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 .
Status of the Application
This Office action is in response to the amendment and remarks filed 08/05/2026. Claims 1, 7, 8, 15, 16, 20 and 21 have been amended; claims 23-28 have been added; claim 22 has been cancelled. Claims 1, 4-21 and 23-28 are pending and are examined herein. Applicant’s amendment adding claims 23-28 necessitated the new grounds of rejection of those claims presented in this Office action. THIS ACTION IS MADE FINAL.
Response to Arguments
Applicant’s arguments filed 08/05/2026 have been fully considered but they are not persuasive.
With respect to Applicant’s argument that “However, the quoted language actually appears in paragraph [0182] of Smith. While this may be a minor clerical error, Applicant respectfully requests that the record be corrected” (Remarks, page 9, paragraph 5). In response, the Examiner agrees and appreciates the correction. The records of cited paragraph numbers are now corrected as noted. However, the correction points to the same teaching the prior Office action quoted and relied on, so the corrections are not new grounds of rejection. The corrected locators appear in the rejections below.
In regard to Applicant’s argument that “These paths synchronize a single physical node with its corresponding redundant digital twin copies for fault tolerance purposes. They do not represent interconnections among different nodes in the network” (Remarks, page 11, paragraph 3), and that the claimed interconnections “refers to the network topology—how different nodes in the radio network are connected to each other to enable data flow” (Remarks, page 10, paragraph 1), the Examiner respectfully disagrees.
Under the broadest reasonable interpretation, the claims are read in light of the specification as one of ordinary skill in the art would read them (MPEP 2111; In re Am. Acad. of Sci. Tech. Ctr., 367 F.3d 1359, 1364 (Fed. Cir. 2004)). Claims 1, 11 and 19 recite a digital twin representation “including virtual representations of the plurality of nodes and interconnections among the plurality of nodes” and do not say what form the interconnections take. The specification uses the same general words (US 2025/0024278 A1, para [0020] “The digital twin 108 also includes virtual representations of interconnections among the plurality of nodes.”) and presents the directed graph with data flow edges as one implementation (US 2025/0024278 A1, para [0020] “In some implementations, the digital twin 108 can include a virtual representation of the network 100 in the form of a directed graph of the various components of the network environment 100.”).
The Examiner does not rely on the synchronization paths of Smith para [0182] alone. Smith’s digital twin model includes a digital twin for each of the physical nodes (Smith, para [0182] “A digital twin model may be generated for physical nodes in an edge network (e.g., at operation 2305). The digital twin model may include a digital twin for a physical node of the physical nodes.”), and its synchronization paths connect a physical node with its digital twins (Smith, para [0182] “the digital twin model may include synchronization paths between the physical node and the digital twin and between a first digital twin of the physical node and a second digital twin of the physical node”). The same framework relates different nodes to one another: a physical node, its peer physical nodes and the digital twins of the peer nodes form a coalition from whose members output metrics are received (Smith, para [0185] “a coalition may be established that includes that includes peer physical nodes, digital twins of the peer nodes, and the physical node or the digital twin for the physical node. Output metrics may be received from members of the coalition.”)
Relations among different nodes of the model, through which the output metrics of the members are received, are virtual representations of interconnections among the plurality of nodes under the broadest reasonable interpretation. The correspondence is set forth in the rejection of claim 1 below.
In response to applicant’s argument that the references fail to show certain features of the invention, it is noted that the features upon which applicant relies (i.e., a network topology in which the edges of a directed graph indicate the direction of data flow among different nodes) are not recited in the rejected claim(s). Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims. See In re Van Geuns, 988 F.2d 1181, 26 USPQ2d 1057 (Fed. Cir. 1993). Those features are recited in new claims 23, 25 and 27, which are addressed below.
As to Applicant’s argument that “Kottkamp does not cure this deficiency” (Remarks, page 11, paragraph 4), the Examiner respectfully disagrees. In response to applicant’s arguments against the references individually, one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986).
The rejection does not rely on Kottkamp for the digital twin representation or its interconnections; Smith supplies those, as set forth above. Kottkamp supplies the trained machine learning model (Kottkamp, para [0013], “a processing unit of the monitoring system executes a spectral anomaly neural network trained by a machine learning algorithm in a training system”), the monitoring of parameters that reveals an out of range condition (Kottkamp, para [0016], “the monitoring system further derives additional parameters of an active communication link in the cellular wireless network, in particular the achieved data rate/bit rate, the signal-to-noise ratio and/or the latency time, and the processing unit further obtains and processes the additional parameters to detect a spectral anomaly information”), and the output of the detected condition (Kottkamp, para [0018], “The monitoring system can further comprise an output unit, in particular a display or a monitor, for presenting the detected spectral anomaly information”).
Turning to Applicant’s argument that “Neither Smith nor Kottkamp teaches or suggests representing the digital twin as a directed graph with edges indicating data flow direction among different nodes, or identifying a representative subset that enables computationally efficient monitoring while reliably representing the entire network’s operating conditions” (Remarks, page 12, paragraph 2), the argument is directed to new claims 23-28, which are rejected below under new grounds of rejection necessitated by the amendment. Dong represents a radio access network as a directed graph whose links indicate the direction of data traffic flow among the network nodes (Dong, para [0046], “converting an undirected graph representation of an IP RAN network to a directed graph representation . . . A directed graph representation may comprise a plurality of CSG links that connect a plurality of adjacent network nodes and the CSG links may indicate the data traffic flow directions among the network node”). Abusch-Magder monitors a selected part of a wireless network that represents the entire network (Abusch-Magder, para [0049], “For the selected neighborhood (at 520), which is representative of the entire network, one or more given parameters thereof may be monitored (530) in an effort to evaluate network performance or determine current status of the network”), with lower cost in computational resources (Abusch-Magder, para [0066], “larger neighborhoods cost more to monitor (due to monitoring equipment, computational resources such as storage and processing servers, increase in communication on backhaul etc.), while smaller neighborhoods might be less accurate”) and in less time than evaluating the full network (Abusch-Magder, para [0055], “results comparable to full-evaluation greedy on a realistic 206-cell network were obtained using an average neighborhood size of 36.7. This simplification provided an algorithm runtime that was roughly 15 times faster than the conventional full-evaluation method”). The combinations are set forth in the rejections of claims 23 and 24 below.
