passive waveguide layers. All elements of the MGVI photonic circuit are realized in one epitaxial growth step and monolithically integrated on one substrate. According to the MGVI design principle, see U.S. Patent 7,532,784 titled "Integrated Vertical Wavelength (De) Multiplexer" by V. Tolstikhin et al. and 7,444,055 titled "Integrated Optics Arrangement for Wavelength (De) Multiplexing in a Multiple-Guide VerticalStack". Band wavelength filters can be implemented internally in the MGVI structure or external to the MGVI structure.

Contents of the invention

It is an object of the invention to obviate or alleviate at least one disadvantage of the prior art.

According to another embodiment of the present invention, a photonic component is provided, comprising:

a) an epitaxial semiconductor structure, when the substrate includes at least one of a commonly assigned waveguide for supporting propagation of an optical signal in a predetermined first wavelength range and a plurality of wavelength assigned waveguides vertically arranged in order of increasing wavelength band gap, at growing in a III-V semiconductor material system in a single growth step, wherein each of the plurality of wavelength-specifying waveguides supports a predetermined second wavelength range, each of the predetermined second wavelength range being in the predetermined first wavelength range;

b) an optical input port for receiving optical signals in the first wavelength range;

c) a first filter comprising at least a first output port and a second output port and characterized by at least a first passband width, the filter is optically coupled to receive optical signals in a first wavelength range and for providing an optical input port with a first predetermined portion of the received optical signal to a first output port, the first predetermined portion of the received optical signal being determined based on at least a first passband width;

d) an optical amplifier comprising at least a gain section formed in one of a plurality of wavelength-specifying waveguides, a first contact for forward biasing the optical amplifier, and a third output port, the optical amplifier being optically coupled to the first an output port for receiving a first predetermined portion of the received optical signal and providing an amplified filtered optical signal to a third output port;

e) a second filter comprising at least a fourth output port and a fifth output port and characterized by at least a second passband width, the filter is optically coupled to the third output port of the optical amplifier and is used to output to the fourth The port provides a first predetermined portion of the amplified filtered optical signal and provides a second predetermined portion of the amplified filtered optical signal to the fifth output port, the first and second predetermined portions of the amplified filtered optical signal according to at least the second The passband width is determined;

f) a first photodetector optically comprising at least a second contact for reverse biasing the first photodetector coupled to the first photodetector for receiving the amplified filtered optical signal a fourth output port of the predetermined portion of the second filter;

g) a second photodetector optically coupled to a fifth output port of a second filter for receiving a second predetermined portion of the amplified filtered optical signal; and

h) a third photodetector optically coupled to the second output port of the first filter for receiving a predetermined portion of the optical signal propagating from the optical amplifier to the first filter, the predetermined portion of the optical signal according to at least the first filter A passband width is determined; where,

The first contact and the second contact are formed on the same layer of the epitaxial semiconductor structure, but are electrically isolated from each other.

According to another embodiment of the present invention, there is provided

A photonic assembly comprising:

a) an epitaxial semiconductor structure, when the substrate includes at least one of a commonly assigned waveguide for supporting propagation of an optical signal in a predetermined first wavelength range and a plurality of wavelength assigned waveguides vertically arranged in order of increasing wavelength band gap, at grown in a III-V semiconductor material system grown in a single growth step, wherein each of the plurality of wavelength-specifying waveguides supports a predetermined second wavelength range, each of the predetermined second wavelength ranges being in the predetermined first wavelength range;

b) an optical input port for receiving optical signals in the first wavelength range;

c) an optical amplifier comprising at least a gain section formed in one of a plurality of wavelength-specifying waveguides, a first contact for forward biasing the optical amplifier, and a first output port, the optical amplifier being optically coupled to a an optical input port that receives an optical signal and provides an amplified optical signal to the first output port;

d) a first filter comprising at least a second output port and characterized at least by a first passband width, the filter is optically coupled to the first output port of the optical amplifier and is used to provide the amplified a first predetermined portion of the optical signal, the first predetermined portion of the amplified optical signal is determined based on at least a first passband width;

e) a first photodetector optically comprising at least a second contact for reverse biasing the first photodetector, the first photodetector being coupled to a first predetermined contact for receiving an amplified optical signal Part of the second output port of the first filter; where

The first contact and the second contact are formed on the same layer of the epitaxial semiconductor structure, but are electrically isolated from each other.

Those skilled in the art will clearly understand other aspects and features of the present invention by reading the following description of specific embodiments of the present invention in conjunction with the accompanying drawings.

Description of drawings

Embodiments of the invention are now described, by way of example only, with reference to the accompanying drawings, in which:

Figure 1A shows a prior art OPAD according to Tolstikhin et al. "Optically Pre-Amplified Detectors for Multi-Guide Vertical Integration in InP" (Proc. IPRM 2009);

Figure 1B shows a schematic diagram of the functionality provided by Tolstikhin's prior art on OPAD;

Figure 2 shows the Q factor of an OPAD receiver versus optical gain with varying filter bandwidth;

Figure 3 shows a schematic diagram of the functionality provided by the present invention;

Figure 4A shows an OPAD according to one embodiment of the present invention, where wavelength filtering is achieved by a thin-film filter with a reflective interface and a waveguide interface with circuit edges;

Figure 4B shows an OPAD according to one embodiment of the present invention, wherein wavelength filtering is achieved by a reflective interface and a thin-film filter combined with a multimode interference coupler;

Figure 4C shows an OPAD according to one embodiment of the present invention, wherein wavelength filtering is achieved by a reflective interface and a thin-film filter with a waveguide horn;

Fig. 5 shows an OPAD according to an embodiment of the present invention, wherein wavelength filtering is realized by the multimode interference filter employed in the passive waveguide layer;

Figure 6 shows an OPAD according to one embodiment of the present invention, wherein wavelength filtering is achieved by a grating-assisted lateral directional coupler formed in the passive waveguide layer; and

7 shows an OPAD according to an embodiment of the present invention, wherein the wavelength filtering between the optical amplifier and the photodetector is realized by a grating-assisted coupling structure, and the second grating-assisted coupling structure combines the corresponding The filtered broadband noise signal is coupled to a monitoring photodetector.

Detailed ways

The present invention is directed to an integrated optical preamplified detector (OPAD) with a bandpass wavelength filter between the amplification and detection sections of the device that is expected to reduce the amplified spontaneous emissions generated in the amplification section of the device to the detection section of the device The effect of broadband noise generated in the optical receiver, thereby enhancing the signal-to-noise ratio and improving the performance of optical receivers featuring optical preamplified detectors.

Reference may be made below to specific elements numbered according to the figures. The following discussion should be considered exemplary in nature, and not limiting on the scope of the invention. The scope of the present invention is defined in the claims and should not be construed as being limited by the implementation details described below, which can be modified by replacing elements with equivalent functional elements, as those skilled in the art will understand.

