T-mobile Usa, Inc.

T-mobile Innovations LLC

Antoine T Tran

Jie Hui

Emile Minh Tran

Doru Culiac

Jong Sung Yoon

Kevin Hinson

Narciso Faustino

Bryan Yang

Yasmin Karimli

Daniel Howard Wilson

Pablo Tapia

Thomas P Lucht

Jeffery A Smith

Kranthi Sontineni

Julio Costa

Nan Xiao

Carl Williams

Himesh Bagley

Morgan MacKaay

Muhammad Khadim

Tony Fung

Chong Kiong

Michael J Mitchell

Vikas Ranjan

Dennis Jarret Guill, Jr

Yunhan Jia

Julio C Costa

Z Morley Mao

Syed Umair Ahmed

Alexander Anh Tran

Ryan Cyrus Khamneian

Yihua Guo

Alex Yoon

Jeremy M Tran

A computer-implemented method generates an estimate of a real-time distribution density of wireless devices in a geographic area. The method includes receiving real-time network activity data collected at cell sites. The method includes processing the real-time network activity data with an ML model. The ML model is trained based on non-real-time network activity data indicating network activities as well as location data of wireless devices served by the cell sites. The ML model is configured to predict a real-time distribution of wireless devices in the geographic area at different points in time. The method includes predicting a real-time distribution of the wireless devices based on a pattern of network activities and locations of wireless devices. The method also includes generating an estimate of a distribution density of the wireless devices in the geographic area based on the predicted real-time distribution and adjusted by the real-time count of network activities.

Dynamic estimation of real-time distribution density of wireless devices using machine learning models

Techniques for locating a user device in an indoor environment include receiving an indication of an acceleration event at a start position and a deceleration event associated with the user device. Based at least on the amount of time elapsed between the acceleration event and the deceleration event, a traveled distance of the user device may be determined. Additionally, sensor data collected from the user device between the acceleration event and the deceleration event may be used to determine the direction of travel of the user device. Following the determination of the traveled distance and the direction of travel, a finish position relative to the start position of the user device in the indoor environment may be mapped using an architectural plan associated with the indoor environment.

Indoor localization based on acceleration events

A method for determining a reliability of a wireless device to estimate its location under a simulated environmental condition includes simulating an environmental condition inside a test chamber by controlling a physical parameter. The test chamber contains the wireless user device configured to estimate a location of the wireless user device based on reference signals received from a signal source. The method includes communicating a pattern of reference signals having varying signal propagation characteristics in the test chamber. The method includes receiving an indication of an estimated location calculated by the wireless user device based on the pattern of reference signals. The reliability of the wireless device to estimate its location under the simulated environmental condition is determined based on a comparison of the estimated location with the simulated location.

Location simulation for wireless devices

Techniques are disclosed for locating a user device in an indoor environment such as a building based at least on a building topology model using one or more Internet of Things (IoT) devices. Certain aspects of the techniques include detecting a presence of a user device in communication with an IoT device in a building, wherein the IoT device is associated with device attributes. The building is identified based at least on a building topology model that is associated with the building and the device attributes. The location of the IoT device is determined based at least on the building topology model. The user device is located relative to the location of the IoT device.

Localization using internet of things devices

Techniques for calculating a location of a position consumer device is disclosed. In one example, a network server may create a fingerprint map from reference data points. Each of the reference data points may include a recorded geo-location of a position source device and signal measurements taken at that recorded geo-location. By initially estimating an initial position of the position consumer device, the network server may apply one or more threshold values to filter reference data points—candidates for interpolation. The network server may then perform an interpolation on one or more pairs of reference data points to find a pair of reference data points that is collinear with the estimated position of the position consumer device. The location of the position consumer device may then be calculated based upon geo-locations of position source devices that are associated with collinear reference data points.

Device position accuracy with network-based crowdsourcing

In some examples, a terminal can establish wireless communication with a base station. The terminal can determine a challenge, transmit the challenge, receive a response, and determine that the response is valid. The terminal can, in response, establish a secure network tunnel to a network node. In some examples, a terminal can determine a first communication parameter associated with communication with the base station. The terminal can receive data indicating a second communication parameter via a secure network tunnel. The terminal can determine that the communication parameters do not match, and, in response, provide an indication that an attack is under way against the network terminal. Some example terminals transmit a challenge, determine a response status associated with the challenge, and determine that an attack is under way based on the response status.

