METHOD AND DEVICE FOR IMPLEMENTING AVERAGE POOLING OF NEURAL NETWORK, AND STORAGE MEDIUM

DETAILED ACTION Notice of Pre-AIA or AIA Status The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA . 2. This action is responsive to the application filed on 06/23/2026. Claims 1-20 are pending in the case. Priority The present application cl08/18/2021 priority under 35 U.S.C. §119 to GB patent Application No. CN202110945903.7, filed on 08/18/2021. Applicant’s claim for the benefit of a prior-filed application under 35 U.S.C. 119 and/or 35 U.S.C. 120 is acknowledged. Information Disclosure Statement 4. As required by MPEP 609 (c), the Applicants’ submission of the Information Disclosure Statement(s) filed on 06/23/2026 are acknowledged by the examiner and the cited references have been considered in the examination of the claims now pending. Examiner Comments 4. In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status. Claim Rejections - 35 USC § 103 The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action: A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made. Claims 1-13 and 17-20 are rejected under 35 U.S.C. 103 as being unpatentable over Mehrabian (PUB. No. US 20200019851 A1, Date: 2020-01-16) in view of Young (Pub. No. US 20180165577 A1, Pub. Date 2018-06-14) Regarding independent Claim 1, Mehrabian teaches a method for performing average pooling of a neural network (see Mehrabian: Fig.9, [0057], “first convolutional layer 902 filters a 224×224×3 input image 918 with 96 kernels of size 11×11×3. The second convolutional layer 904 takes as input the (response-normalized and pooled)”), wherein the method comprises: acquiring a plurality of to-be-treated optical signals with unequal wavelengths (see Mehrabian: Fig.1, [0031], “collection of kernels 110 are processed simultaneously on a single input 120. Each input 120 is assigned to a distinct wavelength of light.”) inputting the to-be-treated optical signals into a micro-ring-resonator array (see Mehrabian:Fig.3, [0033], “illustrates a micro ring resonator (MRR) neuron 300.”… “MRR neuron 300 is a component of a photonic convolutional filter. In the MRR neuron 300 of the photonic convolutional filter, wavelength-division multiplexed input signals 310a (e.g., 120, 230) are weighted through tunable MRRs 340a. Weighted inputs 310a are later incoherently summed using balance detector summation circuit 350.”), wherein the micro-ring-resonator array comprises a plurality of micro-ring resonators that are connected in series (see Mehrabian: Fig.4, [0036], “the MRR 440 is connected to memory 410. In at least one example of this disclosure, memory 410 is a memristor 470. The MRR 440 is connected to memory 410 (470) via p-junction 450. The MRR 440 is also connected to through-waveguide 480. The MMR 440 is connected to waveguide output 430 through n-junction 460.”) applying a corresponding electric current to the micro-ring-resonator array (see Mehrabian: Fig.7, [0052, “generating 718 an analog electrical signal corresponding to the one or more integrated light signals. The method 700 further comprises: receiving 720 the analog electrical signal at an analog-to-digital converter (e.g., 570); converting 722 the analog electrical signal to a digital electrical signal using the analog electrical signal at an analog-to-digital converter (e.g., 570); and outputting 724 the digital electrical signal from the analog-to-digital converter (e.g., 570). In at least one example of the method 700, the one or more voltage signals correspond to one or more analog Winograd domain filter signals.”), to adjust a transfer function of each of the micro-ring resonators to reach a target value (see Mehrabian: Fig.3, [0058], “The electrical signal (e.g., 380) is transformed back by an inverse photonic Winograd Transform circuit (e.g., 520c) and converted to a digital signal using an analog-to-digital converter (e.g., 570). Outputs (e.g., 240) from the analog-to-digital converter (e.g., 570) are placed in output line buffers (e.g., 510b) and stored backed to off-chip memory (e.g., 580). Such output values (e.g., 240) from a first layer (e.g., 902) constitute the inputs (e.g., 120, 230) for the next layer (e.g., 904). The process described above is performed for all five convolutional layers 902, 904, 906, 908, 9”); and feeding an optical signal outputted by the micro-ring-resonator array into a photodiode (see Mehrabian: Fig.3, [0033], “The EWMM circuit 340 is connected to a balance detection summation circuit 350. In at least one example of this disclosure, the balance detection summation circuit 350 includes a first photodetector 350a and a