With respect to Applicant’s argument that “Applicant respectfully submits that claims 1 and 4-22 are patentable over the cited combination of Smith and Kottkamp, and that new claims 23-28 are also patentable” (Remarks, page 12, paragraph 2), the dependent claims are not argued apart from their independent claims and new claims 23-28. They remain rejected for the reasons given for their parent claims and for the additional reasons set forth below.
For at least these reasons, the rejections are maintained and are restated below to address the amended claim language. The rejections of new claims 23-28 over Smith in view of Kottkamp and further in view of Dong, and over Smith in view of Kottkamp and further in view of Abusch-Magder, are new grounds of rejection necessitated by Applicant’s amendment adding those claims.
Claim Rejections - 35 USC § 103
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 (i.e., changing from AIA to pre-AIA ) 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.
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 and 4-21 are rejected under 35 U.S.C. 103 as being unpatentable over Smith (US 2022/0191648 A1) in view of Kottkamp (US 2021/0136603 A1).
Regarding claim 1, (Currently Amended) Smith discloses: A method comprising: accessing, by one or more processing devices, a digital twin representation of a radio network (RN), the radio network including a plurality of nodes configured to enable communications between wireless user equipments (UEs),, because Smith teaches a digital twin framework for an edge network with at least one processor that executes its operations, and a digital twin model generated for the physical nodes of the network, the digital twin being a virtual representation of a physical system. The edge cloud is co-located at access points or base stations close to user equipment, which means that the processor accesses a digital twin representation of a radio network including a plurality of nodes configured to enable communications between wireless user equipments: (Smith, para [0002] “A digital twin (DT) is a virtual representation of elements and dynamics of a physical system/world. The virtual representation is used along with machine learning-based (ML-based) intelligent communication-compute systems” . . . para [0037] “the edge cloud 110 is co-located at an edge location, such as an access point or base station 140 . . . The edge cloud 110 is located much closer to the endpoint (consumer and producer) data sources 160 (e.g., autonomous vehicles 161, user equipment 162” . . . para [0182] “A digital twin model may be generated for physical nodes in an edge network (e.g., at operation 2305). The digital twin model may include a digital twin for a physical node of the physical nodes.” . . . para [0188] “Example 1 is a system for a digital twin framework for an edge network comprising: at least one processor; and memory including instructions that, when executed by the at least one processor, cause the at least one processor to perform operations”).
In addition, Smith discloses: the digital twin representation including virtual representations of the plurality of nodes and interconnections among the plurality of nodes;, as Smith describes a digital twin model that includes a digital twin for a physical node of the physical nodes and synchronization paths between a physical node and its digital twins, and a coalition of peer physical nodes and the digital twins of the peer nodes from whose members output metrics are received (i.e., “interconnections among the plurality of nodes” as claimed), such that the digital twin representation includes virtual representations of the plurality of nodes and interconnections among the plurality of nodes: (Smith, para [0182] “the digital twin model may include synchronization paths between the physical node and the digital twin and between a first digital twin of the physical node and a second digital twin of the physical node” . . . para [0185] “a coalition may be established that includes that includes peer physical nodes, digital twins of the peer nodes, and the physical node or the digital twin for the physical node. Output metrics may be received from members of the coalition”).
Although Smith teaches a digital representation of the physical world built from collected sensor data by an AI/ML algorithm such as a convolutional neural network, modelling the entire physical world or a subset of its objects: (Smith, para [0117]), Smith does not explicitly disclose identifying, using a trained machine learning model, a first subset of the digital twin representation representative of operating conditions of the radio network for a first time range, determining that at least one of the operating conditions is outside of a corresponding expected range, and generating a first output signal.
However, Smith in view of Kottkamp discloses identifying, using a trained machine learning model, a first subset of the digital twin representation, wherein the first subset is identified by the trained machine learning model as being representative of operating conditions of the radio network for a first time range; since Kottkamp explains a spectral anomaly neural network, trained by a machine learning algorithm, that a processing unit executes on the spectrum an RF receiver derives over time and frequency as a waterfall diagram of the cellular network (i.e., “for a first time range” as claimed), whereby the trained model identifies a first subset representative of operating conditions of the radio network for a first time range (Kottkamp, para [0013], “a processing unit of the monitoring system executes a spectral anomaly neural network trained by a machine learning algorithm in a training system” . . . para [0015], “the RF receiver monitors and derives the spectrum over time and frequency to produce a waterfall diagram, wherein the processing unit obtains and processes the waterfall diagram”).
Moreover, Kottkamp discloses determining, by the one or more processing devices via monitoring one or more parameters of the first subset of the digital twin representation, that at least one of the operating conditions of the radio network is outside of a corresponding expected range; and in that Kottkamp shows the monitoring system deriving parameters of an active communication link, such as the achieved data rate, the signal-to-noise ratio and the latency time, and the processing unit processing those parameters to detect spectral anomaly information, with the result that the processing unit, by monitoring one or more parameters, determines that at least one of the operating conditions of the radio network is outside of a corresponding expected range (Kottkamp, para [0016], “the monitoring system further derives additional parameters of an active communication link in the cellular wireless network, in particular the achieved data rate/bit rate, the signal-to-noise ratio and/or the latency time, and the processing unit further obtains and processes the additional parameters to detect a spectral anomaly information”).