Optical waveguides are typically referenced by reference to etched ridge waveguide structures, and are identified by the ridge element in the uppermost layer of each etched ridge waveguide structure. Such references are intended to simplify the description rather than denote any components. The optical waveguide includes only the identified upper etched ridge elements. Those skilled in the art understand that the scope of the present invention is therefore not intended to be limited to such etched ridge waveguides, as they represent only a part of possible embodiments.

Referring to FIG. 1A , there is shown a prior art integrated OPAD 100A according to V. Tolstikhin et al., "Optically Pre-Amplified Detectors for Multi-Guide Vertical Integration in InP" (Proc. IPRM 2009). OPAD 100A includes passive and active waveguides shown by structures 110 and 120 respectively, which are stacked vertically with respect to each other such that passive waveguide 110 is below and designed to be transparent in the specified wavelength range of upper active waveguide 120 of. Laterally, the waveguides are defined by etching the shallow ridgelines of the passive waveguides 110 and the deep ridgelines of the active waveguides 120 . The terms "shallow" and "deep" ridgelines are hereafter used to identify ridgeline waveguide designs in which the etch terminates above and through the guiding layer, respectively. Etching back the active waveguide 120 actually forms a mesa where the N-contact 130 to the waveguide PIN structure is deposited on the top surface of the passive waveguide 110 in addition to the mesa. The P-contact is formed on the top surface of the mesa as the upper surface of the active waveguide 120 . Thus, the electrical isolation between the amplifying section 140 and the detecting section 150 of the respectively forward and reverse biased active waveguide PINs that share a common ground, ie N contact, is achieved by etching away the mesa between these two sections. material to achieve. A similar design as previously described is reported by S. Forrest et al. in US Patent 6,795,622 entitled "Photonic Integrated Circuits".

Although a simple insulating trench between the two PIN structures provides this electrical isolation, it also interrupts the active waveguide abruptly, creating unwanted light loss for the transition from the amplification to the detection segment. Not only is this loss undesirable as it must be compensated with greater gain which will generate greater ASE-related noise, but as it relates to PIC environments, excessive light scattering at the output of the amplified section will also cause damage to other optical circuits Optical crosstalk of components. Therefore, a low insertion loss transition between the amplification and detection sections of the active waveguide via an adiabatic transition from the amplification section to the passive waveguide section and then from the passive waveguide section to the detection section is a more preferred solution. This still provides electrical isolation between the two oppositely biased active waveguide segments. In practice, this adiabatic transition is achieved by appropriate tapers of active waveguides and possibly passive waveguides in the transition region, as shown in Fig. 1, but not clearly identified.

The lateral taper assisted adiabatic vertical transition of the optical signal between the active waveguide 120 and the passive waveguide 110 or vice versa is a design solution that can reduce the insertion loss between the electrically isolated amplification section 140 and the detection section 150 reduced to between 1dB and 2dB (see V. Tolstikhin et al., "Optically Pre-Amplified Detectors for Multi-Guide Vertical Integration in InP" (Proc. IPRM 2009, pp. 155-158, Newport Beach, 2009). However, these The vertical transition is not actually wavelength selective, and therefore passes the broadband ASE generated in the amplification section 140 to the detection section 150, where it will generate signal-ASE and ASE-ASE beat noise, producing them with equal efficiency.

Thus, an effective circuit configuration is shown in FIG. 1B by circuit 100B including optical gain block 180 and photodetector 190 . An optical signal at wavelength _λs is fed into circuit 100B from optical input 170 and coupled to optical gain block 180 . From optical gain block 180, the amplified optical signal at wavelength _λs propagates forward via transition block 185 to photodetector 190, which represents the effect of the two optical transitions between active and passive waveguides. Also coupled from optical gain block 180 are forward and reverse propagating ASE signals to optical input 170 and photodetector 190 respectively, these ASE signals having a wavelength spectrum λ _ASE . The forward propagating ASE signal propagates essentially unaffected with respect to the wavelength spectrum through the two optical transitions between the active and passive waveguides, but with reduced optical power (typically 0.5dB1. 0 dB), each reported by V. Tolstikhin et al. in "Optically Pre-Amplified Detectors for Multi-Guide Vertical Integration in InP" (Proc. IPRM 2009, pp. 155-158, Newport Beach, 2009).

即50nm或以下的光学通带Δλ_PBF的这些计算中采用的市场销售TIA的值时，达到对应于误码率10^-12的超过7的Q的值是可行的。对于在1490nm的中心波长适合于2.5Gb/s传输的1.8GHz的电带宽来执行图2中的计算。如图所示，第一至第三曲线210至230分别表示20nm、30nm和50nm的带宽的增益段噪声因数Fg=7的性能曲线。第四至第六240至260分别表示20nm、30nm和50nm的带宽的增益段噪声因子F_g=7的性能曲线。The effect of ASE on the receiver SNR is shown by the calculations shown in Fig. 2, where the Q factor is given as a function of the net gain of a signal with varying ASE filtered spectral width Δλ _PBF , and based on the analysis shown previously to provide. With reference to the application of OPADs in fiber-optic access to the home, light at the receiver (emitter relative intensity noise, shot and thermal noise in waveguide detectors) and All significant noise contributions in circuits that are electrically (represented by the equivalent input noise i ² _EIN in the front-end amplifier). It can be seen from Figure 2 that although the Q factor tends to saturate as a function of net gain, when the gain stage noise factor F _g is high (i.e., F _g =7), the saturation is at lower Q and net optical gain occur. Reducing the gain stage noise figure (ie, F _g =5) and/or limiting the optical passband of the ASE (ie, filtering out the ASE outside the passband Δλ _PBF ) increases the achievable Q-factor. When the input noise i _EIN of the front-end amplifier is set to the value

Values of Q exceeding 7 corresponding to a bit error rate of 10 ⁻¹² are feasible at the values of commercially available TIAs employed in these calculations of the optical passband Δλ _PBF of 50 nm or less. The calculations in Figure 2 were performed for an electrical bandwidth of 1.8 GHz suitable for 2.5 Gb/s transmission at a center wavelength of 1490 nm. As shown in the figure, the first to third curves 210 to 230 represent the performance curves of the noise factor Fg=7 of the gain section with bandwidths of 20nm, 30nm and 50nm, respectively. The fourth to sixth 240 to 260 represent the performance curves of the noise factor F _g =7 of the gain section with bandwidths of 20nm, 30nm and 50nm respectively.