second photodetector 350b. In at least one example of this disclosure, the first photodetector 350a is a photodiode and the second photodetector 350b is also a photodiode.”) As shown above, Mehrabian teaches all the structural and operational elements using of the claimed method except for explicitly labeling the result as average pooling. Mehrabian teaches the implementation of micro ring resonators to perform average pooling of neural network. Mehrabian does not each the system comprising: average pooling of a neural network (see Young: Fig.6, [0078], “a flow diagram of an example method for instructing a neural network processing system to perform a computation for an average pooling layer of a neural network.” obtain an operation result of the average pooling of the neural network (see Young: Fig.6, [0078], “generates instructions that, when executed by the special-purpose hardware circuit of the system, cause the special-purpose hardware circuit to, during processing of a network input by the neural network, generate a layer output tensor that is equivalent to an output of the average pooling neural network layer (606). For example, in response to the request, the neural network implementation engine 150 of the system of FIG. 1 generates instructions that direct or control the special-purpose hardware circuit 110 to generate an output tensor, i.e., an output vector, that is equivalent to an output that the special-purpose hardware circuit 110 would generate if it processed an input tensor by performing average pooling on the input tensor.”) Because both Mehrabian and Young are in the same/similar field of neural network computation, accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the invention, to modify the photonic neural network of Mehrabian to perform average pooling as taught by Young. After modification of Mehrabian, the photonic neural network that perform the summation of multiple wavelength encoded optical input using micro-ring resonators and also perform average pooling of neural network in order to achieve standard neural function. One would have been motivated to make such a combination in order to improve neural network function by pooling the outputs can also reduce dimensionality of inputs received at the subsequent neural network layers while maintaining desired characteristics of the outputs before pooling, which can improve efficiency without significantly compromising the quality of inferences generated by the neural networks. (see Young: [0025]) Regarding Claim 2, As shown above, Mehrabian and Young teach all the limitation of claim 1. Mehrabian further teach the system wherein: each of the micro-ring resonators comprises one straight waveguide and one micro-ring waveguide (see Mehrabian: Fig.3, [0033], “a micro ring resonator (MRR) neuron 300. The MRR neuron 300 includes an EOM 310 coupled to a wavelength-division multiplexer 330. The wavelength-division multiplexer 330 is connected to an element-wise matrix multiplication (EWMM) circuit 340. The wavelength-division multiplexer 330 is configured to receive (laser) signals 310a having wavelengths λ.sub.1 . . . λ.sub.n, (e.g., 120, 230). The EWMM circuit 340 contains a MRR 340a. The EWMM circuit 340 is connected to a balance detection summation circuit 350. In at least one example of this disclosure, the balance detection summation circuit 350 includes a first photodetector 350a and a second photodetector 350b.”); and micro-ring radii of the micro-ring resonators are unequal (see Mehrabian: Fig.4, [0037], “EOM 420 (310) has a silicon-based p-i-n junction that is side coupled to a wave guide (e.g., 480) or to a p-n junction (e.g., 450, 460). In one or more examples, the maximum allowed optical power flowing in each physical channel of a photonic accelerator (PA) (500) is bound by the optical power that would produce non-linearities in the silicon waveguides (e.g., 480) and the minimum power that a photo-detector (not shown) can distinguish from noise is when the signal-to-noise ratio of the PA (500) is unity (SNR=1).”) Regarding Claim 3, As shown above, Mehrabian and Young teach all the limitation of claim 1. Mehrabian further teach the system wherein: a quantity of types of the wavelengths of the to-be-treated optical signals is equal to a quantity of the micro-ring resonators (see Mehrabian: Fig.1, [0031], “the collection of kernels 110 are processed simultaneously on a single input 120. Each input 120 is assigned to a distinct wavelength of light.”); and the wavelengths of the to-be-treated optical signals correspond to radii of the micro-ring resonators one to one (see Mehrabian: