Hughes Network Systems, LLC

Kaustubh Jain

Chi-Jiun Su

Nagesh Javali

Se Gi Hong

Sriram Vasudevan

Ramya Anusha Durvasula

Ramya Anuha Durvasula

Methods, systems, and apparatus, including computer programs encoded on computer-storage media, for traffic flow classification using machine learning. In some implementations, a communication device includes an anomaly detector comprising a machine learning model trained to predict whether data traffic patterns differ from a set of observed traffic patterns present in a set of training data. The communication device includes a traffic classifier comprising a machine learning model trained to predict a quality of service (QoS) class for network connections or data flows. The communication device is configured to evaluate network connections or data flows using the anomaly detector. The communication device may (i) use the traffic classifier to predict QoS classes for traffic that the anomaly detector predicts to be similar to the observed traffic patterns, and (ii) store data traffic that the anomaly detector predicts to be different from the observed traffic patterns.

Traffic flow classification using machine learning

A network monitoring method includes: monitoring, at a network node, one or more network statistics for each of one or more objects exchanged between a client end point and a server end point through a network connection, the network node being located between the client end point and the server end point; and determining one or more QoS metrics for each of the one or more objects based on the network statistics for respective one of the one or more objects.

Network monitoring method and network monitoring apparatus

Computer-implemented systems and methods for diagnosing and correcting connectivity problems in a network are provided. The method includes obtaining, at a network node disposed at an edge of the network, measurements of one or more attributes of network traffic exchanged between a content server and a client device via the network. The network traffic is associated with streaming video content transmitted from the content server to the client device over the network. The method also includes analyzing the one or more attributes of the network traffic to estimate Quality of Experience (QoE) performance metrics related to the streaming video content at the client device, and adjusting one or more network operating parameters of the network responsive to QoE performance metrics falling below a predetermined threshold.

Method for estimation of quality of experience (QoE) metrics for video streaming using passive measurements

A method wherein resource data for a web resource is generated can include age, confidence level, cacheability level, cache expiration, and/or priority level. Resource readiness for the web page is prepared based on the resource data. This can include preconnecting to a domain having the web resource when the resource data indicates a first level of resource readiness; preloading the web resource from the domain when the browser is open when the resource data indicates a second level of resource readiness, where the second level of resource readiness is greater than the first level of resource readiness; or prefetching the web resource from the domain prior to opening of the browser when the resource data indicates a third level of resource readiness, where the third level of resource readiness is greater than the second level of resource readiness.

Web acceleration via learning

Approaches for determining link conditions within a client network are provided. Network protocol statistics are recorded for network connections over the client network, wherein each connection provides data communications between a client terminal node and a different respective client device, and the client terminal node provides access to a broadband communications network for each of the different respective client device(s). Connection metrics are determined for each of the network connections, wherein each connection metric is determined based on the network protocol statistics associated with the respective network connection. Diagnoses of network conditions regarding connections over the client network that are of interest, wherein each diagnosis is performed based on an analysis of a respective set of the connection metrics for the respective network connection. Results of each diagnosis are provided to a system operator.

Determining link conditions of a client LAN/WAN from measurement point to client devices and application servers of interest

Systems and methods are disclosed, and one includes a computer-implemented method for predicting quality of experience (QoE) for application downloading a webpage from a server stack, including determining a stat data corresponding to a traffic through a network node, generating a feature vector using the stat data, estimating a download performance metric for the application in a downloading of the webpage, using feature vector data, and estimating a QoE value, using the estimated performance metric.

System and method for estimation of quality of experience (QoE) for web browsing using passive measurements

Application layer throughput (ALT) shaping is provided for a streaming media session is provided. A request for a content segment is received. The receipt time of the request is recorded and the request is forwarded to a content server. A response message is received from the server. The receipt time of the response and a segment size is determined. An ALT for the current segment is determined based on the receipt times of the request and response messages and the segment size. The ALT rate is compared to a target ALT rate required for a desired session PBR. When the ALT rate is higher than the target ALT rate, the content response is held for a time period determined to lower the current ALT rate to the target ALT rate, and then released to the client device. Otherwise, the content response is promptly provided to the client device.

Maximizing quality of service for QoS adaptive video streaming via dynamic application-layer throughput rate shaping

Automatically diagnosing operation of a communication network, including collecting, during multiple time periods spanning multiple days, a plurality of Layer-4 statistics for Layer-4 network connections on the network; determining, for each time period in the multiple time periods, a first Layer-4 condition metric for the time period based on the Layer-4 protocol statistics collected during the time period; determining a network performance indicator, based on a number of the time periods that a first threshold condition is met by the first Layer-4 condition metric; determining whether the communication network has experienced a performance issue based on the network performance indicator meeting a second threshold condition; and causing a diagnostic action, a corrective action, or a maintenance action to be performed based on the determination whether the communication network has experienced a performance issue.

Automated network diagnostic techniques

Characterizing transport protocol connections and controlling behavior of a communication network based on the characterizations, including obtaining transport protocol state information for transport protocol packets on a transport protocol connection; identifying request/response periods for the transport protocol connection based on the transport protocol state information, the identifying including identifying a beginning of a first request/response period based on the transport protocol state information for a first forward packet, and identifying an end of the first request/response period and a beginning of a second request/response period based on the transport protocol state information for a second forward packet; associating an application type with a server endpoint based on the identified request/response periods; and controlling behavior of the communication network based on the application type.

Application characterization using transport protocol analysis

Methods and systems of history-based instantaneous QoS class classification of traffic flow. QoS classifications are mapped to target destinations such that, when a first flow parcket packet of a traffic flow requesting access to a resource associated with the target destination, a classifier can instantaneously classify the traffic flow as belonging to a particular QoS class for servicing of the traffic flow.

History-based classification of traffic into QoS class with self-update

Automatically diagnosing operation of a communication network, including collecting, during a plurality of time periods, a plurality of Layer-4 and/or Layer-1/-2 statistics for network activity on the communication network; determining a network performance indicator, based on the statistics collected during a first time period included in the time periods and the statistics collected during a second time period included in the time periods and neither overlapping nor consecutive with the first time period; determining whether the communication network has experienced a performance issue based on the network performance indicator; and causing a diagnostic action, a corrective action, or a maintenance action to be performed based on the determination whether the communication network has experienced a performance issue.

Growing community of inventors

Germantown, MD, United States of America

Kaustubh Jain