For an optical passband filter to have a positive effect on receiver sensitivity by reducing the ASE-ASE beat noise, the filter passband Δλ _PBF should be narrower than the ASE spectral width λ _ASE under operating conditions, but wider than or equal to the pre- The wavelength range width λ _s of the amplified optical signal. In a typical waveguide semiconductor optical amplifier characterized by a large number or quantum well active layer and providing a net gain of -5dB 7dB, the ASE spectral width is wider than 50nm and may exceed 100nm, while the signal wavelength range width is usually narrower, such as in EPON or 20nm in the case of a GPON ONU data receiver or 10nm in the case of a GPON ONU video receiver, thus leaving the designer with the inequalities λ _S ≤ λ _PBF ≤ λ _ASE A space squeezed into. In such a passband, both the signal and the ASE will pass from the amplification into the OPAD's detection section, while all wavelengths outside the passband will be rejected and thus will not contribute to receiver noise.

Thus, it is evident from Fig. 2 that if, while remaining within the framework of a solution to achieve electrical isolation between the amplification and detection sections of the active waveguide with minimal optical losses, also provide passband wavelength filtering to reduce the If the resulting ASE affects the total received noise, then it is an improvement in the integrated OPAD design. It would be further advantageous to make this or such solution compatible with the MGVI platform, and accordingly make the integrated OPAD not only a high-function device, but also an important building block for a cost-effective PIC receiver and transceiver.

Referring to Figure 3, there is a schematic diagram representing an OPAD 300 according to an embodiment of the present invention in which passband filtering is employed between the optical amplification and detection sections of the OPAD 300 and between the leading optical circuit or network and the optical amplification section of the OPAD 300 element. The former is mandatory for any embodiment of the invention, while the latter is optional, and thus serves to prevent unwanted ASE light from entering the leading optical circuit or network.

同时，理想地，它应当不包括信号波长范围λ_s外部的所有波长，表明在适当设计的装置中在没有提供前PBF 320并且相应地在放大段的前侧没有吸收器350的情况下，入局光信号从源310直接传送到增益元件330中。Optical signals in the wavelength range _λs enter OPAD 300 at source 310 and are coupled to front passband filter (PBF) 320 with optical filters employed on both the front and rear sides of the amplification section of OPAD 300 . The front PBF 320 has a wavelength passband width Δλ ^F _PBF such that it includes all signal wavelengths, and thus passes the incoming optical signal into the amplification section (gain element) 330 . Wavelengths outside the passband range Δλ ^F _PBF are rejected and rerouted to absorber 350 , where the optical signals they carry are absorbed and thus prevented from propagating further into OPAD 300 . Since the front PBF 320 must transmit all incoming optical signals to the detector element 360, the Δλ ^F _PBF should include all signal wavelengths, i.e.,

At the same time, ideally, it should exclude all wavelengths outside the signal wavelength range λ _s , indicating that in a properly designed setup In the absence of a front PBF 320 and correspondingly no absorber 350 at the front side of the amplification section, the incoming optical signal is passed directly from the source 310 into the gain element 330 .

In either case, incoming optical signals are amplified in gain element 330 and then forward coupled to rear PBF 350 where they are filtered according to their passband Δλ ^B _PBF such that wavelengths in this passband propagate further to the OPAD 300, the detector element 360, while wavelengths outside this passband range are rejected and optionally routed to a monitoring element 370, which is, for example, another photodetector, thereby providing a net Feedback signal for control of gain (gain element 330). Since the rear PBF 360 must transmit all incoming optical signals to the detector element 370, the Δλ ^B _PBF should include all signal wavelengths, i.e. and preferably excludes all wavelengths outside the signal wavelength range _λs , i.e., in an optimal design

In addition to providing the required amplification of the incoming optical signal prior to its detection in detector element 360, gain element 330 also produces an unintended ASE by adding an ASE in parallel with gain element 330 in the block diagram shown in FIG. Element 380 is represented. This ASE is characterized by a wavelength range λ _ASE , which substantially overlaps the net gain range λ _G of gain element 330 , and propagates forward and backward from ASE element 340 in OPAD 300 .

The backpropagating ASE λ _ASE in the signal wavelength range λ _s is delivered into the leading optical circuit or network (schematically represented by source 310 ) both with the pre-PBF 320 and when no pre-BPF 320 is present. However, with the pre-PBF 320, then signals outside the signal wavelength range _λs can be rejected by the pre-PBF 320 and then absorbed in the absorber element 340, thereby reducing ASE penetration into the preamble circuit or network .

The forward propagating ASE in the signal wavelength range λ _s is transmitted by the rear PBF 360 into the detector element 360 together with the preamplified incoming signal. However, ASE signals outside the signal wavelength range λ _s are rejected in all embodiments of the invention. Optionally, these rejected signals are rerouted to and detected by monitoring element 370 to provide control of gain element 330, or otherwise absorbed, dissipated or routed.

如果Δλ_ASE＞＞Δλ_S，例如在具有宽增益谱和窄信号波长范围的装置中，它可以是显著的，由此改进OPAD300性能，但是如果Δλ_ASE≤λ_S，则ASE滤波没有实际上改进OPAD性能，并且因而没有意义。According to the above formula (3), the impact of post-PBF 350 on receiver noise is estimated as ASE-ASE beat frequency contribution to wideband noise reduction factor

If Δλ _ASE >> Δλ _S , for example in devices with a wide gain spectrum and narrow signal wavelength range, it can be significant, thereby improving OPAD300 performance, but if Δλ _ASE ≤ λ _S , ASE filtering is not actually improved OPAD performance, and thus meaningless.

Accordingly, the block diagram of an OPAD 300 with ASE filtering given in Figure 3 represents the most general solution and approach to this problem, which is not limited to any particular OPAD design, nor does it depend on PBF elements and rerouting waveguide elements the design of. Some specific designs of the rear PBF 360 elements described above for the OPAD 300 of FIG. 3 are shown below for the embodiments shown in FIGS. 4 to 7 . These are achievable using a range of optical waveguide circuit elements and arrangements. These waveguide circuits and arrangements are examples for convenience of illustration only, and do not represent all potential embodiments falling within the scope of the claims.

Referring to Figures 4A-4C, an embodiment of the present invention is shown in which a thin film filter (TFF) provides the required passband filtering in the OPAD. With the functionality previously described for Figure 3, the TFF is designed as a reflective filter for the signal wavelength range _λs and as a transmit filter for ASE wavelengths outside this range, _λASE ≤ λ _{PBF Lower} and λ _ASE ≥ λ _{PBF_Upper} , where λ _{PBF_Lower} and _{λPBF_Upper} represent the lower and upper wavelength limits of the PBF provided by the TFF, which can be set to the signal wavelength range _λs or determined to allow for environmental influences such as temperature. TFF employs, for example, a multilayer dielectric stack design (see, for example, JDS Uniphase Interference Filter Hanadbook, Rev. 2, 2007).

