Neural processing unit and WI-FI scheduling

Abstract

Technology disclosed herein may include an access point including a processing device. The processing device may generate, at an access point, a machine learning model previously trained using training traffic data; identify, at the access point, traffic data; provide, at the access point, the traffic data to the machine learning model; predict, at the access point, a traffic pattern using the machine learning model; and determine, at the access point, a scheduling characteristic based on the traffic pattern.

Description

NEURAL PROCESSING UNIT AND WI-FI SCHEDULINGRELATED APPLICATION

[0001] This application claims the benefit of U.S. Provisional Application No. 63 / 604,896, filed December 1, 2023, the disclosure of which is incorporated herein by reference in its entirety.

[0002] The examples discussed in the present disclosure are related to neural processing units and Wi-Fi® scheduling.BACKGROUND

[0003] Unless otherwise indicated herein, the materials described herein are not prior art to the claims in the present application and are not admitted to be prior art by inclusion in this section.

[0004] An access point (AP), is a networking hardware device that allows other Wi-Fi devices to connect to a wired network. As a standalone device, the AP may have a wired connection to a router, but, in a wireless router, it can also be an integral component of the router itself. There are many wireless data standards that have been introduced for wireless access point and wireless router technology such as 802.11a, 802.11b, 801.11g, 802.1 In (WiFi 4), 802.1 lac (Wi-Fi 5), 802.1 lax (WiFi 6), and so forth.

[0005] The subject matter claimed in the present disclosure is not limited to examples that solve any disadvantages or that operate only in environments such as those described above. Rather, this background is only provided to illustrate one example technology area where some examples described in the present disclosure may be practiced.SUMMARY

[0006] A method may include one or more of: generating, at an access point, a machine learning model previously trained using training traffic data; identifying, at the access point, traffic data; providing, at the access point, the traffic data to the machine learning model; predicting, at the access point, a traffic pattern using the machine learning model; and determining, at the access point, a scheduling characteristic based on the traffic pattern.

[0007] A method may include one or more of: generating, at an access point, a machine learning model previously trained using training power usage data; identifying, at the access point, power usage data; providing, at the access point, the power usage data to the machine learning model; predicting, at the access point, a power usage using the machine learning model; and determining, at the access point, a power setting based on the power usage.

[0008] An access point may include a processing device. The processing device may generate, at an access point, a machine learning model previously trained using training traffic data; identify, at the access point, traffic data; provide, at the access point, the traffic data to the machine learning model; predict, at the access point, a traffic pattern using the machine learning model; and determine, at the access point, a scheduling characteristic based on the traffic pattern.

[0009] The objects and advantages of the embodiments will be realized and achieved at least by the elements, features, and combinations particularly pointed out in the claims.

[0010] Both the foregoing general description and the following detailed description are given as examples and are explanatory and are not restrictive of the invention, as claimed.BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Examples will be described and explained with additional specificity and detail through the use of the accompanying drawings in which:

[0012] FIG. 1 illustrates an example system for training a model.

[0013] FIG. 2A illustrates an example process flow for enhancing Wi-Fi® scheduling.

[0014] FIG. 2B illustrates an example process flow for reducing Wi-Fi® power usage.

[0015] FIG. 3 illustrates a block diagram of an example system configured for Wi-Fi® scheduling.

[0016] FIG. 4 illustrates an example structure of a neural network.

[0017] FIG. 5 illustrates an example structure of a convolutional neural network.

[0018] FIG. 6 illustrates an example process flow of a convolutional neural network.

[0019] FIG. 7 illustrates an example of analog deep learning.

[0020] FIG. 8 illustrates an example of analog deep learning.

[0021] FIG. 9 illustrates an example computing system that may be used for Wi-Fi® scheduling, all arranged in accordance with some examples of the present disclosure.DETAILED DESCRIPTION

[0022] Artificial Intelligence (Al) and Machine Learning (ML) may be integrated in WiFi® routers to enhance network performance and security, automate management, and enhance user experience by adapting to usage patterns. Al and ML may facilitate efficient device connectivity, offer predictive maintenance, and lead to cost savings through optimized resource utilization. The result may be more reliable, personalized, and efficient connectivity.

[0023] There are various other kinds of processing units including central processing units (CPUs), neural processing units (NPUs), graphic processing units (GPUs), tensor processing units (TPUs), neural network processors (NNP), intelligent processing units (IPUs), vision processing units (VPUs). The CPU may be the core processor directing system operations, while the GPU may accelerate image and graphics rendering. The NPUmay specifically boost artificial intelligence tasks by efficiently managing large-scale machine learning computations. While CPUs may provide computations on deep neural networks, CPUs perform at a slower pace compared to GPUs or NPUs. Since an NPU performs Al, it is recommended for Al standards over 1 trillion operations per second (TOPS) to combine a vector digital signal processor (DSP) with an NPU to provide maximum performance with additional programmability.

[0024] The NPU may be used for deep learning. Regular programming is used to write specific instructions for computers to perform tasks, in which logic and algorithms may be defined by programmers. Statistics in programming may involve analyzing data to identify patterns and make predictions to support data-based decisions. Shallow learning may use algorithms to allow computers to learn from data and make informed decisions without complex analysis. Deep learning may employ neural networks to analyze large datasets and excel in tasks like image and speech recognition by learning features directly from the data. The evolution from traditional programming to deep learning signifies a change in computational strategies for solving problems.

[0025] Deep learning algorithms may be used for a wide variety of applications including for artificial intelligence (Al) and / or for machine learning (ML). Deployment of NPUs for Al and ML applications may be broadly applicable. For example, deployment of NPUs may be used for one or more of Wi-Fi scheduling (e.g., predictive traffic management, device fingerprinting, and power control), smart channel selection, and anomaly detection.

[0026] Although a Wi-Fi® access point is illustrated for using machine learning and artificial intelligence, a passive optical network may also be used in conjunction with machine learning and artificial intelligence to enhance scheduling.