Furthermore, Kottkamp discloses in response to determining that at least one of the operating conditions of the radio network is outside of the corresponding expected range, generating a first output signal indicative of a condition of the radio network during the first time range. given that Kottkamp sets out an output unit, such as a display or a monitor, presenting the spectral anomaly information that the processing unit has detected, so that upon the detection a first output signal indicative of a condition of the radio network during the first time range is generated (Kottkamp, para [0018], “The monitoring system can further comprise an output unit, in particular a display or a monitor, for presenting the detected spectral anomaly information”).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to apply the trained spectral anomaly neural network of Kottkamp, with its monitored link parameters and its output unit, to the digital twin representation of Smith, so that the trained model identifies the portion of the digital twin representation that represents the operating conditions of the network over a time window, monitors its parameters, and outputs a signal when a condition falls outside its expected range, in order to detect abnormal operating conditions of the network from its digital twin. Smith uses its digital twin together with machine learning based intelligent communication and compute systems and seeks to identify errors of the physical nodes the twin models, and Kottkamp teaches a trained neural network that detects anomalies in a cellular wireless network from monitored parameters and presents them on an output unit. The combination applies a known machine learning technique for its established purpose to a representation that Smith already builds with machine learning, and it would have predictably yielded detection and reporting of out of range operating conditions for each time range. Thus, one of ordinary skill would have had a reasonable expectation of success, since both references process network data with trained machine learning models and the combination changes only the data on which Kottkamp’s trained network operates.
Regarding claim 4, which depends on claim 1, (Previously Presented) Smith in view of Kottkamp discloses The method of claim 1, further comprising: identifying, using the trained machine learning model, a second subset of the digital twin representation, wherein the second subset is representative of operating conditions of the radio network for a second time range different from the first time range., as Kottkamp further discloses the RF receiver deriving the spectrum over time and frequency to produce a waterfall diagram that the processing unit obtains and processes, the diagram covering the spectrum at different times, with the result that the trained model identifies a second subset representative of operating conditions of the radio network for a second time range different from the first time range (Kottkamp, para [0015], “the RF receiver monitors and derives the spectrum over time and frequency to produce a waterfall diagram, wherein the processing unit obtains and processes the waterfall diagram”).
Consequently, Smith and Kottkamp are combined for the reasons set forth in the rejection of claim 1 above.
Regarding claim 5, which depends on claim 4, (Previously Presented) Smith in view of Kottkamp discloses The method of claim 4, wherein the second subset is different from the first subset., as Kottkamp further discloses the spectrum derived over time and frequency as a waterfall diagram, each time window of which holds the spectrum measured in that window, so that the second subset is different from the first subset (Kottkamp, para [0015], “the RF receiver monitors and derives the spectrum over time and frequency to produce a waterfall diagram, wherein the processing unit obtains and processes the waterfall diagram”).
For these reasons, Smith and Kottkamp are combined for the reasons set forth in the rejection of claim 1 above.
Regarding claim 6, which depends on claim 4, (Previously Presented) Smith in view of Kottkamp discloses The method of claim 4, further comprising: determining, by the one or more processing devices via monitoring one or more parameters of the second subset of the digital twin representation, that at least one of the operating conditions of the radio network is outside of a corresponding expected range; and in response to determining that at least one of the operating conditions of the radio network is outside of the corresponding expected range, generating a second output signal indicative of a condition of the radio network during the second time range., as Kottkamp further discloses the processing unit processing the monitored link parameters, such as the data rate, the signal-to-noise ratio and the latency time, to detect spectral anomaly information, and the output unit presenting the detected information, which means that when the parameters of the second subset are monitored, an operating condition is determined to be outside of a corresponding expected range and a second output signal indicative of a condition of the radio network during the second time range is generated (Kottkamp, para [0016], “the monitoring system further derives additional parameters of an active communication link in the cellular wireless network, in particular the achieved data rate/bit rate, the signal-to-noise ratio and/or the latency time, and the processing unit further obtains and processes the additional parameters to detect a spectral anomaly information” . . . para [0018], “The monitoring system can further comprise an output unit, in particular a display or a monitor, for presenting the detected spectral anomaly information”).
Therefore, Smith and Kottkamp are combined for the reasons set forth in the rejection of claim 1 above.
Regarding claim 7, which depends on claim 1, (Currently Amended) Smith discloses The method of claim 1, further comprising generating a control signal configured to address the at least one of the operating conditions of the radio network., as Smith further discloses identifying an error of a physical node or its digital twin and then updating the digital twin model to halt communication with that node or its twin, such that a control signal is generated that is configured to address the at least one of the operating conditions of the radio network (Smith, para [0188] “identify an error of the physical node or the digital twin for the physical node; update the digital twin model to halt communication with the physical node or the digital twin of the physical node”).
Regarding claim 8, which depends on claim 7, (Currently Amended) Smith discloses The method of claim 7, further comprising: identifying, from the first subset of the digital twin representation, one or more nodes as a root cause of the at least one of the operating conditions of the radio network being outside of the corresponding expected range; and, as Smith further discloses an error of a physical node being identified from the inconsistency of the output metrics of that node, or of its digital twin, with the output metrics of its peer physical nodes and of their digital twins, whereby one or more nodes are identified as a root cause of the at least one of the operating conditions of the radio network being outside of the corresponding expected range (Smith, para [0185] “it may be determined that output metrics from the physical node or the digital twin for the physical node are inconsistent with output metrics from the peer physical nodes and digital twins of the peer physical nodes. The error may be identified based on the inconsistency”).