Fig.2, [0032], “Operating on the inputs 230 by the kernels 210 produces output 240. In at least one CNN, using N wavelengths of light, it is possible to establish up to N.sup.2 independent connections, in accordance with this disclosure. The number of independent connections which are possible enable a CNN to operate on much larger datasets than non-photonic neural networks.”) Regarding Claim 4, As shown above, Mehrabian and Young teach all the limitation of claim 1. Mehrabian further teach the system wherein: before the operation of applying the corresponding electric current to the micro-ring-resonator array (see Mehrabian: Fig.3, [0033], “The wavelength-division multiplexer 330 is connected to an element-wise matrix multiplication (EWMM) circuit 340. The wavelength-division multiplexer 330 is configured to receive (laser) signals 310a having wavelengths λ.sub.1 . . . λ.sub.n, (e.g., 120, 230)”), the method further comprises: by the to-be-treated optical signals with the unequal wavelengths, performing resonance with the corresponding micro-ring resonators (see Mehrabian: Fig.3, [0035], “when a bias voltage V.sub.1 is applied to the EOM 310, the application of the bias voltage causes the transmission spectrum of the MRR 340a to have a resonant frequency f.sub.1. When wavelength-division multiplexed light passes through the coupled waveguide 480, the component of that light having (matching) wavelength λ.sub.1 is coupled into the MRR 340a. In at least one example, raising the bias voltage applied to EOM 310 causes the resonant frequency f.sub.1 of the MRR 340a to shift due to the change in the effective refractive index of the MRR 340a.”) Regarding Claim 5, As shown above, Mehrabian and Young teach all the limitation of claim 1. Mehrabian further teach the system wherein: straight waveguides of all of the micro-ring resonators of the micro-ring-resonator array are a same shared straight waveguide (see Mehrabian: Fig.3, [0035], “When wavelength-division multiplexed light passes through the coupled waveguide 480, the component of that light having (matching) wavelength λ.sub.1 is coupled into the MRR 340a. In at least one example, raising the bias voltage applied to EOM 310 causes the resonant frequency f.sub.1 of the MRR 340a to shift due to the change in the effective refractive index of the MRR 340a.”); the shared straight waveguide is provided with an input port and a direct-passing port (see Mehrabian: Fig.4, [0036], “EWMM circuit 340 includes drop-waveguide 430. The MRR 440 is connected to memory 410. In at least one example of this disclosure, memory 410 is a memristor 470. The MRR 440 is connected to memory 410 (470) via p-junction 450. The MRR 440 is also connected to through-waveguide 480. The MMR 440 is connected to waveguide output 430 through n-junction 460.”); and the photodiode is located at the direct-passing port (see Mehrabian: Fig.4, [0035], “the balance detector summation circuit 350 can include two photodiodes 350a, 350b. The balance detector summation circuit 350 integrates (determines integral sum of the values of) the various wavelengths of light that the balance detector summation circuit 350 receives through waveguide 480 and generates an electrical signal 380 corresponding to sum of the light beam power.”) Regarding Claim 6, As shown above, Mehrabian and Young teach all the limitation of claim 1. Mehrabian further teach the system wherein: light intensities of the to-be-treated optical signals with the unequal wavelengths are unequal (see Mehrabian: Fig.5, [0050], “voltages generated using the stored analog signals of the memristive memory 550 to modulate the laser source 540 for the filters 110. In accordance with at least one example of this disclosure, each signal generated by DAC 530 and DAC 535 are used to modulate a laser beam generated by laser source 540 of a specific wavelength λ”) Regarding Claim 7, As shown above, Mehrabian and Young teach all the limitation of claim 1. Mehrabian and Young further teach the system wherein: when the to-be-treated optical signals with four wavelengths are inputted into the micro- ring-resonator array, determining the target value to be ¼ (see Mehrabian Fig.1, [0031], “ collection of kernels 110 are processed simultaneously on a single input 120. Each input 120 is assigned to a distinct wavelength of light.” … [0032], using N wavelengths of light, it is possible to establish up to N.sup.2 independent connections, in accordance with this disclosure. The number of independent connections which are possible enable a CNN to operate on much larger datasets than non-photonic neural network.”) ; and when the target value