The basic idea of an embodiment that includes a TFF as a filtering element between the amplification and detection stages (ie, post-PBF 360 ) is illustrated in FIG. 4A by the schematic diagram of OPAD 400A. Accordingly, OPAD 400A includes an optical substrate 410 on which MGVI waveguide structures have been grown and patterned, not explicitly shown for clarity. Accordingly, the first passive waveguide 411 receives an amplified optical signal from the amplifying section 412 comprising the passive and waveguide layers of the MGVI receiving the incoming signal from the second passive waveguide 410 . The enlarged section 412 is shown schematically and is not intended to reflect the actual active-passive waveguide integration in the MGVI platform as would be apparent to those skilled in the art. Thus, incoming optical signals and forward propagating ASE in wavelengths _λs and λ _ASE respectively propagate towards the rear facet 416 of the device, where they are incident on the TFF 413 . Optical signals in the predetermined wavelength range _λs are reflected from the TFF 413 and then coupled to a third passive waveguide 414, which is optically connected to the detection section 415 of the OPAD 400A. Like the amplification section 412, the detection section 415 is shown schematically and is not intended to reflect the actual active-passive waveguide integration in the MGVI platform. All wavelengths outside the predetermined wavelength range λ _s are transmitted through the TFF 413 and accordingly out of the PIC comprising the OPAD 400A.

It will be apparent to those skilled in the art that optionally a photodetector (not shown in this schematic for clarity) can be placed behind the TFF 413 to measure transmitted light outside the wavelength range of the signal as forward ASE light , thereby providing the gain control of the amplification section of the OPAD 400A according to the block diagram of the OPAD 300 in FIG. 3 above.

The design of the TFF 413 at the rear facet 416 of the OPAD 400A should be tuned to the angle of incidence of the first passive waveguide 411 such that optical signals in the target wavelength range _λs are coupled into the second passive waveguide 414 after being reflected by the TFF 413 . Unlike e.g. DHCushing in US Patent 6011652 titled "Multi-Layer ThinFilm Dielectric Bandpass Filter" and PJ Gasoli in US Patent 5179468 titled "Interleaving of Similar Thin-Film Stacks for Producing Optical Interference Coatings" is expected to be used for approximation With a conventional TFF design towards (i.e., 0 degrees) incidence, the TFF 413 would be designed to operate at larger angles of incidence, but still remain smaller than the angle corresponding to the total internal reflection angle in wavelengths outside the predetermined wavelength range _λs .

In contrast to the design simplicity of the embodiment depicted in FIG. 4A , the actual implementation is quite tricky, for example, it requires in part to accurately split the coated TFF 413 end faces. If accurate splitting on the micrometer scale is not an option in tolerances comparable to waveguide widths, such as in the case of typical shallow etched ridgeline waveguides in InP-based materials operating in the 1.3 μm or 1.5 μm wavelength range, then Certain design modifications can be implemented to alleviate splitting tolerances.

One such modification is shown in FIG. 4B, which shows another embodiment in accordance with the present invention of an OPAD 400B including a TFF 425 at the end face of the device, wherein the split tolerance is mitigated by separately incorporating a two-port multimode interferometer (MMI ) 421 is inserted between the first and second passive waveguides 422, 423 and the device end face 424 with TFF 425 to achieve. A first passive waveguide 422 couples the signal from the amplification section 426 to the MMI 421 and a second passive waveguide 423 couples the filtered signal from the MMI 421 to the detection section 427 . For wavelengths in the predetermined signal wavelength range λ _s reflected by the TFF 425 into the PIC chip, although wavelengths outside this range pass through the TFF 425 and are transmitted from the PIC chip, the input port 421A and the output port 421B including the same end face 421 The performance of the MMI 421 is equivalent to that of a double-length MMI (such a transmissive MMI is not shown in FIG. 4B ) having the MMI 421 with oppositely facing input and output ports. The design technique of utilizing the MMI 421 thereby achieving the increased tolerance of the two-port MMI to the length of the MMI 421 is well known and practically reduced to providing a flat top passband in the port-to-port transmission spectrum, see e.g. L. Soldano et al. Optical Multi-Mode Interference Devices Based on Self-Imaging: Principles and Applications" (J. Lightwave Tech., vol. 13, No. 4, pp. 615-627, April 1995).

Thus, deviations in TFF 425 and rear facet positions (i.e. device facet 424 relative to MMI 421's front facet) that correspond to wavelength deviations are less significant in their impact on port-to-port transmission, thereby mitigating the impact of split tolerance on device performance Impact. As an added benefit, the MMI-assisted rear TFF solution allows the use of conventional TFF designs intended for normal incidence, such as D.H. Cushing in US Patent 6011652 entitled "Multi-Layer Thin Film Dielectric Bandpass Filter" and P.J. Gasoli in Described in US Patent 5,179,468 entitled "Interleaving of Similar Thin-FilmStacks for Producing Optical Interference Coatings". Those skilled in the art will appreciate that MMI 421 can also be designed to provide at least some aspect of wavelength filtering to function in conjunction with TFF 424.

It will also be apparent to those skilled in the art that the TFF 425 can be provided as a discrete TFF element bonded to the device end face 424 or it can be deposited onto the device end face 424 . Optionally, a third output optical port can also be added to the MMI 421, and accordingly, a third passive waveguide provided leading to the second detection section acts as a monitor in use, such as the monitoring of FIG. 3 described above. Components 380 to provide a gain control loop to the amplifier section of OPAD 400B are not shown in FIG. 4B for clarity. In this case, the resulting PIC's equivalent optical circuit reproduces the backend of the generic OPAD 300 given in Figure 3 previously described, where the TFF 425 and optionally the MMI 421 are used as both rear PBF 360 and active waveguide. A detection section serves as a detector element 370 and a monitor element 380 . In order for this functionality to be feasible, the treeport MMI must then deliver an optical signal in the wavelength range _λs to the output passive waveguide connected to the first detection section of the active waveguide, i.e. detection section 427 serving as detection element 370 , and ASE light in wavelengths outside the wavelength range λ _s is coupled into another passive waveguide section leading to a second detection section of the active waveguide (operating as monitoring element 380).