In addition, Smith discloses generating the control signal such that the control signal is configured to affect operations of the one or more nodes., as Smith further discloses updating the digital twin model to halt communication with the physical node or its digital twin whose error was identified, with the result that the control signal is configured to affect operations of the one or more nodes (Smith, para [0188] “identify an error of the physical node or the digital twin for the physical node; update the digital twin model to halt communication with the physical node or the digital twin of the physical node”).
Regarding claim 9, which depends on claim 1, (Previously Presented) Smith in view of Kottkamp discloses The method of claim 1, wherein the one or more parameters comprise at least one of: a number of beams, a number of slices, a radio power level, or RF characteristics., as Kottkamp further discloses the monitored parameters of an active communication link including the achieved data rate, the signal-to-noise ratio and the latency time, the signal-to-noise ratio being a characteristic of the received radio frequency signal, with the result that the one or more parameters comprise RF characteristics (Kottkamp, para [0016], “the monitoring system further derives additional parameters of an active communication link in the cellular wireless network, in particular the achieved data rate/bit rate, the signal-to-noise ratio and/or the latency time, and the processing unit further obtains and processes the additional parameters to detect a spectral anomaly information”). Furthermore, claim 9 recites its alternatives joined by or, so the claim is satisfied when any one of them is satisfied. The alternative relied on is the recitation of RF characteristics, which is the limitation mapped above.
Consequently, Smith and Kottkamp are combined for the reasons set forth in the rejection of claim 1 above.
Regarding claim 10, which depends on claim 1, (Previously Presented) Smith discloses The method of claim 1, wherein the one or more parameters are determined by the trained machine learning model., as Smith further discloses data collected from the physical product via sensors being provided to a function using an AI/ML algorithm, such as a convolutional neural network, that builds the digital representation of the physical world, so that the parameters the digital twin representation holds are determined by the trained machine learning model (Smith, para [0117] “There may be a large amount of data, collected from the physical product (PP) via sensors, that is provided to a function (e.g., using an AI/ML algorithm such as a convolutional neural network, etc.) to build a digital representation of the physical world”).
Regarding claim 11, (Previously Presented) Smith discloses: One or more non-transitory computer storage media encoded with computer program instructions that when executed by one or more computers cause the one or more computers to perform operations comprising: accessing a digital twin representation of a radio network, the radio network including a plurality of nodes configured to enable communications between wireless user equipments (UEs),, because Smith teaches a digital twin framework for an edge network whose memory holds instructions that, when executed by at least one processor, cause the processor to perform its operations, the memory being a non-transitory storage medium, and a digital twin model generated for the physical nodes of the network, the digital twin being a virtual representation of a physical system. The edge cloud is co-located at access points or base stations close to user equipment, which means that the instructions cause the computers to access a digital twin representation of a radio network including a plurality of nodes configured to enable communications between wireless user equipments: (Smith, para [0002] “A digital twin (DT) is a virtual representation of elements and dynamics of a physical system/world. The virtual representation is used along with machine learning-based (ML-based) intelligent communication-compute systems” . . . para [0037] “the edge cloud 110 is co-located at an edge location, such as an access point or base station 140 . . . The edge cloud 110 is located much closer to the endpoint (consumer and producer) data sources 160 (e.g., autonomous vehicles 161, user equipment 162” . . . para [0182] “A digital twin model may be generated for physical nodes in an edge network (e.g., at operation 2305). The digital twin model may include a digital twin for a physical node of the physical nodes.” . . . para [0188] “Example 1 is a system for a digital twin framework for an edge network comprising: at least one processor; and memory including instructions that, when executed by the at least one processor, cause the at least one processor to perform operations”).
Furthermore, Smith discloses: the digital twin representation including virtual representations of the plurality of nodes and interconnections among the plurality of nodes;, as Smith describes a digital twin model that includes a digital twin for a physical node of the physical nodes and synchronization paths between a physical node and its digital twins, and a coalition of peer physical nodes and the digital twins of the peer nodes from whose members output metrics are received (i.e., “interconnections among the plurality of nodes” as claimed), such that the digital twin representation includes virtual representations of the plurality of nodes and interconnections among the plurality of nodes: (Smith, para [0182] “the digital twin model may include synchronization paths between the physical node and the digital twin and between a first digital twin of the physical node and a second digital twin of the physical node” . . . para [0185] “a coalition may be established that includes that includes peer physical nodes, digital twins of the peer nodes, and the physical node or the digital twin for the physical node. Output metrics may be received from members of the coalition”).
Although Smith teaches a digital representation of the physical world built from collected sensor data by an AI/ML algorithm such as a convolutional neural network, modelling the entire physical world or a subset of its objects: (Smith, para [0117]), Smith does not expressly teach identifying, using a trained machine learning model, a first subset of the digital twin representation representative of operating conditions of the radio network for a first time range, determining that at least one of the operating conditions is outside of a corresponding expected range, and generating an output signal.
However, Smith in view of Kottkamp discloses identifying a first subset of the digital twin representation, wherein the first subset is identified by a trained machine learning model as being representative of operating conditions of the radio network for a first time range; since Kottkamp explains a spectral anomaly neural network, trained by a machine learning algorithm, that a processing unit executes on the spectrum an RF receiver derives over time and frequency as a waterfall diagram of the cellular network (i.e., “for a first time range” as claimed), whereby the trained model identifies a first subset representative of operating conditions of the radio network for a first time range (Kottkamp, para [0013], “a processing unit of the monitoring system executes a spectral anomaly neural network trained by a machine learning algorithm in a training system” . . . para [0015], “the RF receiver monitors and derives the spectrum over time and frequency to produce a waterfall diagram, wherein the processing unit obtains and processes the waterfall diagram”).
