DETAILED ACTION
Status of Claims
This action is in reply to the communication filed on 01/02/2026.
Claims 1, 2, 8, 9, 15, and 16 have been amended.
Claims 1-20 are currently pending and have been examined.
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 .
Response to Arguments
Applicant's amendments and arguments filed 01/02/2026 with respect to the rejections under 35 USC § 101 and 35 USC § 112 to have been fully considered and resolved the prior issues. The rejections have accordingly been withdrawn.
Applicant’s arguments filed 01/02/2026 with respect to the rejections under 35 USC § 102/103 have been considered but are moot in view of the new grounds of rejection.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 2, 6, 8, 9, 13, 15, 16, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Ke et al. (“Deep reinforcement learning-based computation offloading and resource allocation in security-aware mobile edge computing”, 2021) in view of Hong et al. (“Online Distributed Job Dispatching with Outdated and Partially-Observable Information”, 2020) in further view of Ahvar et al. (“DECA: A Dynamic Energy Cost and Carbon Emission-Efficient Application Placement Method for Edge Clouds”, 2021).
Claims 1, 8 and 15:
Ke discloses the limitations as shown in the following rejections:
A system, comprising: a memory; and a processor coupled to the memory, wherein the processor (wireless device (WD) and/or macro base station (MBS)) performs operations, the operations comprising: providing a plurality of edge computing nodes (MEC servers and/or WD) in a multi-access edge computing environment (pg. 3358, § 1, para. 1; pg. 3360-3361, § 3.1 and Fig. 1).
recommending workload allocation policies (identify as optimal) in the multi-access edge computing environment by determining which policy (policy/action) to use to allocate workloads to the plurality of edge computing nodes (pg. 3357, Abstract; pg. 3362, col. 2; pg. 3364-3365, § 4.2) “the algorithm can learn the optimal policy for decision-making under stringent latency and risk constraints [and] can explore and learn the optimal decision-making policy”; “the decision-making scheme for optimal computation offloading and bandwidth allocation to obtain the minimum weighted sum cost of proposed framework can be designed as an MDP. “
to maximize a probability of carbon footprint requirements being satisfied1 (energy usage is minimized) (3363, §4.1) "the optimal offloading decision-making and bandwidth allocation strategy for the minimization of the weighted sum cost, including the latency, energy consumption, bandwidth cost, and privacy and security cost for all WDs and MEC servers."
wherein the workload allocation policies to use include measures of…observed load of the plurality of edge computing nodes (bandwidth/channel state, execution delay, CPU cycles); and [energy consumption], further discussed below) of the plurality of edge computing nodes (pg. 3362-3363, § 3.4; pg. 3364, col. 2).
wherein in response to a workload request, performing: allocating a workload to a current edge computing node; and transferring the workload to another edge computing node based on the workload allocation policies (pg. 3360, § 2.3, para. 3; pg. 3361, col. 1; pg. 3362; pg. 3364-3365, § 4.2) e.g. “When the computation workloads arrive, DOCRRL selected to partially offload workloads to MEC and partially execute locally at WDs”; “The agent observe the current state in the state space by interacting with the proposed MEC framework, then select an action based on the certain policy.”
As noted, above, Ke discloses “The agent observe the current state in the state space by interacting with the proposed MEC framework, then select an action based on the certain policy”. Ke does not specifically account for uncertainty in freshness of observability data of the plurality of edge computing nodes.
Hong, however, discloses (pg. 315, Abstract; and § I, para. 2; pg. 316, § I, para. 2-3) an analogous method for allocating/dispatching workloads amongst access points (APs) and edge nodes based on policies modelled as an MDP, and specifically considers:
“each AP also suffers from signaling latency, which is the time consumed for each AP to collect system state information under some signaling mechanism...the latency will also lead to outdated information at each dispatcher and information inconsistency among different dispatchers, which may introduce ineliminable estimation error on the number of jobs in the system." (pg. 315).