Another design solution is shown in FIG. 4C by OPAD 400C, which allows mitigating the impact of split tolerance on the OPAD performance characterized by the rear facet TFF of the PBF between the amplification and detection sections of the device. Here, first and second passive waveguides 431, 432, respectively coupling the optical signal from the amplification section and to the detection section and not shown for clarity, are equipped with planar focusing elements 433, 434, such as WKBurns et al. Described in "Optical Waveguide Parabolic Coupling Horns" (Appli. Phys. Lett. Vol. 30, pp. 28-30, Jan. 1, 1977). These planar focusing elements 433 and 434 are intended to provide parallel light beams into the slab waveguide 435 at the exit of the ridge waveguide, thereby reducing beam divergence in the plane of the waveguide core. The slab waveguide is terminated to the device end face on which TFF 440 is bonded or deposited. Thus, wavelengths in the predetermined signal range _λs will be reflected into the chip by the TFF 440 at the end facet, while the reflected beam in the slab waveguide will remain nearly parallel, regardless of the exact location of the split end facet relative to the passive waveguide. The planar focusing elements 433 and 434 need not be identical, instead it is entirely conceivable that they have different shapes, e.g. The illustrated launch plane focusing element 433 can be designed to provide a parallel beam at small angles of incidence, while at the beginning of the second passive waveguide 432, leading to the detection section of the OPAD 400C and similarly not shown in Figure 4C, the collection plane Focusing element 434 can be optimized for coupling a broad and divergent two-dimensional light beam, ie along a direction perpendicular to the plane of FIG. 4C , the light rays are confined in and around the passive waveguide core.

It will be apparent to those skilled in the art that in each of the embodiments described above with respect to FIGS. 4A to 4C , the ASE rays propagating through the TFFs respectively as TFFs 413, 425, 440 may optionally pass through detectors disposed behind the TFFs. to monitor, thereby allowing control of the gain in the amplification section of the OPAD. This additional TFF can be placed externally to the MGVI structure, or alternatively, where the TFF is placed in a groove formed in the MGVI structure, it can be placed externally with a waveguide interconnect (such as a planar waveguide structure), or placed in the MGVI Among the additional features implemented in the structure.

It will also be apparent to those skilled in the art that the previously shown embodiments employ reflective TFF, where the optical signals λ _and λ in the passband of the filter are reflected and coupled to the _{photodetector} , and alternatively may employ transmissive TFF filter such that the optical signals _λs and _λASE in the passband of the filter are transmitted and those outside the passband are reflected. Such a transmissive TFF element can be realized in an embodiment according to the invention by a suitable arrangement of the detector element 370, with or without the monitoring element 380 associated with a transmissive TFF with or without a planar waveguide element in between. .

Finally, it is also apparent that other waveguide elements and structures can be used in conjunction with TFFs to achieve wavelength filtering of ASE outside the predetermined signal wavelength range λ _s , see e.g. T. Augustsson entitled "Device and Method for Optical Add/Drop Multiplexing" US Patent 7,423,658 and US Patent 5,596,661 entitled "Monolithic Optical Waveguide Filters Based on Fourier Expansion" by CH Henry et al.

Referring now to FIG. 5 , there is shown a schematic diagram of an OPAD 500 according to one embodiment of the present invention, as previously shown schematically in FIG. 3 , which includes an amplification section 510 (gain element 330 in FIG. 3 ) and a detection section 520 ( detector element 360). ASE filtering is achieved by inserting an MMI 530 with associated first and second passive waveguides 530 and 540, acting as a post-PBF 360, between active waveguide sections 515 and 525, respectively, providing an amplification section 510 and a detection section 520. Similar to other embodiments of the present invention, this integrated assembly OPAD 500 includes a substrate 505 on which the MGVI structure is grown and processed, not explicitly identified for clarity. The MGVI structure and the guided optical signal propagation therein (e.g. lateral taper-assisted vertical conversion between passive and active waveguides) are related to V. Tolstikhin et al. in "Optically Pre-Amplified Detectors for Multi-Guide Vertical Integration in InP" (Proc .Indium Phosphide and Related Materials 2009 Conference, pp. 155-158, Newport Beach, 2009).

It is obvious that the OPAD 500 differs from this prior art in that the MMI 530 has now been incorporated into the section of the passive waveguide between the amplification and detection sections 510 and 520 of the active waveguide, respectively, wherein the MMI 530 is defined on the same vertical layer as the passive waveguide, as shown in FIG. 5 . The two-port MMI 530 is designed to operate as an optical passband filter according to the general description of the invention with reference to FIG. 3 . It receives an amplified optical signal in a predetermined wavelength range λ _s together with ASE light in a wavelength range λ _ASE which is generally wider than the signal range λ _s . However, it only transmits wavelengths in the range λ _PBF that coincides with the signal wavelength range λ _s , so that these wavelengths propagate first through the second passive waveguide 550 between the MMI 530 and the detection section 520, and secondly by means of passive and The vertical taper defined at the active waveguide stage propagates vertically into the active waveguide 525 into the detection section 520 of the OPAD 500 and its active waveguide section 525 .

The design principle of the MMI 530 with estimated wavelength filtering is well known, for example in "Optical Multi-Mode Interference Devices Based on Self-Imaging: Principles and Applications" by L.Soldano et al. (J.Lightwave Tech., Vol.13, No. 4, pp. 615-627, April 1995) and R.M. Jenkins et al. in US Patent 5,428,698 "Optical Routing Device". It should be apparent to those skilled in the art that by proper selection of the MMI shape and size, and by adjusting the layout of the passive waveguides entering and leaving the MMI filter, the passives in the MGVI optimized for efficient and controllable passive-active vertical coupling The source waveguide layer is also suitable for the required MMI passband filtering.

Optionally, a second output optical port can be added to the MMI 560 and accordingly a second passive waveguide leading to the second detection section is provided in use as the monitoring element 308 to provide gain to the amplification section of the OPAD The control loop, not shown in Figure 5 for clarity. In this case, the equivalent optical circuit of the integrated assembly reproduces the backend of the generic OPAD shown in Fig. 3, where the MMI 530 serves as the rear PBF 350 and the two detection segments of the active waveguide serve as the detector element 360 and monitor 370 . In order to make this functionality feasible, the treeport MMI must then deliver optical signals in the wavelength range _λs to the output passive waveguide connected to the first detection section of the active waveguide acting as detector element 360, and the wavelength range λ ASE light in wavelengths outside _s is directed to a passive waveguide operating as a monitor 370 leading to a second detection section of the active waveguide.

Referring now to FIG. 6 , there is shown a schematic diagram of an OPAD 600 according to an embodiment of the present invention, as shown schematically in FIG. 3 above, which includes an amplification section 610 (gain element 330 in FIG. 3 ) and a detection section 620 ( detector element 360). ASE filtering is achieved by inserting a grating-assisted directional coupler 650 with associated first and second coupler waveguides 630, 640 acting as post-PBF 360 between the amplification section 610 and the detection section 620, respectively. The first and second coupler waveguides 630 , 640 each form a first and second grating 635 , 645 on their upper surface, respectively, such that the entire combination acts as a grating-assisted directional coupler 650 . As with other embodiments of the invention, this integrated assembly OPAD 600 includes a substrate 605 on which the MGVI structure is grown and processed, not explicitly identified for clarity. The MGVI structure and the guided optical signal propagation therein, such as the lateral taper-assisted vertical conversion between passive and active waveguides, are discussed with V. Tolstikhin et al. in "Optically Pre-Amplified Detectors for Multi-GuideVertical Integration in InP" (Proc .Indium Phosphide and Related Materials 2009 Conference, pp. 155-158, Newport Beach, 2009).