Moreover, Kottkamp discloses determining, via monitoring one or more parameters of the first subset of the digital twin representation, that at least one of the operating conditions of the radio network is outside of a corresponding expected range; and in that Kottkamp shows the monitoring system deriving parameters of an active communication link, such as the achieved data rate, the signal-to-noise ratio and the latency time, and the processing unit processing those parameters to detect spectral anomaly information, with the result that the processing unit, by monitoring one or more parameters, determines that at least one of the operating conditions of the radio network is outside of a corresponding expected range (Kottkamp, para [0016], “the monitoring system further derives additional parameters of an active communication link in the cellular wireless network, in particular the achieved data rate/bit rate, the signal-to-noise ratio and/or the latency time, and the processing unit further obtains and processes the additional parameters to detect a spectral anomaly information”).
In addition, Kottkamp discloses in response to determining that at least one of the operating conditions of the radio network is outside of the corresponding expected range, generating an output signal indicative of a condition of the radio network during the first time range. given that Kottkamp sets out an output unit, such as a display or a monitor, presenting the spectral anomaly information that the processing unit has detected, so that upon the detection an output signal indicative of a condition of the radio network during the first time range is generated (Kottkamp, para [0018], “The monitoring system can further comprise an output unit, in particular a display or a monitor, for presenting the detected spectral anomaly information”).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to apply the trained spectral anomaly neural network of Kottkamp, with its monitored link parameters and its output unit, to the digital twin representation of Smith, so that the trained model identifies the portion of the digital twin representation that represents the operating conditions of the network over a time window, monitors its parameters, and outputs a signal when a condition falls outside its expected range, in order to detect abnormal operating conditions of the network from its digital twin. Smith uses its digital twin together with machine learning based intelligent communication and compute systems and seeks to identify errors of the physical nodes the twin models, and Kottkamp teaches a trained neural network that detects anomalies in a cellular wireless network from monitored parameters and presents them on an output unit. The combination applies a known machine learning technique for its established purpose to a representation that Smith already builds with machine learning, and it would have predictably yielded detection and reporting of out of range operating conditions for each time range. One of ordinary skill would have had a reasonable expectation of success, since both references process network data with trained machine learning models and the combination changes only the data on which Kottkamp’s trained network operates.
Regarding claim 12, which depends on claim 11, (Previously Presented) Smith in view of Kottkamp discloses The one or more non-transitory computer storage media of claim 11, the operations further comprising: identifying, using the trained machine learning model, a second subset of the digital twin representation, wherein the second subset is representative of operating conditions of the radio network for a second time range different from the first time range., as Kottkamp further discloses the RF receiver deriving the spectrum over time and frequency to produce a waterfall diagram that the processing unit obtains and processes, the diagram covering the spectrum at different times, such that the trained model identifies a second subset representative of operating conditions of the radio network for a second time range different from the first time range (Kottkamp, para [0015], “the RF receiver monitors and derives the spectrum over time and frequency to produce a waterfall diagram, wherein the processing unit obtains and processes the waterfall diagram”).
Accordingly, Smith and Kottkamp are combined for the reasons set forth in the rejection of claim 11 above.
Regarding claim 13, which depends on claim 12, (Previously Presented) Smith in view of Kottkamp discloses The one or more non-transitory computer storage media of claim 12, wherein the second subset is different from the first subset., as Kottkamp further discloses the spectrum derived over time and frequency as a waterfall diagram, each time window of which holds the spectrum measured in that window, whereby the second subset is different from the first subset (Kottkamp, para [0015], “the RF receiver monitors and derives the spectrum over time and frequency to produce a waterfall diagram, wherein the processing unit obtains and processes the waterfall diagram”).
Thus, Smith and Kottkamp are combined for the reasons set forth in the rejection of claim 11 above.
Regarding claim 14, which depends on claim 12, (Previously Presented) Smith in view of Kottkamp discloses The one or more non-transitory computer storage media of claim 12, the operations further comprising: determining, via monitoring one or more parameters of the second subset of the digital twin representation, that at least one of the operating conditions of the radio network is outside of a corresponding expected range; and in response to determining that at least one of the operating conditions of the radio network is outside of the corresponding expected range, generating a second output signal indicative of a condition of the radio network during the second time range., as Kottkamp further discloses the processing unit processing the monitored link parameters, such as the data rate, the signal-to-noise ratio and the latency time, to detect spectral anomaly information, and the output unit presenting the detected information, with the result that when the parameters of the second subset are monitored, an operating condition is determined to be outside of a corresponding expected range and a second output signal indicative of a condition of the radio network during the second time range is generated (Kottkamp, para [0016], “the monitoring system further derives additional parameters of an active communication link in the cellular wireless network, in particular the achieved data rate/bit rate, the signal-to-noise ratio and/or the latency time, and the processing unit further obtains and processes the additional parameters to detect a spectral anomaly information” . . . para [0018], “The monitoring system can further comprise an output unit, in particular a display or a monitor, for presenting the detected spectral anomaly information”).
Consequently, Smith and Kottkamp are combined for the reasons set forth in the rejection of claim 11 above.
Regarding claim 15, which depends on claim 11, (Currently Amended) Smith discloses The one or more non-transitory computer storage media of claim 11, the operations further comprising: generating a control signal configured to address the at least one of the operating conditions of the radio network., as Smith further discloses identifying an error of a physical node or its digital twin and then updating the digital twin model to halt communication with that node or its twin, so that a control signal is generated that is configured to address the at least one of the operating conditions of the radio network (Smith, para [0188] “identify an error of the physical node or the digital twin for the physical node; update the digital twin model to halt communication with the physical node or the digital twin of the physical node”).
