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 .
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 02/25/2026 has been considered by examiner and made of record in the application file.
Response to Arguments
Applicant’s arguments with respect to claim(s) 41, 45-46, 48-49, 51, 56 and 58-60 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
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.
Claim(s) 41, 49, 51, 59, and 60 are rejected under 35 U.S.C. 103 as being unpatentable over VULGARAKIS FELJAN (US-20220417109-A1) in view of Curic (US-20240098568-A1) in further view of Tapia (US-20170353991-A1).
Regarding Claim 41, VULGARAKIS FELJAN discloses a method comprising:
building, based on telemetry data from an open radio access network, a library of machine learning models that match performance indicator degradation signatures to respective known configuration solutions that predictably improve corresponding instances of performance degradation within the open radio access network (paragraph [0029], "In various embodiments, at least one ML model may be adapted and data for a target network converted based on the semantic mapping of the data. Applicable ML models may be selected using a complex set of criteria derived for each situation and may solve problems when the equipment between different vendors produces data that does not look alike." and paragraph [0031], "Presently disclosed embodiments may provide potential advantages. One potential advantage may provide reuse of existing ML models to predict KPIs, outages, monitor SLA, etc." and paragraph [0036], Fig.1, "The network inventory model may include a matching of the equipment (e.g., base station 116a and diesel energy source 118a) in each network (e.g., network 100a) to a ML model that has been trained for that equipment. For example, if a ML model has been trained on a network node trained for three uplinks, the ML model is not matched to another network node that provides only two uplinks (and therefore should not be used the other network node). In a further example, if a ML model has been trained on network equipment with a diesel power supply, the ML model is not matched to network equipment that is powered on solar supply (and therefore should not be used for network equipment that is powered with a solar supply)." (i.e., Open radio access network will be mapped by a different prior art. “Applicable ML model” is reading on a library of model. The machine learning models are selected to solve problems.));
matching, by a radio access network intelligent controller, the telemetry data from the open radio access network to a specific machine learning model from the library of machine learning models to apply to the telemetry data (paragraph [0049], Fig.2, "At operations 222, selector and adaptor node 104 requests from ML model database 108 an identification of ML models that match a filter based on the requested task (e.g., desired high-level KPI such as KPI degradation)." and paragraph [0050], "At operations 226 and 228, selector and adaptor node 104 iterates through the outputs of each ML model in the filtered identification of ML models to select a ML model(s) that is a match for the requested task (e.g., KPI degradation). For each ML model in the filtered identification, at operation 226, selector and adaptor node 104 selects inputs that apply to target network 100a. At operation 228, selector and adaptor node 104 identifies the ML models from the filtered identification of ML models that include inputs that apply to target network 100a (e.g., matched models)."(i.e., a radio access network intelligent controller will be mapped by a different reference. There is a matching between the specific data and a machine learning to solve the KPI degradation as the machine learning was trained on solving the issues.));
performing, by applying the specific machine learning model (paragraph [0057], "FIG. 3 illustrates an exemplary application 300 of a deployed ML model to perform a task. At operation 302, actor device 102 requests that network control node 106 perform a task using the deployed model (e.g., the deployed model from operation 246). Responsive to the request, at operation 304, network control node 106 makes a read request from network database 114 for data or counters of network 100a needed as input(s) to the deployed ML model." (i.e., the in response to detecting… will be clearer with a different reference. Fig.3 is to show a ML performing task to improve performance the ML that was picked in Fig.2.));
wherein:
