# US 2024/0187320 A1 · Estimating Quality Metric for Latency Sensitive Traffic Flows in Communication Networks

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2024/0187320 A1
- Document ID: US20240187320A1
- Status: Published application
- Kind: A1
- Application number: 18/437,655
- Filed: 2024-02-09
- Published: 2024-06-06
- Earliest claimed: 2019-06-30
- Assignees: Telefonaktiebolaget LM Ericsson (publ)
- Applicants: Telefonaktiebolaget LM Ericsson (publ)
- Inventors: Göran Eriksson; Magnus Westerlund; Stefan Håkansson; Gunnar Heikkilä
- Web page: https://patentgrep.com/patent/US20240187320A1

## CPC

- H04L 65/80: Responding to QoS
- H04L 41/147: for predicting network behaviour
- H04L 41/16: using machine learning or artificial intelligence
- H04L 41/5067: Customer-centric QoS measurements
- H04L 43/026: using flow identification
- H04L 43/0864: Round trip delays
- H04L 43/0876: Network utilisation, e.g. volume of load or congestion level
- H04L 43/0888: Throughput
- H04L 43/0894: Packet rate
- H04L 43/10: Active monitoring, e.g. heartbeat, ping or trace-route
- H04L 65/1083: In-session procedures
- H04L 65/61: for supporting one-way streaming services, e.g. Internet radio
- H04L 41/145: {involving simulating, designing, planning or modelling of a network}
- H04L 43/0829: Packet loss
- H04L 43/0852: Delays
- H04L 43/087: Jitter
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

An apparatus in a mobile communication network combines information from monitoring IP flows carrying latency sensitive content passing the apparatus and information about the application behavior and target Quality of Experience (QoE) or target connectivity characteristics such as Quality of Service (QoS) from the application to provide ongoing predictions of QoE/QoS. In some cases, the apparatus exploits a probe on a device to generate traffic for learning flow characteristics not obtained from monitoring application IP flows in the network. Embodiments disclosed herein can be used to predict quality metrics for many applications where jitter/latency is a factor affecting perceived quality, such as QoE for a human consumer or QoS for machine type communications. The embodiments are applicable to the analysis of traffic carrying conversational speech.

## Claims

### Claim 1

A method, in a mobile communication network, of estimating a quality metric for a packet flow associated with an application and carrying latency sensitive content, the method characterized by:
obtaining one or more algorithms for estimating connectivity characteristic variations and the quality metric for the packet flow; and
monitoring network traffic;
classifying packets belonging to the packet flow;
analyzing network traffic parameters for the packets belonging to the packet flow;
based on the obtained algorithm for estimating connectivity characteristic variations, predicting connectivity characteristic variations for the packet flow;
based on the obtained quality metric algorithm and the predicted connectivity characteristic variations, predicting a quality metric for the packet flow; and
reporting the predicted quality metric.

### Claim 2 (depends on 1)

The method of claim 1, wherein the predicted connectivity characteristic variations include late loss from packet jitter.

### Claim 3

A network node operative in a mobile communication network, and operative to estimate a quality metric for a packet flow associated with an application and carrying latency sensitive content, the network node characterized by:
communication circuitry; and
processing circuitry operatively coupled to the communication circuitry and adapted to:
obtain one or more algorithms for estimating connectivity characteristic variations and the quality metric for the packet flow; and
monitor network traffic;
classify packets belonging to the packet flow;
analyze network traffic parameters for the packets belonging to the packet flow;
based on the obtained algorithm for estimating connectivity characteristic variations, predict connectivity characteristic variations for the packet flow;
based on the obtained quality metric algorithm and the predicted connectivity characteristic variations, predict a quality metric for the packet flow; and
report the predicted quality metric.

### Claim 4 (depends on 3)

The network node of claim 3, wherein the predicted connectivity characteristic variations include late loss from packet jitter.

### Claim 5 (depends on 3)

The network node of claim 3 wherein the latency sensitive content is voice, and wherein the quality metric is a Quality of Experience (QoE).