Regarding claim 16, which depends on claim 15, (Currently Amended) Smith discloses The one or more non-transitory computer storage media of claim 15, the operations further comprising: identifying, from the first subset of the digital twin representation, one or more nodes as a root cause of the at least one of the operating conditions of the radio network being outside of the corresponding expected range; and, as Smith further discloses an error of a physical node being identified from the inconsistency of the output metrics of that node, or of its digital twin, with the output metrics of its peer physical nodes and of their digital twins, which means that one or more nodes are identified as a root cause of the at least one of the operating conditions of the radio network being outside of the corresponding expected range (Smith, para [0185] “it may be determined that output metrics from the physical node or the digital twin for the physical node are inconsistent with output metrics from the peer physical nodes and digital twins of the peer physical nodes. The error may be identified based on the inconsistency”).
Moreover, Smith discloses generating the control signal such that the control signal is configured to affect operations of the one or more nodes., as Smith further discloses updating the digital twin model to halt communication with the physical node or its digital twin whose error was identified, such that the control signal is configured to affect operations of the one or more nodes (Smith, para [0188] “identify an error of the physical node or the digital twin for the physical node; update the digital twin model to halt communication with the physical node or the digital twin of the physical node”).
Regarding claim 17, which depends on claim 11, (Previously Presented) Smith in view of Kottkamp discloses The one or more non-transitory computer storage media of claim 11, wherein the one or more parameters comprise at least one of: a number of beams, a number of slices, a radio power level, or RF characteristics., as Kottkamp further discloses the monitored parameters of an active communication link including the achieved data rate, the signal-to-noise ratio and the latency time, the signal-to-noise ratio being a characteristic of the received radio frequency signal, such that the one or more parameters comprise RF characteristics (Kottkamp, para [0016], “the monitoring system further derives additional parameters of an active communication link in the cellular wireless network, in particular the achieved data rate/bit rate, the signal-to-noise ratio and/or the latency time, and the processing unit further obtains and processes the additional parameters to detect a spectral anomaly information”).
Claim 17 recites its alternatives joined by or, so the claim is satisfied when any one of them is satisfied. The alternative relied on is the recitation of RF characteristics, which is the limitation mapped above.
Accordingly, Smith and Kottkamp are combined for the reasons set forth in the rejection of claim 11 above.
Regarding claim 18, which depends on claim 11, (Previously Presented) Smith discloses The one or more non-transitory computer storage media of claim 11, wherein the one or more parameters are determined by the trained machine learning model., as Smith further discloses data collected from the physical product via sensors being provided to a function using an AI/ML algorithm, such as a convolutional neural network, that builds the digital representation of the physical world, whereby the parameters the digital twin representation holds are determined by the trained machine learning model (Smith, para [0117] “There may be a large amount of data, collected from the physical product (PP) via sensors, that is provided to a function (e.g., using an AI/ML algorithm such as a convolutional neural network, etc.) to build a digital representation of the physical world”).
Regarding claim 19, (Previously Presented) the claim recites: A system comprising: one or more computers and one or more storage devices on which are stored instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising: accessing a digital twin representation of a radio network, the radio network including a plurality of nodes configured to enable communications between wireless user equipments (UEs), the digital twin representation including virtual representations of the plurality of nodes and interconnections among the plurality of nodes; identifying a first subset of the digital twin representation, wherein the first subset is identified by a trained machine learning model as being representative of operating conditions of the radio network for a first time range; determining, via monitoring one or more parameters of the first subset of the digital twin representation, that at least one of the operating conditions of the radio network is outside of a corresponding expected range; and in response to determining that at least one of the operating conditions of the radio network is outside of the corresponding expected range, generating an output signal indicative of a condition of the radio network during the first time range. Claim 19 is analogous to claim 11 and is rejected for the same reasons.
Regarding claim 20, which depends on claim 19, (Currently Amended) the claim recites: The system of claim 19, the operations further comprising: identifying, using the trained machine learning model, a second subset of the digital twin representation, wherein the second subset is representative of operating conditions of the radio network for a second time range different from the first time range. Claim 20 is analogous to claim 12 and is rejected for the same reasons.
Regarding claim 21, which depends on claim 20, (Currently Amended) Smith in view of Kottkamp discloses The system of claim 20, the operations further comprising: determining, by the one or more computers via monitoring one or more parameters of the second subset of the digital twin representation, that at least one of the operating conditions of the radio network is outside of a corresponding expected range; and in response to determining that at least one of the operating conditions of the radio network is outside of the corresponding expected range, generating a second output signal indicative of a condition of the radio network during the second time range., as Kottkamp further discloses the processing unit processing the monitored link parameters, such as the data rate, the signal-to-noise ratio and the latency time, to detect spectral anomaly information, and the output unit presenting the detected information, which means that when the one or more computers monitor the parameters of the second subset, an operating condition is determined to be outside of a corresponding expected range and a second output signal indicative of a condition of the radio network during the second time range is generated (Kottkamp, para [0016], “the monitoring system further derives additional parameters of an active communication link in the cellular wireless network, in particular the achieved data rate/bit rate, the signal-to-noise ratio and/or the latency time, and the processing unit further obtains and processes the additional parameters to detect a spectral anomaly information” . . . para [0018], “The monitoring system can further comprise an output unit, in particular a display or a monitor, for presenting the detected spectral anomaly information”).
Therefore, Smith and Kottkamp are combined for the reasons set forth in the rejection of claim 11 above.
Claims 23, 25 and 27 are rejected under 35 U.S.C. 103 as being unpatentable over Smith (US 2022/0191648 A1) in view of Kottkamp (US 2021/0136603 A1) and further in view of Dong (US 2016/0057629 A1).