It will be apparent that the OPAD 600 differs from this prior art in that a grating assisted directional coupler 650 has now been incorporated into the passive waveguide section between the amplification and detection sections 610, 620 of the active waveguide respectively. , where the grating-assisted directional coupler 650 is defined on the same vertical layer as the passive waveguide, as shown in FIG. 6 . The grating assisted directional coupler 650 is designed to operate as an optical passband filter according to the general description of the invention with reference to FIG. 3 . It receives an amplified optical signal in a predetermined wavelength range λ _s together with ASE light in a wavelength range λ _ASE which is generally wider than the signal range λ _s . It transmits wavelengths in the range λ _PBF coincident with the signal wavelength range λ _s such that these wavelengths pass vertically into the active waveguide 625 first at the Propagated in one coupler waveguide 630 and secondly coupled into a second coupler waveguide 640 into the detection section 620 of the OPAD 600 and its active waveguide section 625 . Signals outside the λ _PBF at the output of the grating assisted directional coupler 650 are in the first coupler waveguide 630 in which they are disposed.

The design principle of the grating-assisted directional coupler 650 with predicted wavelength filtering is well known, for example, A.Carenco et al. Grating using Sample Grating" (IEEE Hot. Tech., Lett., Vol. 6, pp. 1222-1224, 1994). It should be apparent to those skilled in the art that by proper selection of the grating structure, directional coupler waveguide, coupler transfer characteristics, and proper design and adjustment of any non-conductive components disposed between the grating-assisted directional coupler 650 and the amplification and detection sections 610 and 620 The layout of the source waveguide section, the passive and active waveguide layers in the MGVI optimized for efficient and controllable passive-active vertical coupling is also suitable for the required grating-assisted directional coupler filtering, this passive waveguide section is in Not shown in FIG. 6 .

Optionally, a second output passive optical waveguide may be added to the output of the first coupler waveguide 630 and correspondingly when properly set up leads to the second detection section in use as the monitoring element 308 for amplification to the OPAD stage provides a gain control loop, not shown in Figure 6 for clarity. In this case, the equivalent optical circuit of the integrated assembly reproduces the backend of the generic OPAD shown in Figure 3, where a grating-assisted directional coupler 650 is used as the rear PBF 350 and the two detection sections of the active waveguide are used as the detection device element 360 and monitor 370. To make this functionality feasible, the three-port directional coupler must then deliver an optical signal in the wavelength range _λs to the output passive waveguide connected to the first detection section of the active waveguide acting as detector element 360, and the wavelength ASE light in wavelengths outside the range _λs is sent to a passive waveguide operating as a monitor 370 leading to a second detection section of the active waveguide

Referring now to FIG. 7, there is shown a schematic diagram of an OPAD 700 according to an embodiment of the present invention, as shown schematically in FIG. 3 above, which includes an amplification section 730 (gain element 330 in FIG. detector element 360) and monitoring section 745. ASE filtering for the detection section 750 is achieved by inserting a first grating assisted coupler 740 acting as a rear PBF 360, while ASE filtering back into the optical network to which the OPAD 700 is connected is assisted by inserting a second grating assisted coupler 740 acting as a front PBF 320 Coupler 725 to achieve. As with the other embodiments of the invention previously described with respect to FIGS. 4A-6 , this fully integrated implementation of OPAD 700 includes a substrate 705 on which MGVI structures have been grown and processed, not explicitly identified for clarity. The MGVI structure and the guided optical signal propagation in it, such as the lateral taper-assisted vertical conversion between passive and active waveguides, and V. Tolstikhin et al. in "Optically Pre-Amplified Detectors for Multi-Guide Vertical Integration in InP" (Proc. Similar to that reported in Indium Phosphide and Related Materials 2009 Conference, pp. 155-158, Newport Beach, 2009).

It will be apparent that the OPAD 700 differs from this prior art in that a second grating assisted coupler 725 now combines the input 710 and amplification section 730 in a segment of the passive waveguide, and the first grating assisted coupler 740 Inserted between the amplification and detection sections 730 and 750 of the active waveguide, respectively, wherein the first and second grating-assisted couplers 740 and 725 are respectively defined on the same vertical layer as the passive waveguide, as shown in FIG. 7 . Each of the first and second grating-assisted couplers 740 and 725, respectively, is designed to operate as an optical passband filter according to the general description of the invention with reference to FIG. 3 . Consider a first grating assisted coupler 740 that receives an amplified optical signal in a predetermined wavelength range λ _s together with ASE light in a wavelength range λ _ASE , which is generally wider than the signal range λ _s from amplification section 730 . However, it transmits to the first output port only those wavelengths in the range λ _PBF coincident with the signal wavelength range λ _s , so that these wavelengths enter the detection section 750 of the OPAD 700 and its active waveguide section, not explicitly identified for clarity. These optical signals in the range λ _PBF are first passed between the first grating assisted coupler 740 and the detection section 750 before passing vertically into the active waveguide of the detection section 750 by means of the vertical taper defined at the passive and active waveguide stages. propagating in the passive waveguide between them.

Similarly, optical signals outside the range λ _PBF are routed to the second output port of the first grating assisted coupler 740 so that these signals enter the monitoring section 745 of the OPAD 700 and its active waveguide section, not explicitly identified for clarity. These optical signals outside the range λ _PBF are first passed between the first grating assisted coupler 740 and the monitoring section 745 before passing vertically into the active waveguide of the monitoring section 745 by means of the vertical taper defined at the passive and active waveguide stages. propagating in the passive waveguide between them.

Considering now the optical signals entering OPAD 700 , they are coupled at input 710 to input passive waveguide 715 and then coupled into second grating assisted coupler 725 that has been incorporated between input passive waveguide 715 and amplification section 730 . Thus, the second grating assisted coupler 725 receives optical signals of the predetermined wavelength range λ _s together with any out-of-band signals from the leading optical network. However, it only transmits to the first output port those wavelengths in the range λ _PBF coincident with the signal wavelength range λ _s , so that these wavelengths enter the amplification section 730 of the OPAD 700 and its active waveguide section, not explicitly identified for clarity. Any signals received from the leading optical network are coupled to the other output of the second grating auxiliary coupler 725 and not coupled to the amplification section 730 .

