# US 12,015,558 B2 · Enhancing classification of data packets in an electronic device > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,015,558 B2 - Document ID: US12015558B2 - Status: Granted - Kind: B2 - Application number: 17/525,069 - Filed: 2021-11-12 - Published: 2022-06-02 - Granted: 2024-06-18 - Earliest claimed: 2020-11-27 - Assignees: ARRIS Enterprises LLC - Applicants: ARRIS Enterprises LLC - Inventors: Sabarinathan Nagarajan; Ian G. Wheelock; Lakshmi Arunkumar - Web page: https://patentgrep.com/patent/US12015558B2 ## CPC - H04L 47/2441: relying on flow classification, e.g. using integrated services [IntServ] - H04L 41/16: using machine learning or artificial intelligence - H04L 45/74: Address processing for routing - H04L 47/20: Traffic policing - G06N 20/00: Machine learning - H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract An electronic device in a network capable of enhancing classification of packets received by the electronic device is provided that includes a network interface, a non-transitory memory having instructions stored thereon, and a hardware processor. The hardware processor executes the instructions to receive a packet in at least one packet flow to be processed using a first service flow or a second service flow, determine whether a packet to be processed using the second service flow is incorrectly classified for processing using the first service flow, and move the incorrectly classified packet to the second service flow. A notification including data is transmitted for the incorrectly classified packet to a machine learning algorithm (MLA). The MLA is trained using the data and the trained MLA categorizes at least one new packet to be processed using the first or second service flow. ## Claims ### Claim 1 An electronic device capable of enhancing classification of packets in a network, the electronic device comprising: a network interface; a non-transitory memory having instructions stored thereon; and a hardware processor configured to execute the instructions to: receive, using the network interface, a packet to be processed using a first service flow queue or a second service flow queue, the first service flow queue comprising at least one service flow; determine whether a packet to be processed using the second service flow queue is incorrectly classified for processing using the first service flow queue based on a delay in service flow in the first service flow queue caused by the packet; move the incorrectly classified packet from the first service flow queue to the second service flow queue; transmit a notification for the incorrectly classified packet to a machine learning algorithm (MLA), each notification including data related to the incorrectly classified packet; train the MLA using the data in the notifications; and categorize, using the trained MLA, at least one new packet to be processed using the first service flow queue or the second service flow queue by modifying packet header information, wherein the at least one new packet is processed based on the category determined by the MLA. ### Claim 2 (depends on 1) The electronic device according to claim 1, wherein the hardware processor further executes the instructions to transmit, using the network interface, the notifications to a cloud-based device for training an updated MLA. ### Claim 3 (depends on 2) The electronic device according to claim 2, wherein the hardware processor further executes the instructions to receive, using the network interface, the trained updated MLA from the cloud-based device. ### Claim 4 (depends on 1) The electronic device according to claim 1, wherein the data included in the notifications comprises: a source Internet Protocol (IP) address; a destination IP address; a communication protocol used to transmit the packet; a source port number; and a destination port number. ### Claim 5 (depends on 1) The electronic device according to claim 1, wherein the first service flow queue is for processing in a low latency service flow and the second service flow queue is for processing in a classic service flow. ### Claim 6 (depends on 1) The electronic device according to claim 1, wherein the hardware processor further executes the instructions to transmit, using the network interface, the notifications to a cloud-based device for forwarding to another electronic device. ### Claim 7 (depends on 1) The electronic device according to claim 1, wherein the hardware processor further executes the instructions to: receive, using the network interface, policy control information; and use the policy control information to process the at least one new packet. ### Claim 8 A method of enhancing classification of data packets in an electronic device comprising: receiving, by an electronic device, a packet to be processed using a first service flow queue or a second service flow queue, the first service flow queue comprising at least one service flow; determining, by the electronic device, whether a packet to be processed using the second service flow queue is incorrectly classified for processing using the first service flow queue based on a delay in service flow in the first service flow queue caused by the packet; moving the incorrectly classified packets from the first service flow queue to the second service flow queue; transmitting a notification for the incorrectly classified packet to a machine learning algorithm (MLA), each notification including data related to the incorrectly classified packet; training the MLA using the data in the notifications; and categorizing, using the trained MLA, at least one new packet to be processed using the first service flow queue or the second service flow queue by modifying packet header information, wherein the at least one new packet is processed based on the category determined by the MLA. ### Claim 9 (depends on 8) The method according to claim 8, further comprising transmitting, by the electronic device, the notifications to a cloud-based device for training an updated MLA. ### Claim 10 (depends on 9) The method according to claim 9 further comprising, receiving, by the electronic device, the trained updated MLA from the cloud-based device. ### Claim 11 (depends on 8) The method according to claim 8, wherein the data included in the notifications comprises: a source Internet Protocol (IP) address; a destination IP address; a communication protocol used to transmit the packet; a source port number; and a destination port number. ### Claim 12 (depends on 8) The method according to claim 8, wherein the first service flow queue is for processing in a low latency service flow and the second service flow queue is for processing in a classic service flow. ### Claim 13 (depends on 8) The method according to claim 8, further comprising transmitting the notifications to a cloud-based device for forwarding to another electronic device. ### Claim 14 (depends on 8) The method according to claim 8, further comprising: receiving policy control information; and using the policy control information to process the at least one packet. ### Claim 15 A non-transitory computer-readable recording medium in an electronic device for enhancing classification of data packets in a network, the non-transitory computer-readable recording medium storing instructions which when executed by a hardware processor cause the non-transitory recording medium to perform steps comprising: receiving a packet to be processed using a first service flow queue or a second service flow queue, the first service flow queue comprising at least one service flow; determining whether a packet to be processed using the second service flow queue is incorrectly classified for processing using the first service flow queue based on a delay in service flow in the first service flow queue caused by the packet; moving the incorrectly classified packets from the first service flow queue to the second service flow queue; transmitting a notification for the incorrectly classified packet to a machine learning algorithm (MLA), each notification including data related to the incorrectly classified packet; training the MLA using the data in the notifications; and categorizing, using the trained MLA, at least one new packet to be processed using the first service flow queue or the second service flow queue by modifying packet header information, wherein the at least one new packet is processed based on the category determined by the MLA. ### Claim 16 (depends on 15) The non-transitory computer-readable recording medium according to claim 15, further comprising transmitting the notifications to a cloud-based device for training an updated MLA. ### Claim 17 (depends on 16) The non-transitory computer-readable recording medium according to claim 16, further comprising receiving the trained updated MLA from the cloud-based device. ### Claim 18 (depends on 15) The non-transitory computer-readable recording medium according to claim 15, wherein the data included in the notifications comprises: a source Internet Protocol (IP) address; a destination IP address; a communication protocol used to transmit the packet; a source port number; and a destination port number. ### Claim 19 (depends on 15) The non-transitory computer-readable recording medium according to claim 15, wherein the first service flow queue is for processing in a low latency service flow and the second service flow queue is for processing in a classic service flow. ### Claim 20 (depends on 15) The non-transitory computer-readable recording medium according to claim 15, further comprising transmitting the notifications to a cloud-based device for forwarding to another electronic device.