[0027] As illustrated in the process flow 100 in FIG. 1, a machine learning model may be trained by inputting training data 102 and an initial model 104 to a trainer 105 togenerate a trained model 106. The trainer 105 may be any combination of hardware and software for training a machine learning model or a neural network. The trainer 105 may receive the type of model that is to be used for training. The trainer 105 may use any appropriate model for training such as a deep learning neural network. The trained model 106 may be used in supervised learning, unsupervised learning, or reinforcement learning. The trained model 106 may be used by the classifier 108 to generate inferences 110. For example, the trained model 106 may be used to generate predictions using data that the trained model has not previously trained on. The training process may be controlled using various parameters such as the maximum model size, the maximum number of passes over the training data, the shuffle type, the regularization type, the learning rate, and the regularization amount.

[0028] Machine learning may be used for Wi-Fi® scheduling. An access point may include a processing device. The processing device (e.g., a neural processing unit) may perform one or more of: (i) generating, at an access point, a machine learning model previously trained using training traffic data; (ii) identifying, at the access point, traffic data; (iii) providing, at the access point, the traffic data to the machine learning model; (iv) predicting, at the access point, a traffic pattern using the machine learning model; or (v) determining, at the access point, a scheduling characteristic based on the traffic pattern.

[0029] The machine learning model may include long short-term memory (LSTM) networks that may be used to accurately forecast traffic patterns. LSTM networks may identify complex temporal dependencies by analyzing historical data which may enhance traffic management. LSTM networks may be scalable (e.g., when handling a large number, e.g., 250, of clients simultaneously).

[0030] Wi-Fi network patterns may be forecast to dynamically fine-tune the medium access control (MAC) scheduler, leveraging orthogonal frequency division multiple access(OFDMA) and multiple resource unit (MRU) settings and calculated bandwidth distribution and modulation and coding scheme (MCS) settings which may enhance overall network throughput, reduce latency, and extend coverage. Therefore, determining the scheduling characteristic may minimize latency relative to a baseline and maximize throughput relative to a baseline. Alternatively or in addition, the scheduling characteristic may include one or more of a MAC scheduler characteristic, an OFDMA setting, an MRU setting, a bandwidth distribution, or an MCS setting.

[0031] The machine learning model may use a deep neural network including one or more of a convolutional neural network or a recurrent neural network. Alternatively or in addition, the machine learning model may use analog deep learning.

[0032] In some examples, a device fingerprint may be determined at the access point.Because Wi-Fi® technology is advancing its security measures by implementing MAC address randomization, device tracking may be used to enhance scheduling in various ways. For example, one or more of the following may be used for device fingerprinting: (i) Signature of station capabilities, (ii) Station ID (SID), (iii) Frequency offset, (iv) Wi-Fi® Packet Rx Time jitter, (v) Intranet / Intemet usage pattern based on ToD (Time of Day), (vi) downstream / upstream usage pattern based on ToD (Time of Day), (vii) virtual private networks (VPNs) used, (viii) packet size used, (ix) presence of a device in the BSS, or (x) RS SI of a fixed device or mobile device.

[0033] The computational load for a machine learning task may result in deferring the computations. A method may include one or more of: identifying, at the access point, a machine learning task; determining, at the access point, a computational load for the machine learning task; and deferring, at the access point, the machine learning task when the computational load is higher than a threshold.

[0034] The selection of machine learning techniques may include a tradeoff between the algorithm’s computational load and the use for real-time processing. Furthermore, privacy may be maintained by performing machine learning techniques locally when privacy is more useful and performing machine learning techniques in the cloud when tasks are less privacysensitive. Integrated AI / ML may facilitate rapid responses to dynamic, real-time conditions. The onboard neural processing unit (NPU) may have sufficient power to execute ML algorithms for local decisions. This approach may boost privacy because the data may remain within the gateway instead of being sent to the cloud. Alternatively or in addition, sophisticated system-wide optimization software may be implemented by third-party applications operating in the cloud.

[0035] Therefore, a processing device (e.g., a neural processing unit) may defer resourceintensive machine learning tasks, splitting computations between cloud-based processing for complex analysis and local Al for privacy-sensitive tasks. By balancing efficiency and privacy, adaptive enhancements may be performed without disrupting real-time operations.

[0036] The processing device may defer certain tasks (e.g., resource-intensive Al computations may be deferred to optimize accelerator usage and system performance). Instead of processing all data points in real-time, the processing device may collect logs and data during active operations and schedule their analysis for a later time when resources are less constrained.

[0037] The deferred tasks may be handled in two ways: (1) cloud-based processing, or (2) local Al processing. For computationally intensive tasks, the system may upload collected data to the cloud, where advanced analysis and model training may occur. To protect user privacy, the data may be compressed, normalized, and quantized before transmission. These transformations may reduce the granularity of the data, making it less identifiable or sensitive while retaining enough detail for effective analysis. The cloud infrastructure may processthis data, identify patterns, and generate optimized models or insights, which may be pushed back to the device for enhanced performance and issue resolution.

[0038] In scenarios where user privacy is to be preserved, computations may occur locally on the device. By keeping data within the device’s ecosystem, this method may eliminate the need for external transmission, reducing privacy risks. Lightweight tasks, such as quick pattern detection or minor adjustments, may be deferred to idle periods (e.g., late at night) to optimize resource usage. Local Al may ensure that even in privacy-sensitive environments, the system may still adapt and improve without external dependencies.

[0039] This dual approach balances advanced processing with privacy and resource efficiency. By compressing and deferring tasks to the cloud or handling them locally, the system may ensure optimal performance while addressing privacy concerns. Over time, these processes enable the device to address issues, such as enhancing traffic scheduling or resolving interference, by learning from past data and applying refined models without disrupting real-time operations.