Hong further discloses (pg. 318, col. 1) wherein the workload allocation policies to use include measures of: uncertainty in freshness (signaling latency/staleness) of observability data (observable state information (OSI)) of the plurality of edge computing nodes; observed load of the plurality of edge computing nodes:
“The observable state information (OSI)… is defined as the aggregation of LSIs of the APs in conflict AP set and the edge servers...The k-th AP is able to collect its OSI Sk(t) at the Dk(t)-th time slots of the t-th broadcast interval, where Dk(t) denotes the signaling latency of the k-th AP at the t-th broadcast interval…dispatching policy of the k-th AP maps from its OSI Sk and its signaling latency Dk to the dispatching action for each job type”
It would have been obvious to one of ordinary skill in the art prior to the filing date of the invention to modify Ke’s model and allocation policies in accordance with Hong’s policy optimization that considers staleness/freshness of observed state information because “the staleness of system state information at the dispatcher of a edge computing systems should be considered. The staleness of information sharing among APs and edge servers may degrade the performance of the job dispatching” (pg. 316, § I), and Hong’s optimization of allocation “policy can achieve obvious and robust performance gain compared with heuristic baselines” (pg. 322, §VII).
As noted above, Ke’s policies include energy consumption of the edge nodes and Ke/Hong do not specifically include carbon emission of the plurality of edge computing nodes.
However, it was known in the art to account and optimize for edge node carbon emission when allocating workloads thereto as evidenced by Ahvar disclosing (pg. 70196; pg. 70198, col. 1) a method for allocating workloads to edge cloud (EC) nodes where workloads are “assigned to an EC, taking into account the geographically varying energy prices and carbon emission rates. The distributed requests in each selected EC are then allocated on appropriate CNs in the EC…our aim is to allocate the application components such that energy costs and carbon emissions are minimized.”
It would have been obvious to one of ordinary skill in the art prior to the filing date of the invention to modify Ke/Hong to consider carbon emissions as taught Ahvar to expand potential optimization objectives to “carbon emissions [that] are becoming pressing issues for the providers of ECs” (pg. 70192) with greater effectiveness than energy consumption minimization alone.
Claims 2, 9, and 16:
The combination of Ke/Hong/Ahvar discloses the limitations as shown in the rejections above. The combination of Ke/Hong/Ahvar further discloses performing operations that act as a plugin to existing orchestration platforms (e.g. Ke macro base station (MBS) Ahvar cloud controller) for allocation policy recommendation (Ke 3358, § 1, para. 2; pg. 3361, col. 1; Ahvar pg. 70195, § III; pg. 70199, IV-B).
Claims 6, 13, and 20:
The combination of Ke/Hong/Ahvar discloses the limitations as shown in the rejections above. Key further discloses wherein entities represent policies as probabilistic transitions determined from historical execution of policy execution logs (traces) (Hong pg. 327; pg. 317, col. 1, para. 2-3). See also Ke pg. 3365, col. 1.
Claims 3, 5, 10, 12, 17, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Ke in view of Hong in view of Ahvar further view of Jurdzinski et al. (“Model Checking Probabilistic Timed Automata with One or Two Clocks”, 2007),
Claims 3, 10, and 17:
The combination of Ke/Hong/Ahvar discloses the limitations as shown in the rejections above. Ke and Hong further discloses modeling workload allocation policies as Probabilistic (Markov Decision Process) (MDP) (Ke pg. 3365, col. 1; Hong pg. 315 Abstract) but does not specifically disclose including clocks in the model.
Jurdzinski, however, shows it was known in the art that PTAs are equivalent to MDPs with clocks, “In this paper, we aim to study model-checking algorithms for probabilistic timed automata, a variant of timed automata extended with discrete probability distributions, or (equivalently) Markov decision processes extended with clocks.” (Jurdzinski, pg. 170).
It would have been obvious to one of ordinary skill in the art prior to the filing date of the invention to modify Ke/Hong/Ahvar’s MDPs to employ clocks as shown by Jurdzinski to allow them to capture real-time dynamics and increase the accuracy of the representation (pg. 170).
Claims 5, 12, and 19:
The combination of Ke/Hong/Ahvar discloses the limitations as shown in the rejections above. Ke and Ahvar further disclose operations model carbon-footprint metric (Ke energy usage; Ahvar carbon emissions model parameters) as labels of a Probabilistic (MDP) and variations in user workloads as probability distributions (Poisson distribution) (Ke 3362, § 3.4; Ahvar pg. 70195-196, § III-A) but does not specifically disclose including clocks in the model.