As previously stated, the amplifying section 730 emits an ASE bi-directionally, and accordingly, if the optical input 710 is connected directly to the amplifying section 730, this ASE is recoupled directly into the leading optical network where it is transmitted into the main optical telecommunications network May or may not be filtered and attenuated before. However, the OPAD 700 includes a second grating assisted coupler 725 as previously described. Accordingly, it transmits to the input passive waveguide 715 that portion of the ASE in the wavelength range λ _PBF that coincides with the signal wavelength range λ _s , so that these wavelengths enter the input passive waveguide 715 and are then coupled into the leading optical network . The ASE outside the lambda _PBF is coupled to the other output of the second grating auxiliary coupler 725 and to the reverse monitoring section 735 . Accordingly, OPAD 700 provides a monolithic implementation of the general description of the invention of FIG. 3 . The signal from the reverse monitoring section 735 can be combined with the signal from the monitoring section 745 to provide control of the amplification section 730 alone or of the OPAD 700 as a whole.

It will be apparent to those skilled in the art that the first and second grating assisted couplers 725 and 740 are shown to have the same passband, ie, λ _PBF , as previously described with respect to FIG. 7 . However, depending on the performance requirements of the overall OPAD 700 and its receiver path elements (ie, amplification section 730 and detection section 750 ), it may be advantageous to design these with different performance characteristics that may include bandpass width, isolation, and the like. It is also apparent that the design of OPAD 700 is shown with input passive waveguide 710, amplification section 730, and detection section 750 formed in the same continuous passive waveguide. Alternatively, the design can be tuned such that the desired optical signal _λs in the λ _PBF is in each directional coupler intersection, the so-called crossover state, rather than straight-through, the so-called through-state. Alternatively, one directional coupler can be designed to be in the through state while the other directional coupler is in the crossover state. In each case, detection and monitoring segments 750 and 745, respectively, are collocated as desired.

Furthermore, in the preceding Figures 6 and 7, each grating assisted coupler was shown as a co-propagating directional coupler such that an optical signal propagates from one end of the directional coupler to the other. Alternatively, with a grating-assisted directional coupler, the design can be implemented as a back-propagating directional coupler, such that the filtered optical signals are not only coupled to the other arm of the directional coupler, but are also reflected so that they originate from the same The end coupling of the directional coupler. In such designs, the outputs at the other end of the directional coupler each contain the undesired signal and each can be coupled to a discrete photodetector, either the repeating monitoring section 745 or the reverse monitoring section 735 or both. single large photodetector. If implemented in the first grating assisted coupler 740, the position of the detection section 750 is also adjusted.

It will be apparent that in the previously shown embodiments the wavelength filtering element, the grating assisted directional coupler and the MMI represent only two of the possible embodiments of a wavelength filtering element possible in a PIC. Optionally, wavelength filtering may include other structures including, but not limited to, Mach-Zehnder interferometers (MZIs), echelle gratings, directional couplers, and arrayed waveguide gratings (AGWs). Furthermore, it will be apparent to those skilled in the art that while the previously shown embodiments employ transmissive waveguide filtering elements, such as MMI 530 and grating-assisted directional coupler 670, alternate design options exist, including the ability to The reflective filter element used.

Alternatively, different structures can be implemented for the front PBF 320 and rear PBF 350 in a fully monolithic waveguide solution, or the monolithic waveguide solution of one of these PBFs can be used in combination with the TFF solution of the other. Alternatively, both PBFs can be implemented using a single TFF or dual TFFs. The specific implementation is determined, for example, by factors including but not limited to: wavelength filtering requirements of the standard or system with which the OPAD is expected to operate, performance limitations of other PIC functions in the PIC of which the OPAD forms a part, cost, footprint, performance, etc. .

Additionally, alternative embodiments of the OPAD are possible without departing from the scope of the invention, including, for example, providing multiple detector elements coupled from a single amplification section for wavelength division multiplexed PON, local area network, metropolitan area network, and long-haul applications , and from a cascade of two or more elements to provide wavelength filtering.

The above-described embodiments of the present invention are meant to be examples only. Alterations, modifications and alterations to the particular embodiment may be effected by those skilled in the art without departing from the scope of the invention, which is defined only by the appended claims.

Claims (19)