[0040] Anomalies may be detected using machine learning and / or artificial intelligence. A method may include detecting, at the access point, an anomaly in the traffic pattern using the machine learning model. Incorporating unsupervised learning algorithms, such as autoencoders, may be beneficial for detecting atypical traffic patterns, which may signify network problems or security threats, including Distributed Denial of Service (DDoS) attacks or virus attacks. Virus threats may be detected using anomaly detection, behavioral analysis, and signature-based detection. These tools may work together to identify and mitigate virus infections in network traffic.

[0041] An algorithm that is suitable to detect anomalies may be used (e.g., an algorithm that may detect traffic patterns). By way of example, a few algorithms are provided. Autoencoders may be trained to reconstruct normal data and may fail to reconstruct anomalieswell, which may be detected through the reconstruction error. Alternatively or in addition, an isolation forest may be used. An isolation forest may be efficient for high-dimensional datasets. This algorithm may isolate anomalies instead of profiling normal data points.

[0042] Equipping routers with the ability to detect DDoS anomalies may enhance network security and alert users to take action. Incorporating DDoS anomaly detection into Wi-Fi, including jamming, may bolster network security and reliability, even in the face of traffic overloads and intentional interference.

[0043] Anomaly detection may be performed in various ways. First, anomalies in traffic patterns may be detected. Second, new device behaviors may be learned. Third, evolving norms may be adapted to. For example, time of day analysis and voice-based security may be used for enhanced control and protection.

[0044] Anomaly detection may involve monitoring traffic for unusual spikes, protocols, or device activities. For example, time of day analysis may be used to flag unexpected behaviors (e.g., high traffic late at night). Alternatively or in addition, location-based traffic may be monitored. For example, high traffic late at night may be normal when a particular user is at home, but not when the high traffic comes from outside of the house. Voice fingerprinting and / or beamforming and / or channel state information may be used to determine a location of a user and determine whether the traffic is coming from the user, who may be located inside the house) or from a different source (which may be outside the house). Therefore, time of day analysis may be used in conjunction with other methods to determine anomalies.

[0045] These other methods (e.g., voice fingerprinting) may also be used to authenticate devices that are being on-boarded. For example, profiles of new devices may be generated to validate behaviors over time, integrating legitimate usage into non-anomalous behavior. Voice-based commands for secure onboarding (e.g., “Approve John’s Laptop”) may be used.Adaptive learning may be used to refine traffic baselines with machine learning, distinguishing between benign changes (e.g., a new device) and threats (e.g., rogue devices). Preventative actions may be taken by automatically restricting or quarantining suspicious traffic and devices while notifying users via app. Therefore, secure, adaptive network performance may be obtained by learning from behaviors, leveraging voice authentication, and dynamically addressing anomalies.

[0046] Alternatively or in addition, voice fingerprinting may be used to authenticate users. In some examples, voice fingerprinting may be used with one or more of channel state information or beamforming to authenticate a user. That is, a processing device may be used to integrate voice fingerprinting with Wi-Fi® sensing (using CSI and beamforming) to enhance security, manage network settings, and enable speaker localization for personalized, location-aware services.

[0047] Voice authentication and control may be used to identify users through voice fingerprints for managing Wi-Fi settings (e.g., changing passwords, enabling guest networks). Authentication may be extended to smart devices like locks or video doorbells. In addition, Wi-Fi® sensing and speaker localization may be performed by one or more of using CSI (Channel State Information) to detect speaker location by analyzing Wi-Fi signal reflections; using beamforming to optimize signal delivery to the speaker's area for better interaction, and using Location-awareness to prevent spoofing by validating the speaker's position.

[0048] Other services may be provided including proximity-based commands to adjust system behavior based on speaker location (e.g., directing commands to nearby devices), and personalized settings for individual users based on voice profiles and location. Security may be provided by combining voice authentication with location awareness for robust security, and / or using machine learning to refine voice and location recognition over time, ensuringcontinuous improvement. Therefore, the processing device may be used to provide a seamless, secure, and location-aware method for managing Wi-Fi® and connected devices while leveraging modem Wi-Fi® sensing capabilities.

[0049] Voice fingerprinting may also be used to on-board guest users. For example, a processing device may use voice fingerprinting for secure and seamless guest network access, ensuring only authorized guests may join the network efficiently. Voice authentication may be provided when guests request access by speaking a predefined phrase. The processing device may compare their voice with authorized profiles and grants access if verified. Temporary access may be provided when hosts set time-limited access via voice commands (e.g., “Grant access to Sarah for two hours”), ensuring restricted guest permissions. Onboarding may allow users to use a secure quick response (QR) code or one-time password to connect without manual setup. The combination of voice and device validation may prevent unauthorized access while adapting to returning guest profiles for faster authentication.

[0050] Machine learning and / or artificial intelligence may be used for smart channel selection. A method may include one or more of: predicting, at the access point, a channel characteristic using the machine learning model; or determining, at the access point, an operating channel based on the channel characteristic, wherein the channel characteristic includes one or more of real-time interference or a usage pattern. Deep Q-Networks are a type of deep neural network (DNN) that may combine Q-Leaming, a form of reinforcement learning, with deep neural networks to create systems that can learn to make decisions to maximize a reward signal. Deep Q networks may be used dynamically learn from the environment to identify the optimal channels with minimal interference.

[0051] Channel selection may be performed by monitoring real-time interference and learning usage patterns over time, adapting to peak usage scenarios. For example, a processing device may dynamically optimize WiFi channel selection to minimize interferenceand maximize performance. By continuously monitoring the network environment, the processing device may identify the best available channels to ensure seamless connectivity, even in high-density scenarios such as multi-dwelling units or crowded office spaces.

[0052] Using traffic analysis and environmental scanning, the processing device may detect sources of interference, such as overlapping WiFi® networks or non-WiFi devices (e.g., microwave ovens or baby monitors). Advanced algorithms may be used to select the channel or frequency band that offers the least congestion.