Jurdzinski, however, shows it was known in the art that PTAs are equivalent to MDPs with clocks, “In this paper, we aim to study model-checking algorithms for probabilistic timed automata, a variant of timed automata extended with discrete probability distributions, or (equivalently) Markov decision processes extended with clocks.” (Jurdzinski, pg. 170).
It would have been obvious to one of ordinary skill in the art prior to the filing date of the invention to modify Ke/Hong/Ahvar’s MDPs to employ clocks as shown by Jurdzinski to allow them to capture real-time dynamics and increase the accuracy of the representation (pg. 170).
Claims 4, 11, and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Ke in view of Hong in view of Ahvar further view of Zhou et al. (“FAS-DQN: Freshness-Aware Scheduling via Reinforcement Learning for Latency-Sensitive Applications”, 2022)
Claims 4, 11, and 18:
Ke/Hong/Ahvar discloses the limitations as shown in the rejections above. Hong measures staleness of observability data via a latency metric, not via a clock, and Ke/Hong/Ahvar does not specifically disclose modeling freshness of observability data using clocks of a Probabilistic Timed Automata.
Zhou, however, discloses an analogous probabilistic model of a distributed system which uses “Age of Information (AoI) to quantify the freshness of data by combining the AoI metric with real-time constraints” (pg. 2381, Abstract) and includes modeling freshness of observability data using clocks (variable tracking AoI) of a Probabilistic Timed Automata (MDP + AoI tracking) (see at least pg. 2381, Abstract; pg. 2383-2384, § 3.1, and Fig. 1).
It would have been obvious to one of ordinary skill in the art prior to the filing date of the invention to modify Ke/Hong/Ahvar to model data freshness to validate the modelled systems ability to meet real-time requirements (pg. Zhou 2381).
Claims 7 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Ke in view of Hong in view of Ahvar further view of Norman et al. (“Model checking for probabilistic timed automata”, 2012).
Claims 7 and 14:
Ke/Hong/Ahvar discloses the limitations as shown in the rejections above. Ke/Hong/Ahvar’s implementation employs MDPs, a form of probabilistic automata, and does not specifically disclose wherein the method utilizes PTA models with a Probabilistic Model Checker to determine a probability of an allocation policy's adherence to a carbon-footprint metric.
Norman, however, discloses (pg. 164-165) utilizing PTA models with a Probabilistic Model Checker to analyze particular model properties, that determine a probability of an allocation policy's adherence to an objective metric, e.g.: “the maximum probability of an airbag failing to deploy within 0.02 seconds…the minimum probability that a packet is correctly delivered with 1 second” (pg. 165); and including energy usage (carbon metric) as an exemplary objective/reward metric (pg. 184-186).
It would have been obvious to one of ordinary skill in the art prior to the filing date of the invention to modify Ding to use PTA models with a PMC to determine if a modeled system has the desired properties, as taught by Norman, to increase confidence the it would work correctly if deployed (Norman pg. 164-165; pg. 187, § 7).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant’s disclosure:
The following are directed to carbon emission aware workload management: US 20210373973 A1; US 20210342185 A1; US 20110282982 A1.
US 20160057039 A1 is directed to methods to determine carbon footprint of a network service.
“On the impact of stale information on distributed online load balancing protocols for edge computing” model based evaluation of staleness in edge computing.
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 of a general nature or relating to the status of this application or concerning this communication or earlier communications from the Examiner should be directed to Paul Mills whose telephone number is 571-270-5482. The Examiner can normally be reached on Monday-Friday 11:00am-8:00pm. If attempts to reach the examiner by telephone are unsuccessful, the Examiner’s supervisor, April Blair can be reached at 571-270-1014.
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.
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.
/P. M./
Paul Mills
04/17/2026
/APRIL Y BLAIR/Supervisory Patent Examiner, Art Unit 2196
1 Examiner notes this limitation is non-limiting intended result as written (MPEP 2111.02; 2111.04)
Prosecution Insights