the radio access network intelligent controller performs, by applying the specific machine learning model (paragraph [0050], Fig.2, "At operations 226 and 228, selector and adaptor node 104 iterates through the outputs of each ML model in the filtered identification of ML models to select a ML model(s) that is a match for the requested task (e.g., KPI degradation). For each ML model in the filtered identification, at operation 226, selector and adaptor node 104 selects inputs that apply to target network 100a. At operation 228, selector and adaptor node 104 identifies the ML models from the filtered identification of ML models that include inputs that apply to target network 100a (e.g., matched models)." and paragraph [0050], "At operation 230, if selector and adaptor node 104 finds an exact match, selector and adaptor node 104 communicates a request to network control node 106 to deploy the ML model that is an exact match. Responsive to the request, at operation 232, network control node 106, deploys the ML model that is an exact match." (i.e., examiner notes the in response to detecting… will be clearer by a different prior art, this citation is to show a selection of the specific model is being performed wherein perform tasks. Fig.2 shows a specific machine learning model is being applied.));
the specific performance indicator comprises a key performance indicator (paragraph [0031], "Presently disclosed embodiments may provide potential advantages. One potential advantage may provide reuse of existing ML models to predict KPIs, outages, monitor SLA, etc." and paragraph [0050], Fig.2, "At operations 226 and 228, selector and adaptor node 104 iterates through the outputs of each ML model in the filtered identification of ML models to select a ML model(s) that is a match for the requested task (e.g., KPI degradation). (i.e., discloses KPI.));
However, VULGARAKIS FELJAN does not explicitly disclose based on telemetry data from an open radio access network; a radio access network intelligent controller; detecting that a specific performance indicator of the open radio access network has degraded; and performing, by applying the specific machine learning model in response to detecting that the specific performance indicator of the open radio access network has degraded, a corresponding solution indicated by the specific machine learning model such that the specific performance indicator is improved; wherein: the radio access network intelligent controller performs, by applying the specific machine learning model in response to detecting that the specific performance indicator of the open radio access network has degraded, the corresponding solution indicated by the specific machine learning model such that the specific performance indicator is improved; the library of machine learning models comprises a library of classifiers that classify telemetry data as matching one or more of the performance indicator degradation signatures to predict the respective known configuration solutions; and the specific machine learning model is built at least in part by: applying a configuration parameter change for a predetermined period of time; monitoring the specific performance indicator before and after the configuration parameter change; comparing a first value of the key performance indicator prior to the configuration parameter change with a second value of the key performance indicator after the configuration parameter change such that an improvement to the key performance indicator is verified; and building the machine learning model such that the machine learning model indicates the configuration parameter change as part of the corresponding solution based on verifying that the configuration parameter change improved the key performance indicator during the predetermined period of time.
Curic discloses based on telemetry data from an open radio access network (paragraph [0045], "…data collected from the near-RT RIC, O-DU 120, and O-RU 116. The near-RT RIC 114 is a logical function that enables near-real-time (sub 1 second operation times) control and optimization of RAN elements and resources via fine-grained, data collection and actions over the E2 interface." and paragraph [0062], "Such improvements include detecting and mitigating conflicts, such as xApp conflicts in near-RT RICs 104 in the O-RAN architecture…Further, embodiments of the present invention provide a practical application in the field of computing technology, particularly O-RAN, by establishing awareness in non-RT RIC 112 of xApp activities across all near-RT RICs 114 associated with that non-RT RIC 112." and paragraph [0068], "The conflict detector 304 periodically or on-demand retrieves logged data from)(App Activity Log available on the near-RT RICs 114 to detect occurrence (or not) of a conflict…" and paragraph [0076], "The xApps 132 only log activities, i.e., network updates, which were approved by the conflict mitigator 402 in one or more embodiments of the present invention." and paragraph [0089], "This data includes both network and user key performance indicators (KPIs) and are all tagged with corresponding timestamps when the data is generated.");
a radio access network intelligent controller (paragraph [0068], "The AI/ML model(s) 312 are pre-trained using a training dataset before being deployed on the non-RT RIC 112 in one or more embodiments of the present invention. The AI/ML model(s) 312 can be continuously updated as the non-RT RIC 112 is used.");