[0053] The processing device may identify patterns in interference over time. For example, the processing device may detect when WiFi® usage increases in the evening as users return home and adjust channel selection proactively to mitigate anticipated interference. This predictive approach may allow the system to stay ahead of congestion trends rather than reacting only to immediate conditions. To adapt to real-time changes, the system periodically re-evaluates the channel configuration, ensuring optimal performance as new interference sources emerge or user traffic patterns shift.

[0054] In one example, a multi -dwelling unit (MDU) may avoid congestion. A processing device may reduce congestion in MDUs using channel scanning, time-of-day analysis, and proactive scheduling to optimize Wi-Fi performance and minimize disruptions. For example, channel scanning may be used to continuously identify and switch to the least congested channel to mitigate interference from neighboring networks. Time-of-Day Analysis may be used to anticipate peak usage periods (e.g., evenings) and preemptively adjust channels and resources to reduce conflicts. Proactive scheduling may be used to balance high-priority traffic across users during congestion while deprioritizing other tasks. Adaptive monitoring may be used to dynamically adjust settings in real time and update congestion strategies based on evolving network conditions.

[0055] In other examples, a processing device (e.g., a neural processing unit) may be used to reduce power usage for Wi-Fi®. A method may include one or more of: generating, at an access point, a machine learning model previously trained using training power usage data; identifying, at the access point, power usage data; providing, at the access point, the power usage data to the machine learning model; predicting, at the access point, a power usage using the machine learning model; or determining, at the access point, a power setting based on the power usage.

[0056] Efficient scheduling may be performed using inference from DNN in the Wi-Fi® MAC layer to enhance system power efficiency. Many settings may be adjusted. For example, sleep modes may be used to turn off one or more of: (a) unused blocks / RAM slices / front end memory (FEM), (b) clock frequencies, or (c) voltage control. Traffic aggregation may also be adjusted. For example, bandwidth may be reduced to an optimal size and bands may be turned off. Packet collisions may be reduced which may reduce wasted power.

[0057] A processing device may use AI / ML to analyze device power usage patterns, tune energy consumption, and comply with various regulations. For example, the European Code of Conduct (CoC) on Energy Consumption of Broadband Equipment specifies power consumption limits for devices like Wi-Fi® gateways and cable modems. These include: (1) Active Modes: devices may not exceed power usage of 6 to 8 watts, depending on functionality; and (2) Standby / Off Modes: Power consumption may be below 0.5 watts, or 0.8 watts if displaying status information.

[0058] Artificial intelligence and machine learning may be used in various ways to reduce power consumption. Energy Pattern Analysis may be used to have AI / ML models analyze real-time and historical device power usage patterns to identify inefficiencies. For example, AI / ML models may detect peak and idle times for better energy allocation.

[0059] Dynamic tuning may be used. Machine learning algorithms may adjust power settings dynamically based on network traffic, user behavior, and environmental factors. This allows devices to switch components into low-power states during idle periods without compromising performance during active use.

[0060] Predictive compliance monitoring may be performed. AI / ML may anticipate scenarios that may breach power limits and proactively adjusts device operations to prevent violations. For instance, AI / ML may preemptively reduce power draw during high-load periods by rebalancing processes or optimizing thermal management.

[0061] Cloud integrated model updates may be provided. That is, Al models may be periodically refined using cloud-based training, incorporating data from similar devices to improve power management strategies. Updates may be pushed back to the device for continuous enhancement.

[0062] Various machine learning models may be used to reduce power usage including a deep neural network including one or more of a convolutional neural network or a recurrent neural network. Alternatively or in addition, analog deep learning may be used.

[0063] FIG. 2A illustrates a process flow of an example method 200 of Wi-Fi scheduling, in accordance with at least one example described in the present disclosure. The method 200 may be arranged in accordance with at least one example described in the present disclosure. The method 200 may be performed by processing logic that may include hardware (circuitry, dedicated logic, etc.), software (such as is run on a computer system or a dedicated machine), or a combination of both, which processing logic may be included in the processing device 902 of FIG. 9, the communication system 300 of FIG. 3, or another device, combination of devices, or systems.

[0064] The method 200 may begin at block 205 where the processing logic may generate, at an access point, a machine learning model previously trained using training traffic data.

[0065] At block 210, the processing logic may identify, at the access point, traffic data.

[0066] At block 215, the processing logic may provide, at the access point, the traffic data to the machine learning model.

[0067] At block 220, the processing logic may predict, at the access point, a traffic pattern using the machine learning model.

[0068] At block 225, the processing logic may determine, at the access point, a scheduling characteristic based on the traffic pattern.

[0069] The processing logic may further determine, at the access point, a device fingerprint.

[0070] The processing logic may further: identify, at the access point, a machine learning task; determine, at the access point, a computational load for the machine learning task; and defer, at the access point, the machine learning task when the computational load is higher than a threshold.

[0071] The processing logic may further detect, at the access point, an anomaly in the traffic pattern using the machine learning model.

[0072] The processing logic may further authenticate, at the access point, a user using voice fingerprinting and one or more of channel state information of beamforming.

[0073] The processing logic may further predict, at the access point, a channel characteristic using the machine learning model; and determine, at the access point, an operating channel based on the channel characteristic in which the chancel characteristic may include one or more of real time interference or a usage pattern.

[0074] In one example, determining the scheduling characteristic may minimize latency and maximize throughput. In another example, the scheduling characteristic may include one or more of medium access control (MAC) scheduler characteristic, an orthogonal frequency division multiple access (OFDMA) setting, a multiple resource unit (MRU) setting, abandwidth distribution, or a modulation coding scheme (MCS) setting. In one example, the machine learning model may use a deep neural network including one or more of a convolutional neural network or a recurrent neural network. In one example, the machine learning model may use analog deep learning.

[0075] The machine learning model may use a deep neural network including one or more of a convolutional neural network or a recurrent neural network. The machine learning model may use analog deep learning. In one example, prioritizing the data packet may minimize latency and maximize throughput. In another example, the traffic data may include one or more of packet size, flow duration, or protocol behavior.

