# US 12,120,604 B2 · Cross-node deep learning methods of selecting machine learning modules in wireless communication systems > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,120,604 B2 - Document ID: US12120604B2 - Status: Granted - Kind: B2 - Application number: 18/496,761 - Filed: 2023-10-27 - Published: 2024-04-18 - Granted: 2024-10-15 - Earliest claimed: 2021-02-03 - Assignees: QUALCOMM Incorporated - Applicants: QUALCOMM Incorporated - Inventors: Tianyang Bai; Hua Wang; Junyi Li - Web page: https://patentgrep.com/patent/US12120604B2 ## CPC - H04W 24/02: Arrangements for optimising operational condition - H04W 48/18: Selecting a network or a communication service - G06N 3/045: Combinations of networks - G06N 3/0464: Convolutional networks [CNN, ConvNet] - G06N 3/09: Supervised learning - G06N 3/092: Reinforcement learning - G06N 3/098: Distributed learning, e.g. federated learning - G06N 20/20: Ensemble learning - H04W 88/02: Terminal devices - H04W 88/08: Access point devices - G06N 3/044: Recurrent networks, e.g. Hopfield networks - G06N 3/048: Activation functions - G06N 3/084: Backpropagation, e.g. using gradient descent - H04W: WIRELESS COMMUNICATION NETWORKS (broadcast communication H04H; communication systems using wireless links for non-selective communication, e.g. wireless extensions H04M1/72) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract A method of wireless communication is performed by a user equipment (UE). The method receives, from a network, a configuration for a decision making module. The method also executes the decision making module to determine a selection parameter for a configuration of at least one machine learning module. The method selects the configuration of the at least one machine learning module based on the selection parameter. Further, the method reports, to the network, a decision resulting from executing the decision making module. ## Claims ### Claim 1 A method of wireless communication by a user equipment (UE), comprising: receiving, from a network, a configuration for a decision making module; executing the decision making module to identify a selection parameter from a plurality of selection parameters for configuring at least one machine learning module at the UE for predicting a future beam for communication with the network; configuring the at least one machine learning module for predicting the future beam in accordance with the selection parameter; and reporting, to the network, a decision resulting from executing the decision making module. ### Claim 2 (depends on 1) The method of claim 1, wherein the selection parameter enables the UE to predict the future beam in accordance with prior beam measurements. ### Claim 3 (depends on 1) The method of claim 1, wherein the selection parameter corresponds to a region in which the UE operates. ### Claim 4 (depends on 1) The method of claim 1, wherein the at least one machine learning module corresponds to a UE moving speed and/or UE trajectory. ### Claim 5 (depends on 1) The method of claim 1, wherein the at least one machine learning module comprises a plurality of machine learning modules, each corresponding to a type of UE mobility. ### Claim 6 (depends on 5) The method of claim 5, wherein the decision making module estimates a Doppler condition and/or a UE speed in accordance with SSB measurements and/or CSI-RS measurements, in order to identify the selection parameter. ### Claim 7 (depends on 1) The method of claim 1, wherein the at least one machine learning module comprises a single machine learning module, and the selection parameter comprises an updated machine learning module configuration. ### Claim 8 (depends on 7) The method of claim 7, wherein the decision making module estimates a Doppler condition in accordance with at least one of TRS measurements, CSI-RS measurements, and location information associated with a PRS, in order to identify the selection parameter. ### Claim 9 (depends on 7) The method of claim 7, further comprising requesting a model update in response to the decision making module identifying a need for a new machine learning model. ### Claim 10 (depends on 7) The method of claim 7, wherein the selection parameter enables the UE to predict the future beam in accordance with more frequent beam measurements in accordance with the UE having a higher mobility level, and in accordance with less frequent beam measurements in accordance with the UE having a lower mobility level. ### Claim 11 (depends on 10) The method of claim 10, wherein the more frequent beam measurements include CSI-RS measurements and SSB measurements. ### Claim 12 (depends on 1) The method of claim 1, wherein the decision making module comprises a reinforcement learning module. ### Claim 13 (depends on 1) The method of claim 1, further comprising transmitting information to the network for training the decision making module. ### Claim 14 (depends on 1) The method of claim 1, wherein the information comprises input to the decision making module, a decision in accordance with input, and performance resulting from the decision. ### Claim 15 An apparatus for wireless communication at a user equipment (UE), comprising: at least one memory; and at least one processor coupled to the at least one memory and configured to receive, from a network, a configuration for a decision making module; execute the decision making module to identify a selection parameter from a plurality of selection parameters for configuring at least one machine learning module at the UE for predicting a future beam for communication with the network; configure the at least one machine learning module for predicting the future beam in accordance with the selection parameter; and report, to the network, a decision resulting from executing the decision making module. ### Claim 16 (depends on 15) The apparatus of claim 15, wherein the selection parameter enables the UE to predict the future beam in accordance with prior beam measurements. ### Claim 17 (depends on 15) The apparatus of claim 15, wherein the selection parameter corresponds to a region in which the UE operates. ### Claim 18 (depends on 15) The apparatus of claim 15, wherein the at least one machine learning module corresponds to a UE moving speed and/or UE trajectory. ### Claim 19 (depends on 15) The apparatus of claim 15, wherein the at least one machine learning module comprises a plurality of machine learning modules, each corresponding to a type of UE mobility. ### Claim 20 (depends on 19) The apparatus of claim 19, wherein the decision making module estimates a Doppler condition and/or a UE speed in accordance with SSB measurements and/or CSI-RS measurements, in order to identify the selection parameter.