detecting that a specific performance indicator of the open radio access network has degraded (paragraph [0056], "There are various types of xApps 132. For example, a first type of xApps 132 can subscribe to rApps 130 and implement the policies that the rApps 130 issue over the interface A1. Second type of xApps 132 can execute independently from rApps 130 and govern the network behavior according to their own logic. xApps 132 execute in parallel of each other and conflicts can occur. Such conflicts can cause network instability or performance degradation. Conflicts also cause a security risk in O-RAN 100 because attackers may use such a conflict as a vulnerability to attack the network."); and
performing, by applying the specific machine learning model in response to detecting that the specific performance indicator of the open radio access network has degraded, a corresponding solution indicated by the specific machine learning model such that the specific performance indicator is improved (paragraph [0072], "In one or more embodiments of the present invention, the analytics module 308 correlates the data from the one or more sources. Based on the correlated result the analytics module 308 creates one or more conflict-avoidance policies that are sent further over A1 interface to the near-RT RICs 114. In one or more embodiments of the present invention, the non-RT RIC 112 creates a respective set of policies for each near-RT RIC 114." and paragraph [0107], "Further yet, an Analytics Module is used for generating policy for each near-RT RIC and/or each xApp on each near-RT RIC. The policy is sent and applied to the respective near-RT RIC to facilitate detection, resolution, and mitigation of direct, indirect, and implicit conflicts intra- and inter-domain in programmable networks with hierarchically organized operation plane." (i.e., VULGARAKIS FELJAN discloses the library of machine learning models and this limitation is reading that each of the machine learning in the library comprises classifiers that can match telemetry data to known degradation in order to obtain respective solution as disclosed in Curic, wherein "policy" are the known configuration solutions.));
wherein:
the radio access network intelligent controller performs, by applying the specific machine learning model in response to detecting that the specific performance indicator of the open radio access network has degraded, the corresponding solution indicated by the specific machine learning model such that the specific performance indicator is improved (paragraph [0068], "The AI/ML model(s) 312 can be continuously updated as the non-RT RIC 112 is used. In one or more embodiments of the present invention, the trained AI/ML model(s) 312 classify the conflict-related patterns into different conflict types, i.e., direct, indirect, and implicit. By performing the conflict detection in the non-RT RIC 112, embodiments of the present invention can facilitate detecting inter-domain conflicts, because data/information required to make such a distinction is only available at the non-RT RIC 112.");
the library of machine learning models comprises a library of classifiers that classify telemetry data as matching one or more of the performance indicator degradation signatures to predict the respective known configuration solutions (paragraph [0068], "The AI/ML model(s) 312 can be continuously updated as the non-RT RIC 112 is used. In one or more embodiments of the present invention, the trained AI/ML model(s) 312 classify the conflict-related patterns into different conflict types, i.e., direct, indirect, and implicit. By performing the conflict detection in the non-RT RIC 112, embodiments of the present invention can facilitate detecting inter-domain conflicts, because data/information required to make such a distinction is only available at the non-RT RIC 112." and paragraph [0072], "In one or more embodiments of the present invention, the analytics module 308 correlates the data from the one or more sources. Based on the correlated result the analytics module 308 creates one or more conflict-avoidance policies that are sent further over A1 interface to the near-RT RICs 114. In one or more embodiments of the present invention, the non-RT RIC 112 creates a respective set of policies for each near-RT RIC 114." and paragraph [0107], "Further yet, an Analytics Module is used for generating policy for each near-RT RIC and/or each xApp on each near-RT RIC. The policy is sent and applied to the respective near-RT RIC to facilitate detection, resolution, and mitigation of direct, indirect, and implicit conflicts intra- and inter-domain in programmable networks with hierarchically organized operation plane." (i.e., VULGARAKIS FELJAN discloses the library of machine learning models and this limitation is reading that each of the machine learning in the library comprises classifiers that can match telemetry data to known degradation in order to obtain respective solution as disclosed in Curic, wherein "policy" are the known configuration solutions.)); and
the specific machine learning model is built at least in part by:
monitoring the specific performance indicator before and after the configuration parameter change (paragraph [0059], Fig.1, "in which the near-RT RIC 104 checks the parameters that certain xApp 132 is attempting to modify before the update is implemented in the network. In some cases, post-action verification is performed," and paragraph [0068], " The AI/ML model(s) 312 can be continuously updated as the non-RT RIC 112 is used." (i.e., Curic discloses of monitoring before and after applying the modification));