be 1/4, obtaining the operation result of 2x2 average pooling of the neural network (see Mehrabian Fig.1, [0031], “The wavelength-division multiplexer 330 is configured to receive (laser) signals 310a having wavelengths λ.sub.1 . . . λ.sub.n, (e.g., 120, 230). The EWMM circuit 340 contains a MRR 340a. The EWMM circuit 340 is connected to a balance detection summation circuit 350.”) average pooling of the neural network (see Young Fig.1, [0031], “after multiplication by the factor equal to the inverse of the number of elements in the window, i.e., 1/9, rescaling corner elements by a scale of 9/4 will result in the corner value of the summer output tensor being divided by 4, which is appropriate since applying the 3×3 pooling kernel to corner elements of the input tensor results in 4 elements of the input tensor being summed to produce a corner element of the summed output tensor.”) It would have been obvious to a person of ordinary skill in the art, before the effective filing date of the invention, to modify the photonic neural network of Mehrabian to perform average pooling as taught by Young. After modification of Mehrabian, the photonic neural network that perform the summation of multiple wavelength encoded optical input using micro-ring resonators and also perform average pooling of neural network in order to achieve standard neural function. One would have been motivated to make such a combination in order to improve neural network function by pooling the outputs can also reduce dimensionality of inputs received at the subsequent neural network layers while maintaining desired characteristics of the outputs before pooling, which can improve efficiency without significantly compromising the quality of inferences generated by the neural networks. (see Young: [0025]) Regarding Claim 8, Claim 8 is directed to a computer device and has similar /same claim limitation as claim 1 and is rejected under the same rationale. Regarding Claim 9, Claim 9 is directed to a non-transitory computer-readable storage mediums claim and has similar /same claim limitation as claim 1 and is rejected under the same rationale. Regarding Claim 10, As shown above, Mehrabian and Young teach all the limitation of claim 1. Mehrabian further teach the system wherein: in the average pooling, an average value of neurons in a pooling core is calculated and used as an output, (see Young: Fig.7, [0087], “vided by a number of elements in the window of the average pooling neural network layer to generate an initial output tensor (704). For example, control signals may be generated by the sequencer 306 based on instructions received at the host interface 302 from the neural network implementation engine for multiplying each element of the summed output tensor by a factor equal to an inverse of the number of elements in the window of the average pooling neural network layer.) Regarding Claim 11, As shown above, Mehrabian and Young teach all the limitation of claim 10. Mehrabian and Young further teach the limitation wherein: one time of operation in the average pooling is expressed as: Y = 1 n 2 ∑ i = 1 n 2 x i , where nxn is a size of the pooling core, x, is a neuron inputted into a pooling layer, and y is a neuron outputted by the pooling layer (see Young: Fig.9, [0111], “The masking tensor of FIG. 9 may be determined by first determining a least common denominator of the number of elements in the window specified for the average pooling layer of the neural network, and a numbers of elements of the input tensor that are used in generating each element of the summed output tensor. For the example of FIG. 9, the least common denominator will therefore be 36, i.e., the least common denominator of 9, i.e., the number of elements in the 3×3 window, 4, i.e., the number of elements of the input tensor summed to generate corner elements of the summed output tensor, and 6, i.e., the number of elements of the input tensor summed to generate edge elements of the summed output tensor.”) It would have been obvious to a person of ordinary skill in the art, before the effective filing date of the invention, to modify the photonic neural network of Mehrabian to perform average pooling as taught by Young. After modification of Mehrabian, the photonic neural network that perform the summation of multiple wavelength encoded optical input using micro-ring resonators and also perform average pooling of neural network in order to achieve standard neural function. One would have been motivated to make such a combination in order to improve neural network function by pooling the outputs can also reduce dimensionality of inputs