[0076] Modifications, additions, or omissions may be made to the method 200 without departing from the scope of the present disclosure. For example, in some examples, the method 200 may include any number of other components that may not be explicitly illustrated or described.

[0077] FIG. 2B illustrates a process flow of an example method 250 of reducing Wi-Fi® power usage, in accordance with at least one example described in the present disclosure. The method 250 may be arranged in accordance with at least one example described in the present disclosure. The method 250 may be performed by processing logic that may include hardware (circuitry, dedicated logic, etc.), software (such as is run on a computer system or a dedicated machine), or a combination of both, which processing logic may be included in the processing device 902 of FIG. 9, the communication system 300 of FIG. 3, or another device, combination of devices, or systems.

[0078] The method 250 may begin at block 255 where the processing logic may generate, at an access point, a machine learning model previously trained using training power usage data.

[0079] At block 260, the processing logic may identify, at the access point, power usage data.

[0080] At block 265, the processing logic may provide, at the access point, the power usage data to the machine learning model.

[0081] At block 270, the processing logic may predict, at the access point, a power usage using the machine learning model.

[0082] At block 275, the processing logic may determine, at the access point, a power setting based on the power usage.

[0083] The processing logic may further activate a sleep mode based on the power usage. The processing logic may further aggregate traffic by one or more of reducing bandwidth or turning off bands. The processing logic may further reduce, at the access point, packet collisions.

[0084] The machine learning model may use a deep neural network including one or more of a convolutional neural network or a recurrent neural network. The machine learning model may use analog deep learning.

[0085] Modifications, additions, or omissions may be made to the method 250 without departing from the scope of the present disclosure. For example, in some examples, the method 250 may include any number of other components that may not be explicitly illustrated or described.

[0086] For simplicity of explanation, methods and / or process flows described herein are depicted and described as a series of acts. However, acts in accordance with this disclosure may occur in various orders and / or concurrently, and with other acts not presented and described herein. Further, not all illustrated acts may be used to implement the methods in accordance with the disclosed subject matter. In addition, those skilled in the art will understand and appreciate that the methods may alternatively be represented as a series ofinterrelated states via a state diagram or events. Additionally, the methods disclosed in this specification are capable of being stored on an article of manufacture, such as a non- transitory computer-readable medium, to facilitate transporting and transferring such methods to computing devices. The term article of manufacture, as used herein, is intended to encompass a computer program accessible from any computer-readable device or storage media. Although illustrated as discrete blocks, various blocks may be divided into additional blocks, combined into fewer blocks, or eliminated, depending on the desired implementation.

[0087] FIG. 3 illustrates a block diagram of an example communication system 300 configured for coordinated scheduling, in accordance with at least one example described in the present disclosure. The communication system 300 may include a digital transmitter 302, a radio frequency circuit 304, a device 314, a digital receiver 306, and a processing device 308. The digital transmitter 302 and the processing device may be configured to receive a baseband signal via connection 310. A transceiver 316 may comprise the digital transmitter 302 and the radio frequency circuit 304.

[0088] In some examples, the communication system 300 may include a system of devices that may be configured to communicate with one another via a wired or wireline connection. For example, a wired connection in the communication system 300 may include one or more Ethernet cables, one or more fiber-optic cables, and / or other similar wired communication mediums. Alternatively, or additionally, the communication system 300 may include a system of devices that may be configured to communicate via one or more wireless connections. For example, the communication system 300 may include one or more devices configured to transmit and / or receive radio waves, microwaves, ultrasonic waves, optical waves, electromagnetic induction, and / or similar wireless communications. Alternatively, or additionally, the communication system 300 may include combinations of wireless and / or wired connections. In these and other examples, the communication system 300 may includeone or more devices that may be configured to obtain a baseband signal, perform one or more operations to the baseband signal to generate a modified baseband signal, and transmit the modified baseband signal, such as to one or more loads.

[0089] In some examples, the communication system 300 may include one or more communication channels that may communicatively couple systems and / or devices included in the communication system 300. For example, the transceiver 316 may be communicatively coupled to the device 314.

[0090] In some examples, the transceiver 316 may be configured to obtain a baseband signal. For example, as described herein, the transceiver 316 may be configured to generate a baseband signal and / or receive a baseband signal from another device. In some examples, the transceiver 316 may be configured to transmit the baseband signal. For example, upon obtaining the baseband signal, the transceiver 316 may be configured to transmit the baseband signal to a separate device, such as the device 314. Alternatively, or additionally, the transceiver 316 may be configured to modify, condition, and / or transform the baseband signal in advance of transmitting the baseband signal. For example, the transceiver 316 may include a quadrature up-converter and / or a digital to analog converter (DAC) that may be configured to modify the baseband signal. Alternatively, or additionally, the transceiver 316 may include a direct radio frequency (RF) sampling converter that may be configured to modify the baseband signal.

[0091] In some examples, the digital transmitter 302 may be configured to obtain a baseband signal via connection 310. In some examples, the digital transmitter 302 may be configured to up-convert the baseband signal. For example, the digital transmitter 302 may include a quadrature up-converter to apply to the baseband signal. In some examples, the digital transmitter 302 may include an integrated digital to analog converter (DAC). The DAC may convert the baseband signal to an analog signal, or a continuous time signal. Insome examples, the DAC architecture may include a direct RF sampling DAC. In some examples, the DAC may be a separate element from the digital transmitter 302.

[0092] In some examples, the transceiver 316 may include one or more subcomponents that may be used in preparing the baseband signal and / or transmitting the baseband signal. For example, the transceiver 316 may include an RF front end (e.g., in a wireless environment) which may include a power amplifier (PA), a digital transmitter (e.g., 302), a digital front end, an Institute of Electrical and Electronics Engineers (IEEE) 1588v2 device, a Long-Term Evolution (LTE) physical layer (L-PHY), an (S-plane) device, a management plane (M-plane) device, an Ethernet media access control (MAC) / personal communications service (PCS), a resource controller / scheduler, and the like. In some examples, a radio (e.g., a radio frequency circuit 304) of the transceiver 316 may be synchronized with the resource controller via the S-plane device, which may contribute to high-accuracy timing with respect to a reference clock.