comparing a first value (paragraph [0059], “parameter” and paragraph [0101], “snapshot…KPI…before”) of the key performance indicator prior to the configuration parameter change with a second value (paragraph [0059], “parameter” and paragraph [0101], “snapshot…KPI…after”) of the key performance indicator after the configuration parameter change such that an improvement to the key performance indicator is verified (paragraph [0059], Fig.1, "in which the near-RT RIC 104 checks the parameters that certain xApp 132 is attempting to modify before the update is implemented in the network. In some cases, post-action verification is performed," and paragraph [0068], " The AI/ML model(s) 312 can be continuously updated as the non-RT RIC 112 is used." and paragraph [0101], “For example, the near-RT RIC 114 creates snapshots of the KPIs of an xApp 132 before and after that xApp 132 has made certain updates. Thus, the delta of the impact can be analyzed and assessed if it is meeting the operation targets. Here, a “snapshot of KPIs” can include a digital collection of the parameter values that are considered as KPIs of the xApp 132. The snapshot can be a data structure of file stored electronically.” and paragraph [0102], “Analytics Module 308 can also revert the policy if it identifies that it did not result with the desired network behavior. The policy revert activity is also logged for future reference.” (i.e., a comparison is being performed being two key performance indicator in order to see the improvement, and will revert the changes of the improvement is not verified.)); and
building the machine learning model such that the machine learning model indicates the configuration parameter change as part of the corresponding solution based on verifying that the configuration parameter change improved the key performance indicator during the predetermined period of time (paragraph [0045], "In some embodiments of the present invention, the non-RT RIC 112 can be an ML training host to host the training of one or more ML models. ML training can be performed offline using data collected from the near-RT RIC, O-DU 120, and O-RU 116. The near-RT RIC 114 is a logical function that enables near-real-time (sub 1 second operation times) control and optimization of RAN elements and resources via fine-grained, data collection and actions over the E2 interface." and paragraph [0059], " A technical challenge exists to detect and mitigate such conflicts in the O-RAN 100. Typically, direct and (some) indirect conflicts can be detected by leveraging pre-action resolution, in which the near-RT RIC 104 checks the parameters that certain xApp 132 is attempting to modify before the update is implemented in the network." and paragraph [0101], “For example, the near-RT RIC 114 creates snapshots of the KPIs of an xApp 132 before and after that xApp 132 has made certain updates. Thus, the delta of the impact can be analyzed and assessed if it is meeting the operation targets. Here, a “snapshot of KPIs” can include a digital collection of the parameter values that are considered as KPIs of the xApp 132. The snapshot can be a data structure of file stored electronically.” and paragraph [0102], “Analytics Module 308 can also revert the policy if it identifies that it did not result with the desired network behavior. The policy revert activity is also logged for future reference.” (i.e., The non-RT RIC builds ML models that checks for conflict and implement updates to resolve the conflicts. There is also some sort of predetermined period of time as Curic discloses of reverting the changes when no improvement is seen as disclosed in par.101-102, but during the predetermined period of time will be clearer by a different reference.)).
VULGARAKIS FELJAN and Curic are considered to be analogous to the claimed invention because they are in the same Supervisory, monitoring or testing arrangements. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention to have modified VULGARAKIS FELJAN to implemented the method of Curic because Curic discloses an improvement to O-RAN, such as facilitating automated conflict detection (Curic, paragraph [0031], “One or more embodiments of the present invention facilitate improvements to radio access networks (RANs), particularly open-RAN (O-RAN) networks. Embodiments of the present invention provide technical solutions that facilitate automated conflict detection. One or more embodiments of the present invention detect indirect and implicit conflicts among operations of a near-RT RICs, regardless of whether the conflicts occur on the same or different (neighboring) Near-RT RICs. Further, embodiments of the present invention detect and mitigate any type of conflict that occur inter- or intra-domain (as defined herein).”).
However, VULGARAKIS FELJAN in view of Curic do not disclose applying a configuration parameter change for a predetermined period of time.