received at the subsequent neural network layers while maintaining desired characteristics of the outputs before pooling, which can improve efficiency without significantly compromising the quality of inferences generated by the neural networks. (see Young: [0025]) Regarding Claim 12, As shown above, Mehrabian and Young teach all the limitation of claim 11. Mehrabian and Young further teach the limitation wherein: when n=2, one time of operation in the pooling layer is expressed as: y = x 1 +   x 2 + +   x 3     +   x 4     4 (See Young: Abstract : “average pooling neural network layer and composed of elements that are each an identity matrix to generate a first tensor, and performing operations to cause each element of the first tensor to be divided by a number of elements in the window of the average pooling neural network layer to generate an initial output tensor.”) It would have been obvious to a person of ordinary skill in the art, before the effective filing date of the invention, to modify the photonic neural network of Mehrabian to perform average pooling as taught by Young. After modification of Mehrabian, the photonic neural network that perform the summation of multiple wavelength encoded optical input using micro-ring resonators and also perform average pooling of neural network in order to achieve standard neural function. One would have been motivated to make such a combination in order to improve neural network function by pooling the outputs can also reduce dimensionality of inputs received at the subsequent neural network layers while maintaining desired characteristics of the outputs before pooling, which can improve efficiency without significantly compromising the quality of inferences generated by the neural networks. (see Young: [0025]) Regarding Claim 13, As shown above, Mehrabian and Young teach all the limitation of claim 2. Mehrabian further teach the system wherein: the micro-ring resonators are silicon- based micro-ring resonators of an All-pass type (see Mehrabian: Fig.4, [0037], “EOM 420 (310) has a silicon-based p-i-n junction that is side coupled to a wave guide (e.g., 480) or to a p-n junction (e.g., 450, 460)”) Regarding Claim 17, Claim 17 is directed to a computer device claim and has similar/same claim limitation as Claim 2 and is rejected under the same rationale. Regarding Claim 18, Claim 18 is directed to a computer device claim and has similar/same claim limitation as Claim 3 and is rejected under the same rationale. Regarding Claim 19, Claim 19 is directed to a non-transitory computer-readable storage medium claim and has similar/same claim limitation as Claim 2 and is rejected under the same rationale. Regarding Claim 20, Claim 19 is directed to a non-transitory computer-readable storage medium claim and has similar/same claim limitation as Claim 3 and is rejected under the same rationale. Claims 14-16 are rejected under 35 U.S.C. 103 as being unpatentable over Mehrabian and Young as applied to claims 1-13 and 17-20 as shown above and in further view of in further view of Bangari (NPL: Title: Digital Electronics and Analog Photonics for Convolutional Neural Networks (DEAP-CNNs), Pub. Data; 23 Apr 2019) Regarding Claim 14, As shown above, Mehrabian and Young teach all the limitation of claim 13. Mehrabian and Young does not teach explicitly the transfer function of an intensity of light expression as stated in claim 14. However as shown below, Bangari teaches the method according to claim 13, wherein: the transfer function of an intensity of light exiting from a through hole at a direct-passing end and an intensity of light entering a input port of the all-pass micro-ring resonator is expressed as: T n ( ∅ i ) = a 2 - 2 r a c o s ∅ i +   r 2 1 -   2 r a c o s ∅ i + ( a r ) 2 , where ∅ i , is a phase of the micro-ring resonator, r is a self-coupling coefficient, and a defines a propagation loss of a ring-type directional coupler. See Bangari: Equation 6 PNG media_image1.png 460 1036 media_image1.png Greyscale It would have been obvious to a person of ordinary skill in the art, before the effective filing date of the invention, to modify the photonic neural network of Mehrabian to perform average pooling as taught by Young and apply transfer function of an intensity of light exiting from a through hole at a direct-passing end and an intensity of light entering a input port of the all-pass micro-ring resonator is expressed as taught by Bangari. After modification of Mehrabian, the photonic neural network that perform the summation of multiple wavelength encoded optical input using micro-ring resonators and also perform average pooling