Regarding claim 23, which depends on claim 1, (New) even though Smith in view of Kottkamp teaches the digital twin representation of the radio network, with virtual representations of its nodes and of the relations among them, a subset of which a trained machine learning model identifies and monitors: (Smith, para [0182], Kottkamp, para [0013]), Smith in view of Kottkamp stops short of disclosing that the digital twin representation comprises a directed graph having vertices representing the plurality of nodes and edges indicating a direction of data flow among the plurality of nodes in the radio network.
Yet, Smith in view of Kottkamp and further in view of Dong discloses The method of claim 1, wherein the digital twin representation comprises a directed graph having vertices representing the plurality of nodes and edges indicating a direction of data flow among the plurality of nodes in the radio network. since Dong explains an IP radio access network represented as a directed graph, in which links connect adjacent network nodes and indicate the data traffic flow directions among those nodes, whereby the representation comprises a directed graph having vertices representing the plurality of nodes and edges indicating a direction of data flow among the plurality of nodes in the radio network (Dong, para [0046], “converting an undirected graph representation of an IP RAN network to a directed graph representation . . . A directed graph representation may comprise a plurality of CSG links that connect a plurality of adjacent network nodes and the CSG links may indicate the data traffic flow directions among the network node”).
Consequently, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to form the digital twin representation of Smith, as combined with Kottkamp, as a directed graph whose vertices are the network nodes and whose edges indicate the direction of data traffic flow among them, as Dong represents its radio access network, in order to give the twin the topology and traffic flow directions of the network it models. Smith’s digital twin model represents the physical nodes of the network together with relations among them, and Dong teaches that a centralized controller with a global view of a radio access network obtains its topology information and works on a directed graph representation whose links indicate the data traffic flow directions among the network nodes. Recording each relation of the twin as a directed edge is the use of a known technique to improve a similar system in the same way, and it would have predictably yielded a twin that shows which way data flows between its nodes. One of ordinary skill would have had a reasonable expectation of success, since a directed graph is a standard way to represent a network topology and adds only a direction to each relation the twin already holds.
Regarding claim 25, which depends on claim 11, (New) even though Smith in view of Kottkamp teaches the digital twin representation of the radio network, with virtual representations of its nodes and of the relations among them, a subset of which a trained machine learning model identifies and monitors: (Smith, para [0182], Kottkamp, para [0013]), Smith in view of Kottkamp does not expressly teach that the digital twin representation comprises a directed graph having vertices representing the plurality of nodes and edges indicating a direction of data flow among the plurality of nodes in the radio network.
However, Smith in view of Kottkamp and further in view of Dong discloses The one or more non-transitory computer storage media of claim 11, wherein the digital twin representation comprises a directed graph having vertices representing the plurality of nodes and edges indicating a direction of data flow among the plurality of nodes in the radio network. given that Dong sets out an IP radio access network represented as a directed graph, in which links connect adjacent network nodes and indicate the data traffic flow directions among those nodes, so that the representation comprises a directed graph having vertices representing the plurality of nodes and edges indicating a direction of data flow among the plurality of nodes in the radio network (Dong, para [0046], “converting an undirected graph representation of an IP RAN network to a directed graph representation . . . A directed graph representation may comprise a plurality of CSG links that connect a plurality of adjacent network nodes and the CSG links may indicate the data traffic flow directions among the network node”).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to form the digital twin representation of Smith, as combined with Kottkamp, as a directed graph whose vertices are the network nodes and whose edges indicate the direction of data traffic flow among them, as Dong represents its radio access network, in order to give the twin the topology and traffic flow directions of the network it models. Smith’s digital twin model represents the physical nodes of the network together with relations among them, and Dong teaches that a centralized controller with a global view of a radio access network obtains its topology information and works on a directed graph representation whose links indicate the data traffic flow directions among the network nodes. Recording each relation of the twin as a directed edge is the use of a known technique to improve a similar system in the same way, and it would have predictably yielded a twin that shows which way data flows between its nodes. One of ordinary skill would have had a reasonable expectation of success, since a directed graph is a standard way to represent a network topology and adds only a direction to each relation the twin already holds.
Regarding claim 27, which depends on claim 19, (New) the claim recites: The system of claim 19, wherein the digital twin representation comprises a directed graph having vertices representing the plurality of nodes and edges indicating a direction of data flow among the plurality of nodes in the radio network. Claim 27 is analogous to claim 25 and is rejected for the same reasons.
Claims 24, 26 and 28 are rejected under 35 U.S.C. 103 as being unpatentable over Smith (US 2022/0191648 A1) in view of Kottkamp (US 2021/0136603 A1) and further in view of Abusch-Magder (US 2007/0225023 A1). (The Examiner notes that Abusch-Magder qualifies as prior art under 35 U.S.C. 102(a)(1) because it was published on September 27, 2007, before the effective filing date of the claimed invention.
Regarding claim 24, which depends on claim 1, (New) even though Smith in view of Kottkamp teaches the trained machine learning model identifying a first subset of the digital twin representation for a first time range and monitoring its parameters: (Smith, para [0117], Kottkamp, para [0015]), Smith in view of Kottkamp does not explicitly disclose that the first subset enables monitoring of the operating conditions with lower computational resources and latency compared to processing the digital twin representation in its entirety while reliably representing operating conditions of the radio network.
Conversely, Smith in view of Kottkamp and further in view of Abusch-Magder discloses The method of claim 1, wherein the first subset enables monitoring of the operating conditions with lower computational resources in that Abusch-Magder shows that a larger part of a wireless network costs more to monitor, in monitoring equipment and in computational resources such as storage and processing servers, than a smaller part, with the result that monitoring a smaller subset enables monitoring of the operating conditions with lower computational resources (Abusch-Magder, para [0066], “larger neighborhoods cost more to monitor (due to monitoring equipment, computational resources such as storage and processing servers, increase in communication on backhaul etc.), while smaller neighborhoods might be less accurate”).