Tapia discloses applying a configuration parameter change for a predetermined period of time (paragraph [0032], "the network fix application 118 may leverage one or more trained machine learning model via the artificial intelligence module 122 to analyze the user device performance data and the network performance data within the network to determine likely root causes for a quality of service issues for subscribers and to determine the most optimal order of providing network fix to address the root causes…In another example, a network fix prioritization is optimal if the order in which network fixes are made according to the network fix prioritization can increase the network performance level within a predetermined period of time." and paragraph [0076], “At the decision block 409, the feedback module 218 of the action tracker 121 analyzes if the expected performance improvement was achieved, wherein the expected performance improvement is achieved when the network performance level is equal to or greater than a predetermined threshold…” (i.e., Curic discloses of applying the changes and monitoring performance after the changes see Curic par.101, Tapia discloses of a predetermined period of time.));
building the machine learning model such that the machine learning model indicates the configuration parameter change as part of the corresponding solution based on verifying that the configuration parameter change improved the key performance indicator during the predetermined period of time (paragraph [0032], "the network fix application 118 may leverage one or more trained machine learning model via the artificial intelligence module 122 to analyze the user device performance data and the network performance data within the network to determine likely root causes for a quality of service issues for subscribers and to determine the most optimal order of providing network fix to address the root causes…In another example, a network fix prioritization is optimal if the order in which network fixes are made according to the network fix prioritization can increase the network performance level within a predetermined period of time." and paragraph [0076], “At the decision block 409, the feedback module 218 of the action tracker 121 analyzes if the expected performance improvement was achieved, wherein the expected performance improvement is achieved when the network performance level is equal to or greater than a predetermined threshold…” (i.e., Curic discloses of building the ML model to indicate the configuration parameter change as a solution see office action above, and Tapia discloses a predetermined time period, and will apply a different solution as disclosed in paragraph 76.)).
VULGARAKIS FELJAN in view of Curic and Tapia are considered to be analogous to the claimed invention because they are in the same Supervisory, monitoring or testing arrangements. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention to have modified VULGARAKIS FELJAN to implemented the method of Tapia of implementing a solution for a predetermined period of time as Tapia discloses of implementing another solution when one solution does not improve the performance in order to maximize reliability and efficient way of solving a network issue by implementing different solutions and Tapia further discloses of increase in accuracy of finding a root cause in order to implement a solution (Tapia, paragraph [0032], “a network fix prioritization is optimal if the order in which network fixes are made according to the network fix prioritization can increase the network performance level within a predetermined period of time.” and paragraph [0033], “Further, the machine learning model may incorporate a combination of multiple machine learning algorithms, such as a Bayesian algorithm, a decision tree algorithm, an SVM algorithm, to increase the accuracy of the root cause predictions and issue prioritization generated by the module 122.”).
Regarding Claim 49, VULGARAKIS FELJAN in view of Curic in further view Tapia discloses the limitations of claim 41.
Curic further discloses wherein the telemetry data is continuously streamed from the open radio access network to a centralized data platform (paragraph [0045], "ML training can be performed offline using data collected from the near-RT RIC, O-DU 120, and O-RU 116. The near-RT RIC 114 is a logical function that enables near-real-time (sub 1 second operation times) control and optimization of RAN elements and resources via fine-grained, data collection and actions over the E2 interface." and paragraph [0053], “In O-RAN 100, SMO 102, non-RT RIC 102, and near-RT RICs 104 continuously collect the network state.” and paragraph [0089], “The matrix 310 is used to keep track about the performance and fault data from the controlled E2 nodes 202. This data includes both network and user key performance indicators (KPIs) and are all tagged with corresponding timestamps when the data is generated.” (i.e., Examiner pointes to Fig.3 of NF being CU, DU, or RU and data is collected to a radio analysis system 300.)).
Tapia discloses a centralized data platform (paragraph [0013], "a network fix application can continuously monitor performance data related to…network components in a wireless carrier network…" and paragraph [0016], Fig.1 "Further, in a networked deployment, new computing nodes 126 may be added on the fly without affecting the operational integrity of the data adaptor platform 116," (i.e., Examiner pointes to Fig.1 where the telemetry data of radio access network such as Fig.1:111 is streamed to the Fig.1:126 and all the operation of providing a solution happens at the computing nodes using Fig.4.)).
The proposed combination as well as the motivations for combining the references presented in the rejection of the parent claim apply to this claim and are incorporated herein by reference.
Regarding Claim 51, which is similar in scope to claim 41, thus rejected under the same rationale.