[0093] In some examples, the transceiver 316 may be configured to obtain the baseband signal for transmission. For example, the transceiver 316 may receive the baseband signal from a separate device, such as a signal generator. For example, the baseband signal may come from a transducer configured to convert a variable into an electrical signal, such as an audio signal output of a microphone picking up a speaker’s voice. Alternatively, or additionally, the transceiver 316 may be configured to generate a baseband signal for transmission. In these and other examples, the transceiver 316 may be configured to transmit the baseband signal to another device, such as the device 314.

[0094] In some examples, the device 314 may be configured to receive a transmission from the transceiver 316. For example, the transceiver 316 may be configured to transmit a baseband signal to the device 314.

[0095] In some examples, the radio frequency circuit 304 may be configured to transmit the digital signal received from the digital transmitter 302. In some examples, the radio frequency circuit 304 may be configured to transmit the digital signal to the device 314 and / or the digital receiver 306. In some examples, the digital receiver 306 may be configured to receive a digital signal from the RF circuit and / or send a digital signal to the processing device 308.

[0096] In some examples, the processing device 308 may be a standalone device or system, as illustrated. Alternatively, or additionally, the processing device 308 may be a component of another device and / or system. For example, in some examples, the processing device 308 may be included in the transceiver 316. In instances in which the processing device 308 is a standalone device or system, the processing device 308 may be configured to communicate with additional devices and / or systems remote from the processing device 308, such as the transceiver 316 and / or the device 314. For example, the processing device 308 may be configured to send and / or receive transmissions from the transceiver 316 and / or the device 314. In some examples, the processing device 308 may be combined with other elements of the communication system 300.Examples

[0097] The following provide examples of machine learning algorithms according to embodiments of the present disclosure.Example 1A: Deep Neural Network - Inference

[0098] As illustrated in the neural network 400 in FIG. 4, the number of input features may be three neurons and the number of output layers may be two neurons. The input features may include a red neuron 410, a green neuron 420, and a blue neuron 430. The red neuron 410 may be connected to a Stop Bl neuron via a weight Wl_l and to a Go B2 neuron via a weight W 1 2, with values provided in Table 1. The green neuron 420 may be connected to a Stop B 1neuron 440 via a weight W2_l and a Go B2 neuron 450 via a weight W2_2, with values provided in Table 1. The blue neuron 430 may be connected to a Stop Bl neuron 440 via a weight W3_l and a Go B2 neuron 450 via a weight W3_2, with values provided in Table 1. Weights and biases are hardcoded for simplicity. Each neuron may use a sigmoid / tanh and / or rectified linear unit (ReLU) as its activation function. There may not be training but there may be a forward pass. The EPOCH may be 5000.Table 1: Weights for Stop and Go

[0099] Values may be provided by multiplying and adding (i.e., a dot product). A ‘stop’ value may be provided by (Red*Wl_l) + (Green*W2_l) + (Blue*W3_l) + Bl. A ‘go’ value may be provided by (Red*Wl_2) + (Green*W2_2) + (Blue*W3_2) + B2. A ‘stop’ value operated on by the rectified linear unit may be provided by max(iStop, 0). A ‘go’ value operated on by the rectified linear unit may be provided by max(iGo, 0).

[0100] The results of the deep neural network (DNN) are provided in Table 2.Table 2: DNN Results

[0101] In some examples, Regular DNN may be used with more layers and neurons. Alternatively or in addition, training algorithms may be used (e.g., with gradient descent with backpropagation) to adjust the weights and biases based on a dataset.Example IB: Deep Neural Network - Traffic Lights for Japan

[0102] The previous example may be used to model traffic lights in Japan. The weights for stop and go may be as shown in Table 3.Table 3: Weights for Stop and Go

[0103] The results of the deep neural network are shown in Table 4.Table 4: DNN ResultsExample 2: Convolutional Neural Networks (CNNs) for Image

[0104] As illustrated in the convolutional neural network 500 in FIG. 5 and the corresponding process flow 600 in FIG. 6, an image 610 may be processed using a convolutional neural network 500. An input layer 501 (i.e. receives the input image) may include inputs II 502, 12 504, 13 506, and In 508. A hidden layer 510 may include weights Hl 1 512, H12 514, and Hix 516. Another hidden layer k 520 may include weights Hkl 522, Hk2 524, and Hky 526. An output layer 530 (i.e. produces the final output) may include an output 01 532 and an output Oz 534.

[0105] As illustrated in the process flow 600 in FIG. 6, an input image 610 may be subject to convolution and ReLU 620 in a convolutional layer. The convolutional layer may apply a learned filter to extract features from the image. The ReLU may be applied to introduce nonlinearity to allow the network to handle the complex patterns. The resulting output may be subject to pooling 630 in a pooling layer. The pooling layer may reduce the spatial dimensions of the feature maps, which may summarize the features. The resulting output may be subject to convolution and ReLU 640 and pooling 650. The resulting output may be flattened in a flatten 660 operation. The resulting output may be a fully connected 670 layer which may integrate the learned features and assist in making a final decision or classification. The final output may be a class label in classification tasks using an activation function (e.g., softmax 680) appropriate for the operation.Example 3: Analog Deep Learning

[0106] Analog deep learning may use analog hardware (e.g., programmable resistors) for neural network computations, offering advantages in energy efficiency and processing speed due to its ability to handle continuous signals directly.

[0107] As illustrated in the diagram 700 in FIG. 7, an input layer 710 having inputs xl 712, x2 714, and xm 716 may be used with weights wl 722, w2 724, wm 726 to provide a sum 730 with a bias 740. The sum 730 may be subject to an activation function 750 toproduce an output Y 760. The weights may be modeled using memristors or transistors for synaptic weights, operational amplifiers for summing inputs, and diodes for non-linear activation functions which may allow the analog deep learning structure to process signals using electrical currents and voltages for faster and more energy efficient computations when compared to digital approaches.