Furthermore, Abusch-Magder discloses and latency compared to processing the digital twin representation in its entirety given that Abusch-Magder sets out a neighborhood of about 37 cells giving results comparable to evaluating the full 206 cell network while running roughly fifteen times faster than the full evaluation, so that the smaller subset is processed with lower latency compared to processing the digital twin representation in its entirety (Abusch-Magder, para [0055], “results comparable to full-evaluation greedy on a realistic 206-cell network were obtained using an average neighborhood size of 36.7. This simplification provided an algorithm runtime that was roughly 15 times faster than the conventional full-evaluation method”).
In addition, Abusch-Magder discloses while reliably representing operating conditions of the radio network. because Abusch-Magder teaches that the selected neighborhood whose parameters are monitored is representative of the entire network, which means that the monitored subset reliably represents operating conditions of the radio network (Abusch-Magder, para [0049], “For the selected neighborhood (at 520), which is representative of the entire network, one or more given parameters thereof may be monitored (530) in an effort to evaluate network performance or determine current status of the network”).
For these reasons, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have the trained model of Smith and Kottkamp identify, as the first subset, a part of the network that is representative of the whole network and to monitor that part instead of the whole, as Abusch-Magder monitors a selected neighborhood of cells, in order to reduce the computational resources and time that monitoring the entire network requires. Smith’s digital twin may model the entire physical world or a subset of its objects, and Abusch-Magder teaches that monitoring cell specific data over a broad network area carries a high overhead cost in data collection and processing, that a smaller neighborhood representative of the entire network may be monitored instead, and that doing so ran roughly fifteen times faster than a full evaluation. Applying that known technique to the digital twin of Smith and Kottkamp would have predictably yielded monitoring of the operating conditions with less computation and delay while the monitored subset still represents the network. One of ordinary skill would have had a reasonable expectation of success, since choosing which part of the representation to monitor changes only the input of the monitoring that the combination already performs.
Regarding claim 26, which depends on claim 11, (New) even though Smith in view of Kottkamp teaches the trained machine learning model identifying a first subset of the digital twin representation for a first time range and monitoring its parameters: (Smith, para [0117], Kottkamp, para [0015]), Smith in view of Kottkamp stops short of disclosing that the first subset enables monitoring of the operating conditions with lower computational resources and latency compared to processing the digital twin representation in its entirety while reliably representing operating conditions of the radio network.
Nevertheless, Smith in view of Kottkamp and further in view of Abusch-Magder discloses The one or more non-transitory computer storage media of claim 11, wherein the first subset enables monitoring of the operating conditions with lower computational resources because Abusch-Magder teaches that a larger part of a wireless network costs more to monitor, in monitoring equipment and in computational resources such as storage and processing servers, than a smaller part, which means that monitoring a smaller subset enables monitoring of the operating conditions with lower computational resources (Abusch-Magder, para [0066], “larger neighborhoods cost more to monitor (due to monitoring equipment, computational resources such as storage and processing servers, increase in communication on backhaul etc.), while smaller neighborhoods might be less accurate”).
Furthermore, Abusch-Magder discloses and latency compared to processing the digital twin representation in its entirety as Abusch-Magder describes a neighborhood of about 37 cells giving results comparable to evaluating the full 206 cell network while running roughly fifteen times faster than the full evaluation, such that the smaller subset is processed with lower latency compared to processing the digital twin representation in its entirety (Abusch-Magder, para [0055], “results comparable to full-evaluation greedy on a realistic 206-cell network were obtained using an average neighborhood size of 36.7. This simplification provided an algorithm runtime that was roughly 15 times faster than the conventional full-evaluation method”).
Moreover, Abusch-Magder discloses while reliably representing operating conditions of the radio network. since Abusch-Magder explains that the selected neighborhood whose parameters are monitored is representative of the entire network, whereby the monitored subset reliably represents operating conditions of the radio network (Abusch-Magder, para [0049], “For the selected neighborhood (at 520), which is representative of the entire network, one or more given parameters thereof may be monitored (530) in an effort to evaluate network performance or determine current status of the network”).
Accordingly, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have the trained model of Smith and Kottkamp identify, as the first subset, a part of the network that is representative of the whole network and to monitor that part instead of the whole, as Abusch-Magder monitors a selected neighborhood of cells, in order to reduce the computational resources and time that monitoring the entire network requires. Smith’s digital twin may model the entire physical world or a subset of its objects, and Abusch-Magder teaches that monitoring cell specific data over a broad network area carries a high overhead cost in data collection and processing, that a smaller neighborhood representative of the entire network may be monitored instead, and that doing so ran roughly fifteen times faster than a full evaluation. Applying that known technique to the digital twin of Smith and Kottkamp would have predictably yielded monitoring of the operating conditions with less computation and delay while the monitored subset still represents the network. One of ordinary skill would have had a reasonable expectation of success, since choosing which part of the representation to monitor changes only the input of the monitoring that the combination already performs.
Regarding claim 28, which depends on claim 19, (New) the claim recites: The system of claim 19, wherein the first subset enables monitoring of the operating conditions with lower computational resources and latency compared to processing the digital twin representation in its entirety while reliably representing operating conditions of the radio network. Claim 28 is analogous to claim 26 and is rejected for the same reasons.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to CHONGSUH (John) PARK whose telephone number is 408-918-7574. The examiner can normally be reached Monday - Friday 8:00-5:30 PST
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, Avellino, Joseph can be reached at 571-272-3905 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.
/CHONGSUH PARK/Examiner, Art Unit 2478
/JOSEPH E AVELLINO/Supervisory Patent Examiner, Art Unit 2478