Examiner notes VULGARAKIS FELJAN discloses a physical computing processor; and a non-transitory computer-readable medium (paragraph [0007], “a computer program product can be provided that includes a non-transitory computer readable medium storing instruction that, when executed on at least one processor, cause the at least one processor to carry out methods performed by the first network node.”).
Regarding Claim 59, which is similar in scope to claim 49, thus rejected under the same rationale.
Regarding Claim 60, which is similar in scope to claim 41, thus rejected under the same rationale.
Examiner notes VULGARAKIS FELJAN discloses a non-transitory computer-readable medium (paragraph [0007], “a computer program product can be provided that includes a non-transitory computer readable medium storing instructions that, when executed on at least one processor, cause the at least one processor to carry out methods performed by the first network node.”).
Claim(s) 45 is rejected under 35 U.S.C. 103 as being unpatentable over VULGARAKIS FELJAN (US-20220417109-A1) in view of Curic (US-20240098568-A1) in view of Tapia (US-20170353991-A1) in further view of MEDITHE (US-20230091638-A1).
Regarding Claim 45, VULGARAKIS FELJAN in view of Curic in further view Tapia discloses the limitations of claim 41.
Tapia further discloses wherein: the machine learning model is built in response to detecting that the specific performance indicator of the open radio access network has degraded at (paragraph [0041], “The network fix application 118 may process real time or non-real time data from various geographic locations, in which data from multiple data sources may be aggregated, converged, or otherwise consolidated.” and paragraph [0049], “The artificial intelligence module 122 comprises at least one machine learning training module for issue prioritization and predicting root cause 219, at least one machine learning logic 221, and one or more machine learning trained models 220A, 220N. The model training module 219 may train machine learning models 220A, 220N to analyze the performance data from the data sources 110-114 to determine root causes for the quality of service issues for subscribers and to prioritize network fix for each problem related to the root causes.” and paragraph [0050], "In the initial training data input phase of the machine learning training pipeline, it is contemplated that the machine learning training module 219 may receive a training corpus comprised of one or more input datasets from the data adaptor platform 116. The training corpus may include training data that emulates data collected from the multiple data sources 110-114 and optionally a set of desired outputs for the training data. For example, the data that is received during the training data input phase may include Operating Support System (OSS) counters, KPIs, network coverage details…" (i.e., the machine learning is built on data from adaptor platform 116 wherein includes RAN data OSS counters as disclosed in par.21, and is built in order to detect KPIs issues or degradation.));
detecting that the specific performance indicator of the open radio access network has degraded is subsequently performed also at a second and distinct site (paragraph [0065], Fig.5:503, " At block 501, the issue investigation module 210 conducts issue investigation to identify symptoms regarding service quality in poor performing areas…At block 502, the issue investigation module 210 correlates one or more complex symptoms to a service quality problem by matching the symptoms with previously stored performance pattern or symptoms that corresponds to a particular type of service quality issue and/or root cause. Accordingly, at block 503, the issue investigation module 210 identifies specific long-term and/or short-term problems in poor performing areas. At block 504, the root cause analysis module predicts a root cause for the identified wireless carrier network issues based on a set of live input data using the machine learning model." (i.e., Examiner reading "previously stored performance pattern" as the performance pattern on the first site and "poor performing areas" as the second and distinct site. Because there is a plurality of “poor performing areas” that is reading as distinct areas but with similar or matching symptoms e.g., the specific performance indicator.)).
Curic further discloses and the radio access network intelligent controller performs, by applying the machine learning model in response to detecting that the specific performance indicator of the open radio access network has degraded at the second and distinct site, the corresponding solution indicated by the machine learning model such that the specific performance indicator is improved (paragraph [0045], Fig.1, "The non-RT RIC 112 is a logical function within the SMO framework 102 that enables non-real-time (>1 second operation times) control and optimization of RAN elements and resources; AI/machine learning (ML) workflow(s) including model training, inferences, and updates; and policy-based guidance of applications/features in the near-RT RIC 114." and paragraph [0093], "In general, the xApp actions at any time must be such that they maintain the satisfactory KPIs or improve the degraded KPIs, while keeping intent into consideration. The analytics module 308 is used for this purpose. The analytics module 308 is aware of the optimization goals of each xApp 132 and parameters that each xApp 132 can affect. Only then the analytics module 308 can tune the xApp activity according to the intent requirements." (i.e., Similar to Claim 41, the near-RT RIC performs the changes of xApp as indicated by the non-RT RIC.)).