of neural network in order to achieve standard neural function. One would have been motivated to make such a combination in order to improve neural network function by pooling the outputs can also reduce dimensionality of inputs received at the subsequent neural network layers while maintaining desired characteristics of the outputs before pooling, which can improve efficiency without significantly compromising the quality of inferences generated by the neural networks. Regarding Claim 15, As shown above, Mehrabian, Young and Bangari teach all the limitation of claim 14. Bangari further teaches the system wherein: a value range of the transfer function is [0,1] (see Bangari: Sec. 11.2, Page 17, “In the case where the coupling losses are negligible, a ≈ 1, the relationship between the add-drop through and drop transfer functions is Tp = Td − 1. In addition, if we connect the through and drop ports into a balanced photodiode and TIA as in Fig. 3(a), we get an effective transfer function of g(Tp −Td) where g is the gain of the TIA.”) See motivation to combine in Claim 14. Regarding Claim 16 As shown above, Mehrabian, Young and Bangari teach all the limitation of claim 13. Mehrabian and Young does not teach explicitly amplitude of an inputted optical signal is Ein and a light intensity of the inputted optical signal is |Ein|2, a light intensity outputted from the micro-ring resonator as stated in claim 16. However as shown below, Bangari teaches the method according to claim 13, wherein: when an amplitude of an inputted optical signal is Ein and a light intensity of the inputted optical signal is |Ein|2, a light intensity outputted from the micro-ring resonator is: |Eout|2= Tn( ∅ i )|Ein|2; See Bangari: Equation 8 and Equation 9 PNG media_image2.png 476 982 media_image2.png Greyscale And the phase ∅ i , is expressed as: ( ∅ i ) = 4 π 2 .   R .   n e f f   λ i , where λ i , is a wavelength, R is the radius of the micro-ring resonator, and neff is an effective refractive index of light. See Bangari: Equation 7 PNG media_image3.png 292 912 media_image3.png Greyscale It would have been obvious to a person of ordinary skill in the art, before the effective filing date of the invention, to modify the photonic neural network of Mehrabian to perform average pooling as taught by Young and apply light intensity outputted from the micro-ring resonator from a through hole at a direct-passing end and an intensity of light entering a input port of the all-pass micro-ring resonator is expressed as taught by Bangari. After modification of Mehrabian, the photonic neural network that perform the summation of multiple wavelength encoded optical input using micro-ring resonators and also perform average pooling of neural network in order to achieve standard neural function. One would have been motivated to make such a combination in order to improve neural network function by pooling the outputs can also reduce dimensionality of inputs received at the subsequent neural network layers while maintaining desired characteristics of the outputs before pooling, which can improve efficiency without significantly compromising the quality of inferences generated by the neural networks. Conclusion The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. PGPUB NUMBER: INVENTOR-INFORMATION: TITLE / DESCRIPTION US 11546077 B1 Parker; Charles Title: Scalable, ultra-low-latency photonic tensor processor Description: Deep neural networks (DNNs) have become very popular in many areas, especially classification and prediction. However, as the number of neurons in the DNN increases to solve more complex problems, the DNN becomes limited by the latency and power consumption of existing hardware. A scalable, ultra-low latency photonic tensor processor can compute DNN layer outputs in a single shot. US 20240127040 A1 Shiflett, Kyle Title: Photonic accelerator for deep neural networks Description: The present invention relates generally to neural networks and, more particularly, to a neural network accelerator including photonic circuits. Any inquiry concerning this communication or earlier communications from the examiner should be directed to ZELALEM W SHALU whose telephone number is (571)272-3003. The examiner can normally be reached M- F 0800am- 0500pm. Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice. If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Cesar Paula can be reached on (571) 272-4128. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300. Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000. /Zelalem Shalu/Examiner, Art Unit 2145 /CESAR B PAULA/Supervisory Patent Examiner, Art Unit 2145