### Claim 6 (depends on 3)

The network node of claim 3 wherein the processing circuitry is further operative to perform the monitoring, classifying, analyzing, predicting, predicting, and reporting steps iteratively over two or more of a plurality of monitoring periods.

### Claim 7 (depends on 3)

The network node of claim 3 wherein the processing circuitry is operative to obtain one or more algorithms for estimating connectivity characteristic variations and quality metric for the packet flow by obtaining the algorithms from an Operations and Maintenance (OAM) function in the mobile network.

### Claim 8 (depends on 7)

The network node of claim 7 wherein the processing circuitry is operative to obtain the algorithms from an OAM function in the mobile network by obtaining the algorithms from the application as part of a Service Level Agreement (SLA) between the application and the network.

### Claim 9 (depends on 8)

The network node of claim 8 wherein the algorithms for estimating connectivity characteristic variations describes the maximum allowed jitter before a decoder will consider a data frame lost.

### Claim 10 (depends on 8)

The network node of claim 8 wherein the quality metric algorithm specifies the criteria for connectivity characteristic variation statistics and characteristics that is allowed during a specified duration to maintain a determined quality level for one session.

### Claim 11 (depends on 10)

The network node of claim 10 wherein the quality metric algorithm further specifies one or more mappings between quality levels and network latency or throughput.

### Claim 12 (depends on 10)

The network node of claim 10 wherein the processing circuitry is further characterized by being adapted to obtain from the application packet flow information that facilitates the identification of packet flows carrying latency sensitive content.

### Claim 13 (depends on 3)

The network node of claim 3 wherein the processing circuitry is adapted to obtain one or more algorithms for estimating connectivity characteristic variations and quality metric for the packet flow by training one or more machine learning functions based on network traffic metrics including measured latency, throughput, and packet loss.

### Claim 14 (depends on 13)

The network node of claim 13 wherein the training is based on feedback from the application of the connectivity characteristic variations and quality metric as calculated by the application.

### Claim 15 (depends on 14)

The network node of claim 14 wherein the packet flow comprises a QUIC transport protocol connection, and wherein training of a machine learning function for algorithm for estimating connectivity characteristic variations is based on QUIC packet characteristics and unencrypted header information, including one or more of QUIC Connection ID, spin bit, and IP header information.

### Claim 16 (depends on 14)

The network node of claim 14 wherein the application has provided decryption information, and wherein training of the machine learning function for the algorithm for estimating connectivity characteristic variations is further based on information in encrypted headers.

### Claim 17 (depends on 16)

The network node of claim 16 wherein the processing circuitry is further characterized by being adapted to execute a trusted QUIC proxy operative to at least read encrypted QUIC header information in the packet flow, based on the received decryption information.

### Claim 18 (depends on 3)

The network node of claim 3 wherein the network node comprises a Network Data Analytics Function (NWDAF) and wherein the packet flow is generated by a Web Real-Time Communication (WebRTC) speech connection traversing the mobile network.

### Claim 19 (depends on 18)

The network node of claim 18 wherein the NWDAF receives, from a device in the WebRTC connection, end point observed statistics of connectivity characteristic variations, including one or more of late loss, throughput, Round Trip Time (RTT) packet loss, and quality metric.

### Claim 20 (depends on 3)

The network node of claim 3 wherein the processing circuitry is adapted to analyze network traffic parameters for the packets belonging to the packet flow by, where one or more desired network traffic parameters are not available or visible in the packet flow:
establishing side car network traffic between a probe server and a probe application on a device, the side car network traffic sharing at least part of the packet flow's path through the mobile network;
monitoring network traffic parameters exposed in the side car network traffic that are not available or visible in the packet flow; and
analyzing network traffic parameters for the packets belonging to the packet flow by analogy to the network traffic parameters in the side car network traffic.

### Claim 21 (depends on 20)

The network node of claim 20 wherein network traffic parameters exposed in the side car network traffic that are not available or visible in the packet flow include a QUIC spin bit, enabling analysis of downstream Round Trip Timing (RTT).

### Claim 22 (depends on 20)

The network node of claim 20 wherein network traffic parameters exposed in the side car network traffic that are not available or visible in the packet flow include packet sequence, enabling analysis of packet loss and reordering.