However, VULGARAKIS FELJAN in view of Curic in further view Tapia do not explicitly disclose a first site.
MEDITHE discloses First site (par.24, Fig.1, cell) (paragraph [0024], "At operation 205, data of a plurality of features of a plurality of cells are collected over a period of time…the collected data correspond to the training data 142 described with respect to FIG. 1." and paragraph [0025], "Examples of cell features for which data are collected include key performance indicators (KPIs)." and paragraph [0035], "At operation 235, one or more trained machine learning models obtained by one or more machine learning techniques performed at operation 225 are saved…the one or more trained machine learning models are provided to a cell anomaly detecting device for use in cell anomaly detection, as described with respect to FIG. 1 where one or more machine learning models 144 are provided to the cell anomaly detecting device 150." (i.e., Fig.1 shows plurality of cells and the data such as KPI are collected to train machine learning to detect anomalies or degradation.)).
VULGARAKIS FELJAN in view of Curic in further view Tapia and MEDITHE are considered to be analogous to the claimed invention because they are in the same field wireless communication. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention to have modified VULGARAKIS FELJAN to implement the method of MEDITH of collecting data from a first cite as MEDITH enables to collect data from a plurality of cells by clustering and reducing the amount of processing and time required for cell anomaly detection (MEDITH, paragraph [0026], “when a group of two or more KPIs are related to each other most of the times, there is a high probability that if one KPI in the group shows that the cell is anomalous, the other KPI(s) in the group will also show that the cell is anomalous. In such situations, it is sufficient to consider one KPI in the group and ignore the other KPI(s) for cell anomaly detection, thereby reducing the amount of processing and time required for cell anomaly detection… As a result, the number of KPIs and the corresponding of KPI data to be considered for cell anomaly detection are significantly reduced, in one or more embodiments.”).
Claim(s) 46, 48, 56, and 58 are rejected under 35 U.S.C. 103 as being unpatentable over VULGARAKIS FELJAN (US-20220417109-A1) in view of Curic (US-20240098568-A1) in view of Tapia (US-20170353991-A1) in further view of Arora (US-20240022923-A1).
Regarding Claim 46, VULGARAKIS FELJAN in view of Curic in further view Tapia discloses the limitations of claim 41.
Curic further discloses wherein the corresponding solution indicated by the machine learning model is performed as part of is autonomously applied by the open radio access network in repeated iterations (paragraph [0092], Fig.10, “FIG. 10 depicts an example impact matrix 1000 that establishes an awareness in the non-RT RIC 112 about the xApp activities in all of the associated near-RT RICs 114. Correlation of the global xApp activity log 800 and performance and fault data matrix 310 facilitates obtaining the information about the network response that followed each action performed by the xApps 132…” and paragraph [0094], “the analytics module 308 uses reinforcement learning to determine the policies (224).” and paragraph [101], “the near-RT RIC 114 creates snapshots of the KPIs of an xApp 132 before and after that xApp 132 has made certain updates. Thus, the delta of the impact can be analyzed and assessed if it is meeting the operation targets.” and paragraph [0102], “Analytics Module 308 can also revert the policy if it identifies that it did not result with the desired network behavior. The policy revert activity is also logged for future reference.” (i.e., Fig.10 and par.92-103 discloses continuously monitor xApp activity and using machine learning that is indicated by the analytics module to improve the degraded KPIs, and examiner reading as a closed-loop since par.92-103 is implying its continuously updating the parameter of xApp behavior.)).
However, VULGARAKIS FELJAN in view of Curic in further view Tapia do not disclose wherein the corresponding solution indicated by the machine learning model is performed as part of a closed-loop radio access network optimization loop in which the corresponding solution is autonomously applied by the open radio access network in repeated iterations.