[0108] As illustrated in the device 800 in FIG. 8, an operational amplifier 850 may be used to model a deep neural network. A first voltage VI 802 may provide a first current II 812 to a first resistor Rin 822. A second voltage 804 may provide a second current 12 814 to a second resistor Rin 824. A third voltage 806 may provide a third current 13 816 to a third resistor Rin 826. A feedback current If 818 from Rf 842 may be provided to a virtual earth summing point 830 along with currents II 812, 12 814, and 13 816. The virtual earth summing point 830 may be input to the negative terminal of the operational amplifier 840. The positive terminal of the operational amplifier 840 may be connected to ground 850. The output terminal of the operational amplifier 840 may be vout 860. Therefore, the device 800 may include inputs, weights, sums, biases, and outputs which may be used to model a deep neural network.

[0109] Therefore, deep neural networks may be used with little computational delay. The input and output may use analog to digital converters and / or digital to analog converters to interface. The weights and / or biases may be controlled using the analog to digital converters and / or digital to analog converters. Capacitors may be used to store the weights and to store the biases derived during backpropagation.

[0110] Figure 9 illustrates a diagrammatic representation of a machine in the example form of a computing device 900 within which a set of instructions, for causing the machine to perform any one or more of the methods discussed herein, may be executed. The computing device 900 may include a rackmount server, a router computer, a server computer, a mainframecomputer, a laptop computer, a tablet computer, a desktop computer, or any computing device with at least one processor, etc., within which a set of instructions, for causing the machine to perform any one or more of the methods discussed herein, may be executed. In alternative examples, the machine may be connected (e.g., networked) to other machines in a local area network (LAN), an intranet, an extranet, or the Internet. The machine may operate in the capacity of a server machine in client-server network environment. Further, while only a single machine is illustrated, the term “machine” may also include any collection of machines that individually or jointly execute a set (or multiple sets) of instructions to perform any one or more of the methods discussed herein.

[0111] The example computing device 900 includes a processing device (e.g., a processor) 902, a main memory 904 (e.g., read-only memory (ROM), flash memory, dynamic random access memory (DRAM) such as synchronous DRAM (SDRAM)), a static memory 906 (e.g., flash memory, static random access memory (SRAM)) and a data storage device 916, which communicate with each other via a bus 908.

[0112] Processing device 902 represents one or more general -purpose processing devices such as a microprocessor, central processing unit, or the like. More particularly, the processing device 902 may include a complex instruction set computing (CISC) microprocessor, reduced instruction set computing (RISC) microprocessor, very long instruction word (VLIW) microprocessor, or a processor implementing other instruction sets or processors implementing a combination of instruction sets. The processing device 902 may also include one or more special-purpose processing devices such as an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a digital signal processor (DSP), network processor, or the like. The processing device 902 is configured to execute instructions 926 for performing the operations and steps discussed herein.

[0113] The computing device 900 may further include a network interface device 922 which may communicate with a network 918. The computing device 900 also may include a display device 910 (e.g., a liquid crystal display (LCD) or a cathode ray tube (CRT)), an alphanumeric input device 912 (e.g., a keyboard), a cursor control device 914 (e.g., a mouse) and a signal generation device 920 (e.g., a speaker). In at least one example, the display device 910, the alphanumeric input device 912, and the cursor control device 914 may be combined into a single component or device (e.g., an LCD touch screen).

[0114] The data storage device 916 may include a computer-readable storage medium 924 on which is stored one or more sets of instructions 926 embodying any one or more of the methods or functions described herein. The instructions 926 may also reside, completely or at least partially, within the main memory 904 and / or within the processing device 902 during execution thereof by the computing device 900, the main memory 904 and the processing device 902 also constituting computer-readable media. The instructions may further be transmitted or received over a network 918 via the network interface device 922.

[0115] Some portions of the detailed description refer to different modules configured to perform operations. One or more of the modules may include code and routines configured to enable a computing system to perform one or more of the operations described therewith.Additionally or alternatively, one or more of the modules may be implemented using hardware including any number of processors, microprocessors (e.g., to perform or control performance of one or more operations), DSPs, FPGAs, ASICs or any suitable combination of two or more thereof. Alternatively or additionally, one or more of the modules may be implemented using a combination of hardware and software. In the present disclosure, operations described as being performed by a particular module may include operations that the particular module may direct a corresponding system (e.g., a corresponding computing system) to perform. Further, the delineating between the different modules is to facilitateexplanation of concepts described in the present disclosure and is not limiting. Further, one or more of the modules may be configured to perform more, fewer, and / or different operations than those described such that the modules may be combined or delineated differently than as described.

[0116] Some portions of the detailed description are presented in terms of algorithms and symbolic representations of operations within a computer. These algorithmic descriptions and symbolic representations are the means used by those skilled in the data processing arts to convey the essence of their innovations to others skilled in the art. An algorithm is a series of configured operations leading to a desired end state or result. In example implementations, the operations carried out require physical manipulations of tangible quantities for achieving a tangible result.

[0117] Unless specifically stated otherwise, as apparent from the discussion, it is appreciated that throughout the description, discussions utilizing terms such as detecting, determining, analyzing, identifying, scanning or the like, can include the actions and processes of a computer system or other information processing device that manipulates and transforms data represented as physical (electronic) quantities within the computer system’s registers and memories into other data similarly represented as physical quantities within the computer system’s memories or registers or other information storage, transmission or display devices.

[0118] Example implementations may also relate to an apparatus for performing the operations herein. This apparatus may be specially constructed for the required purposes, or it may include one or more general -purpose computers selectively activated or reconfigured by one or more computer programs. Such computer programs may be stored in a computer readable medium, such as a computer-readable storage medium or a computer-readable signal medium. Computer- executable instructions may include, for example, instructionsand data which cause a general- purpose computer, special-purpose computer, or specialpurpose processing device (e.g., one or more processors) to perform or control performance of a certain function or group of functions.