1. A photonic assembly, comprising: a) an epitaxial semiconductor structure, when the substrate includes at least one of a commonly assigned waveguide for supporting propagation of an optical signal in a predetermined first wavelength range and a plurality of wavelength assigned waveguides vertically arranged in order of increasing wavelength band gap, at grown in a III-V semiconductor material system in a single growth step, wherein each of said plurality of wavelength-specifying waveguides supports a predetermined second wavelength range, each of said predetermined second wavelength ranges being in said predetermined first wavelength range ; b) an optical input port for receiving optical signals in said first wavelength range; c) a first filter comprising at least a first output port and a second output port and characterized by at least a first passband width, said filter being optically coupled to receive light in said first wavelength range an optical input port for providing a first predetermined portion of the received optical signal to the first output port, the first predetermined portion of the received optical signal being determined based on at least the first passband width; d) an optical amplifier comprising at least a gain section formed in one of said plurality of wavelength-specifying waveguides, a first contact for forward biasing said optical amplifier, and a third output port, said optical amplifier at optically coupled to the first output port for receiving the first predetermined portion of the received optical signal and providing an amplified filtered optical signal to the third output port; e) a second filter comprising at least a fourth output port and a fifth output port and characterized by at least a second passband width, said filter being optically coupled to said third output port of said optical amplifier and for providing a first predetermined portion of the amplified filtered optical signal to the fourth output port and a second predetermined portion of the amplified filtered optical signal to the fifth output port, the amplified filtered optical signal The first and second predetermined portions of are determined based at least on said second passband width; f) a first photodetector optically comprising at least a second contact for reverse biasing said first photodetector, said first photodetector coupled to a said fourth output port of said second filter for filtering a first predetermined portion of an optical signal; g) a second photodetector optically coupled to a fifth output port of said second filter for receiving a second predetermined portion of said amplified filtered optical signal; and h) a third photodetector optically coupled to a second output port of said first filter for receiving a predetermined portion of the optical signal propagating from said optical amplifier to said first filter, said The predetermined portion of the optical signal is determined based on at least said first passband width; wherein, The first contact and the second contact are formed on the same layer of the epitaxial semiconductor structure, but are electrically isolated from each other. 2. The photonic assembly of claim 1, wherein: At least one of said first photodetector, said second photodetector, said third photodetector, and said optical amplifier includes a vertical element for combining multiple wavelengths from a commonly designated waveguide an optical signal in a predetermined second wavelength range of one of the designated waveguides is coupled to one of the plurality of wavelength designated waveguides, the vertical element comprising at least one lateral taper formed in the common designated waveguide by at least one semiconductor etching process formed in each of the plurality of wavelength-specifying waveguides, one of the plurality of wavelength-specifying waveguides, and any intermediate waveguides of the plurality of wavelength-specifying waveguides between the common-specifying waveguide and one of the plurality of wavelength-specifying waveguides. 3. The photonic assembly of claim 1, wherein, At least one of the first filter and the second filter includes at least one of a thin film filter and a waveguide filter implemented in the epitaxial semiconductor structure. 4. The photonic assembly of claim 3, wherein, When the at least one is a thin-film filter, it is at least one of the following: abutting on the end face of the epitaxial semiconductor structure, deposited on the end face of the epitaxial semiconductor structure, and disposed on the surface of the epitaxial semiconductor structure Among the features formed in . 5. The photonic assembly of claim 3, wherein, When said at least one is a waveguide filter, it includes at least a first element selected from the group consisting of multimode interference filters, directional couplers, Mach-Zehnder interferometers, arrayed waveguide gratings, echelle gratings, and Bragg gratings and ring resonators. 6. The photonic assembly of claim 3, wherein, When said at least one is a waveguide filter, a first predetermined portion of said waveguide filter is implemented in said common assigned waveguide and another one of said plurality of wavelength assigned waveguides, said one and said One of the plurality of wavelength-designating waveguides is adjacent. 7. A photonic assembly comprising: a) an epitaxial semiconductor structure, when the substrate includes at least one of a commonly assigned waveguide for supporting propagation of an optical signal in a predetermined first wavelength range and a plurality of wavelength assigned waveguides vertically arranged in order of increasing wavelength band gap, at Grown in a III-V semiconductor material system grown in a single growth step, wherein each of said plurality of wavelength-specifying waveguides supports a predetermined second wavelength range, each of said predetermined second wavelength range being at said predetermined first wavelength in range; b) an optical input port for receiving optical signals in said first wavelength range; c) an optical amplifier comprising at least a gain section formed in one of said plurality of wavelength-specifying waveguides, a first contact for forward biasing said optical amplifier, and a first output port, said optical amplifier at optically coupled to an optical input port for receiving the optical signal and providing an amplified optical signal to the first output port; d) a first filter comprising at least a second output port and characterized by at least a first passband width, said filter being optically coupled to the first output port of said optical amplifier and configured to provide input to said second an output port providing a first predetermined portion of an amplified optical signal determined based on at least said first passband width; e) a first photodetector optically comprising at least a second contact for reverse biasing said first photodetector, said first photodetector coupled to a a second output port of said first filter of a first predetermined portion of an optical signal; wherein The first contact and the second contact are formed on the same layer of the epitaxial semiconductor structure, but are electrically isolated from each other. 8. The photonic assembly of claim 7, wherein, At least one of the optical amplifier, the first filter, and the first photodetector further includes a vertical element for coupling at least one of the signals from and to the commonly assigned waveguide and to the and optical signals in said predetermined second wavelength range from one of said plurality of wavelength-specifying waveguides between at least one of said plurality of wavelength-specifying waveguides, said vertical element comprising at least one transverse taper, It passes at least one semiconductor etching process between each of said commonly assigned waveguides, one of said plurality of wavelength assigned waveguides, and said plurality of wavelengths between said commonly assigned waveguide and one of said plurality of wavelength assigned waveguides Formed in any intermediate waveguides of the specified waveguide. 9. The photonic assembly of claim 7, further comprising: f) a second photodetector optically connected to said first filter and adapted to receive a second predetermined portion of said amplified optical signal, said second predetermined portion of said amplified optical signal being at least Determined according to the first passband width. 10. The photonic assembly of claim 7, further comprising: f) a second filter, disposed between said input port and said optical amplifier, and characterized at least by a second passband width, said filter for providing said optical amplifier with a third predetermined part and one of the first predetermined part of the noise generated by the optical discharger is provided to the input port, the third predetermined part of at least one of the received optical signal and the noise generated by the optical discharger At least one of the first predetermined portions of noise is determined based at least on the second passband width. 11. The photonic assembly of claim 9, further comprising: g) a third photodetector optically connected to said second filter and adapted to receive a second predetermined portion of said noise generated by said optical amplifier, a portion of said noise generated by said optical amplifier The second predetermined portion is determined based on at least the second passband width. 12. The photonic assembly of claim 7, wherein, The first filter includes at least one of a thin film filter and a waveguide filter implemented in the epitaxial semiconductor structure. 13. The photonic assembly of claim 10, wherein, The second filter includes at least one of a thin film filter and a waveguide filter implemented in the epitaxial semiconductor structure. 14. The photonic assembly of claim 12, wherein, When the at least one is a thin-film filter, it is at least one of the following: abutting on the end face of the epitaxial semiconductor structure, deposited on the end face of the epitaxial semiconductor structure, and disposed on the surface of the epitaxial semiconductor structure Among the features in . 15. The photonic assembly of claim 13, wherein, When the at least one is a thin-film filter, it is at least one of the following: abutting on the end face of the epitaxial semiconductor structure, deposited on the end face of the epitaxial semiconductor structure, and disposed on the surface of the epitaxial semiconductor structure Among the features in . 16. The photonic assembly of claim 12, wherein, When said at least one is a waveguide filter, it includes at least one of the following: a first of said waveguide filters implemented in one of said commonly assigned waveguides and the other of said plurality of wavelength assigned waveguides. The predetermined portion, said one being adjacent to one of said plurality of wavelength-specifying waveguides, and comprising a multimode interference filter, a directional coupler, a Mach-Zehnder interferometer, an arrayed waveguide grating, an echelle grating, a Bragg grating, and a ring At least one first element selected from the set of resonators. 17. The photonic assembly of claim 13, wherein, When said at least one is a waveguide filter, it includes at least one of the following: a first one of said waveguide filter implemented in another one of said common assigned waveguide and said plurality of wavelength assigned waveguides. a predetermined portion, said one being adjacent to one of said plurality of wavelength-specifying waveguides, and comprising a multimode interference filter, a directional coupler, a Mach-Zehnder interferometer, an arrayed waveguide grating, an echelle grating, a Bragg grating, and At least one first element selected from the group of ring resonators. 18. The photonic assembly of claim 7, further comprising: f) at least one optical element of a plurality of optical elements disposed between the optical amplifier and the first filter and between at least one of the first filter and the first photodetector, the optical element comprising At least one of a waveguide filter element, a predetermined portion of a multimode interference waveguide element, a predetermined portion of a directional coupler, a predetermined portion of a Mach-Zehnder interferometer, a parabolic waveguide element, a planar waveguide, and a lens. 19. The photonic assembly of claim 11, further comprising: h) at least one optical element of a plurality of optical elements disposed between the optical amplifier and the second filter, between the input port and the second filter, and between the second filter and the second filter Between at least one of the third photodetectors, the optical element comprises a waveguide filter element, a predetermined portion of a multimode interference waveguide element, a predetermined portion of a directional coupler, a predetermined portion of a Mach-Zehnder interferometer, a parabolic waveguide At least one of an element, a planar waveguide, and a lens.

Images