Arora discloses wherein the corresponding solution indicated by the machine learning model is performed as part of a closed-loop radio access network optimization loop in which the corresponding solution is autonomously applied by the open radio access network in repeated iterations (paragraph [0066], Fig.5, "At operation 504, the process flow 500 can comprise enforcing a configuration that can account for and/or resolve a mismatch, misconfiguration and/or the like…executing the configuration enforcement…in the case of an inconsistency (e.g., mismatch and/or misconfiguration) among the NFs…a configuration can be retrieved from the SMO 360 and transmitted to the affected NF, such as by the controller component 312." and paragraph [0067], “At operation 505, the process flow 500 can comprise analysis of the knowledge base 328 and/or analytical model 330 for updates and/or unknown or unspecified context. Likewise, operation 505 can comprise updating of the knowledge base 328 and/or training of the analytical model 330 where a new context is determined.” (i.e., Arora Fig.5 discloses monitoring and configuring network functions (NFs) of open ran.)).
VULGARAKIS FELJAN in view of Curic in further view Tapia and Arora are considered to be analogous to the claimed invention because they are in the same Supervisory, monitoring or testing arrangements. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention to have modified VULGARAKIS FELJAN to implement the method of closed-loop as shown in Fig.5 as Arora discloses of implementing a solution for a mismatch before the degradation of KPI thus improving network connectivity (Arora, paragraph [0094], “An advantage of the one or more embodiments of the aforementioned system, method and/or non-transitory machine-readable medium can be allowing for dynamic analysis of NFs, configurations, interfaces and the like, such of mismatch and/or misconfigurations that have occurred after provisioning, or that were not caught in provisioning. This diagnosis can be employed proactively, prior to degradation of KPIs, such as based on known context that can lead to such degradation.”).
Regarding Claim 48, VULGARAKIS FELJAN in view of Curic in further view Tapia discloses the limitations of claim 41.
However, VULGARAKIS FELJAN in view of Curic in further view Tapia do not explicitly disclose wherein the telemetry data comprises at least two of performance management data, fault management data.
Arora further discloses wherein the telemetry data comprises at least two of performance management data (paragraph [0052], Fig.3:328, "such knowledge base 328 can comprise a set of events that have correlation or causality with configuration mismatch… additional context can comprise user entity (UE) events, key logs, metadata, telemetry, traces" (i.e., Par.50-53 radio data analysis system 300 has a knowledge base 328 and that stores data of O-RAN.)),
fault management data (paragraph [0054], Fig.3:328, "Generally, the AAs 314 can report to the controller component 312 upon detection of a context relating to possible KPI degradation. Such context can be any specified event, message and/or the like of a knowledge base 328." (i.e., Par.53 also talks about updated the knowledge base when there is a degradation of KPI.)).
VULGARAKIS FELJAN in view of Curic in further view Tapia and Arora are considered to be analogous to the claimed invention because they are in the same Supervisory, monitoring or testing arrangements. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention to have modified VULGARAKIS FELJAN to implement the method of closed-loop as shown in Fig.5 as aroara discloses of implementing a solution for a mismatch before the degradation of KPI thus improving network connectivity (Arora, paragraph [0094], “An advantage of the one or more embodiments of the aforementioned system, method and/or non-transitory machine-readable medium can be allowing for dynamic analysis of NFs, configurations, interfaces and the like, such of mismatch and/or misconfigurations that have occurred after provisioning, or that were not caught in provisioning. This diagnosis can be employed proactively, prior to degradation of KPIs, such as based on known context that can lead to such degradation.”).
Regarding Claim 56, which is similar in scope to claim 46, thus rejected under the same rationale.
Regarding Claim 58, which is similar in scope to claim 48, thus rejected under the same rationale.
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 Erkin S. Abdullaev whose telephone number is (571)272-4135. The examiner can normally be reached Monday - Friday - 8:00 am - 5:00 pm.
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, Wesley Kim can be reached at (571)272-7867. 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.
ERKIN S. ABDULLAEV
Examiner
Art Unit 2648
/ERKIN ABDULLAEV/Examiner, Art Unit 2648
/WESLEY L KIM/Supervisory Patent Examiner, Art Unit 2648
Prosecution Insights