[0119] Although the subject matter has been described in language specific to structural features and / or methodological acts, it is to be understood that the subject matter configured in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims.

[0120] Unless specific arrangements described herein are mutually exclusive with one another, the various implementations described herein can be combined in whole or in part to enhance system functionality and / or to produce complementary functions. Likewise, aspects of the implementations may be implemented in standalone arrangements. Thus, the above description has been given by way of example only and modification in detail may be made within the scope of the present disclosure.

[0121] With respect to the use of substantially any plural or singular terms herein, those having skill in the art can translate from the plural to the singular or from the singular to the plural as is appropriate to the context or application. The various singular / plural permutations may be expressly set forth herein for sake of clarity. A reference to an element in the singular is not intended to mean “one and only one” unless specifically stated, but rather “one or more.” Moreover, nothing disclosed herein is intended to be dedicated to the public regardless of whether such disclosure is explicitly recited in the above description.

[0122] In general, terms used herein, and especially in the appended claims (e.g., bodies of the appended claims) are generally intended as “open” terms (e.g., the term “including” should be interpreted as “including but not limited to,” the term “having” should be interpreted as “having at least,” the term “includes” should be interpreted as “includes but is not limitedto,” etc.). Furthermore, in those instances where a convention analogous to “at least one of A, B, and C, etc.” is used, in general, such a construction is intended in the sense one having skill in the art would understand the convention (e.g., “a system having at least one of A, B, and C” would include but not be limited to systems that include A alone, B alone, C alone, A and B together, A and C together, B and C together, or A, B, and C together, etc.). Also, a phrase presenting two or more alternative terms, whether in the description, claims, or drawings, should be understood to include one of the terms, either of the terms, or both terms. For example, the phrase “A or B” will be understood to include the possibilities of “A” or “B” or “A and B.”

[0123] Additionally, the use of the terms “first,” “second,” “third,” etc., are not necessarily used herein to connote a specific order or number of elements. Generally, the terms “first,” “second,” “third,” etc., are used to distinguish between different elements as generic identifiers. Absence a showing that the terms “first,” “second,” “third,” etc., connote a specific order, these terms should not be understood to connote a specific order.Furthermore, absence a showing that the terms first,” “second,” “third,” etc., connote a specific number of elements, these terms should not be understood to connote a specific number of elements.

[0124] The present disclosure may be embodied in other specific forms without departing from its spirit or essential characteristics. The described implementations are to be considered in all respects only as illustrative and not restrictive. The scope of the disclosure is, therefore, indicated by the appended claims rather than by the foregoing description. All changes which come within the meaning and range of equivalency of the claims are to be embraced within their scope.

Claims

CLAIMSWhat is claimed is:

1. A method, comprising: generating, at an access point, a machine learning model previously trained using training traffic data; identifying, at the access point, traffic data; providing, at the access point, the traffic data to the machine learning model; predicting, at the access point, a traffic pattern using the machine learning model; and determining, at the access point, a scheduling characteristic based on the traffic pattern.

2. The method of claim 1, wherein determining the scheduling characteristic minimizes latency and maximizes throughput.

3. The method of claim 1, wherein the scheduling characteristic includes one or more of a medium access control (MAC) scheduler characteristic, an orthogonal frequencydivision multiple access (OFDMA) setting, a multiple resource unit (MRU) setting, a bandwidth distribution, or a modulation coding scheme (MCS) setting.

4. The method of claim 1, wherein the machine learning model uses a deep neural network including one or more of a convolutional neural network or a recurrent neural network.

5. The method of claim 1, wherein the machine learning model uses analog deep learning.

6. The method of claim 1, further comprising: determining, at the access point, a device fingerprint.

7. The method of claim 1, further comprising: identifying, at the access point, a machine learning task; determining, at the access point, a computational load for the machine learning task; and deferring, at the access point, the machine learning task when the computational load is higher than a threshold.

8. The method of claim 1, further comprising: detecting, at the access point, an anomaly in the traffic pattern using the machine learning model.

9. The method of claim 1, further comprising: authenticating, at the access point, a user using voice fingerprinting and one or more of channel state information or beamforming.

10. The method of claim 1, further comprising: predicting, at the access point, a channel characteristic using the machine learning model; and determining, at the access point, an operating channel based on thechannel characteristic, wherein the channel characteristic includes one or more of real-time interference or a usage pattern.

11. A method, comprising: generating, at an access point, a machine learning model previously trained using training power usage data; identifying, at the access point, power usage data; providing, at the access point, the power usage data to the machine learning model; predicting, at the access point, a power usage using the machine learning model; and determining, at the access point, a power setting based on the power usage.

12. The method of claim 11, further comprising: activating a sleep mode based on the power usage.

13. The method of claim 11, further comprising: aggregating traffic by one or more of reducing bandwidth or turning off bands.

14. The method of claim 11, further comprising: reducing, at the access point, packet collisions.

15. The method of claim 11, wherein the machine learning model uses a deep neural network including one or more of a convolutional neural network or a recurrent neural network.

16. The method of claim 11, wherein the machine learning model uses analog deep learning.

17. An access point, comprising: a processing device operable to: generate, at an access point, a machine learning model previously trained using training traffic data; identify, at the access point, traffic data; provide, at the access point, the traffic data to the machine learning model; predict, at the access point, a traffic pattern using the machine learning model; and determine, at the access point, a scheduling characteristic based on the traffic pattern.

18. The access point of claim 17, further comprising: determine, at the access point, a device fingerprint.

19. The access point of claim 17, further comprising: identify, at the access point, a machine learning task; determine, at the access point, a computational load for the machine learning task; and defer, at the access point, the machine learning task when the computational load is higher than a threshold.

20. The access point of claim 17, further comprising: detect, at the access point, an anomaly using the machine learning model.

Images