Hughes Network Systems, LLC

Rajeev Kubba

Shivaram Sitaram

Ying Wang

Hongyangyang Shi

Junyu Gong

Will Highfield

An apparatus and method for predicting the impact of various weather conditions on signal quality in a satellite communication system. A first model is generated for predicting RF signal degradation, and trained with link quality metrics from selected terminals over a first time interval. A second model is generated for predicting RF signal degradation resulting from weather conditions, and trained with both the link quality metrics from the selected terminals and weather conditions over a second predetermined time interval. Real time link quality metrics are analyzed using the trained first model and the trained second model in order to identify changes in current RF signal quality. At least one recommendation is subsequently generated to indicate a cause for changes in the current RF signal quality.

Weather impact modeling of satellite based services

A method and system for service management of a complex network including: computing, at a computer, a weather impact score for geographic areas within a coverage area of a satellite; predicting, based on the weather impact score for each of the geographic areas, a degradation of at least one of the satellite links serving a respective geographic area; and sending a notification about the degradation. The method may include calculating, with a computer, a peak Quality of Service (QoS) for each of the satellite links; aggregating, for a duration, transmission errors to calculate an actual QoS for each of the satellite links; and displaying a drill-down dashboard comprising a color-code for each of the satellite links, wherein the color-code corresponds to a severity of a respective discrepancy between a respective peak QoS and a respective actual QoS of a respective satellite link.

Adaptive rule based engine for QoS computations of internet services over satellite

A system and method for identifying devices with WiFi connection incompatibility are described. Data related to connectivity for at least one customer premise equipment (CPE) connected to each terminal in a satellite network is collected in order to generate a set of metrics associated with each of the at least CPE using the collected data. A first set of results indicative of network incompatibility is determined for each of the CPEs. Next, a second set of results indicative of network incompatibility is determined for each CPE using at least one machine learning algorithm. Information indicative of network incompatibility is subsequently generated for each CPE based on the first set of results and the second set of results.

WiFi device compatibility analyzer for satellite networks

An apparatus and method for optimizing selection of data cap limited service plans. A model is created in order to predict bandwidth usage by existing subscribers in a satellite communication system. The model is trained with usage data for all subscribers of the satellite communication system over a predetermined time interval, and used to analyze usage patterns of each subscriber. Bandwidth usage is predicted for each subscriber relative to an existing service plan based, at least one recommendation is generated for optimizing use of bandwidth in the satellite communication system based on the analysis and predicted bandwidth usage.

Optimizing data cap limited service plans

An apparatus and method for data cap management of users exceeding or violating bandwidth policies in a communication system. A model is created for identifying super heavy users, and trained usage data for all current users of the communication system for a predetermined time interval. Usage patterns of each user are analyzed over a second duration using the trained model in order to identify current super heavy users based, at least in part, on the analysis. Traffic flow control restrictions are then applied to some or all of the traffic associated with identified super heavy users.

Data cap management of super heavy users based on monthly usage and service plan limits

A system and method for reducing waste of a network resource is disclosed. The method including: providing a group of active subscribers; determining an underutilization level of the network resource for an upcoming allocation interval; calculating a throttle and a resource weight for the group of active subscribers to decrease the underutilization level; allocating the network resource based on the throttle and the resource weight; and adjusting, based on a feedback underutilization level, the throttle and the resource weight. In some embodiments, the throttle is based on congestion metrics including measuring available channel capacity, a latency, a queue depth, a count of subscribers in an outroute channel or the like. In some embodiments, the active subscribers may include under subscribers, over subscribers and premium subscribers, wherein each of the over subscribers have exceeded a respective network resource usage allowance for a respective subscription interval.

Growing community of inventors

Germantown, MD, United States of America

Rajeev Kubba