Wireless-network attack detection

Techniques are described for locating user devices in an indoor environment based at least on a building topology model. The techniques include detecting a presence of a user device in a building. The building may include ingress and egress points connecting defined spaces within the building. The locations of the individual ingress and egress points are identified based at least on the movement data of the user device. A building topology model may be generated using the locations of the individual ingress and egress points. To locate the user device at a given timestamp, one or more of the ingress and egress points passed through by the user device may be identified. The location of the user device may be determined based at least on the locations of the one or more ingress and egress points passed through by the user device mapped to the building topology model.

Constrained user device location using building topology

Described herein are techniques, devices, and systems for training a machine learning model(s) and/or artificial intelligence algorithm(s) to determine where a mobile device (and, hence, a user of the mobile device) is located based on audio data associated with the mobile device and/or contextual data associated with the mobile device. The machine learning techniques may be used to determine contextual information about users, such as determining that a particular location is likely to be a user's home, office, or the like, based on movement patterns exhibited in the data associated with a user's mobile device. Once trained, the machine learning model(s) is usable to classify a mobile device as having been located at one of multiple candidate locations, such as indoors or outdoors, at a particular time. The described techniques can improve the accuracy of determining a mobile device's location, among other technical benefits.

Using machine learning to locate mobile device

Techniques are described for filtering location data in a dataset based at least on its representation of the actual location of a user device. In some embodiments, sensor data including location data associated with a user device located in an environment may be used to identify one or more prediction factors. The one or more prediction factors may be associated to embed a user behavior vector of a user based at least on the associations. The user behavior vector may be a representation of an environment class in which the environment may be classified. The location data is associated with the environment class.

High fidelity geolocation using machine learning

Techniques are described herein for developing a fingerprint map that may be used for 3D indoor localization. In one example, a network server may leverage a building footprint from an open source database with signal measurements taken by probing user devices from signal sources such as access point (AP) devices. The network server may use the signal measurements to remotely calculate corresponding 3D positions of the AP devices in a particular building. Further, the network server may use the building footprint and the calculated 3D positions of the AP devices as references for developing the fingerprint map for 3D indoor localization.

Developing a fingerprint map for determining an indoor location of a wireless device

A telemetry data computing engine may receive the telemetry data of a user device, via a network, and it may apply a machine learning algorithm to at least one telemetry data and generate a predicted location for the user device. The predicted location may be used to generate a location pattern for the user device and a user device profile for the user device. The user device profile may be routed to the wireless carrier core network.

Profiling location information and network traffic density from telemetry data

Techniques directed to utilizing networked devices to enhance user experience are described. In an example, sensor data received from sensor(s) associated with a facility can be used to determine a presence of a user equipment (UE) within or proximate to the facility. Based at least in part on the sensor data, an account associated with the UE can be accessed. Based at least in part on account data associated with the account, a distribution of network technology can be modified to temporarily enable the UE to access a service that is not otherwise available to at least one other UE within or proximate to the facility.

Enhanced algorithms for improved user experience using internet of things sensor integration

A method for a base station in communication with a user device through a wireless communications network is described. The method includes communicating with a user device on a current frequency through a wireless communications network. The method may also include determining, based at least in part on a model, a probability that a higher capacity may be acquired by the user device on a frequency other than the current frequency; predicting a first recommended frequency for the user device that is likely to have a higher capacity than the current frequency; instructing the user device to scan the first recommended frequency; receiving an indication from the user device that the first recommended frequency has a higher capacity than the current frequency; and communicating with the user device on the first recommended frequency.

Predictive model for high-capacity frequencies in mobile broadband

User equipment (UE) can include a spectrum analyzer to monitor characteristics of transmission channels. The user equipment can monitor a 600 MHz spectrum and associated channels, for example, to determine if the spectrum is free of interference or is currently occupied. The UE can analyze a received signal strength indication (RSSI), a reference signal received power (RSRP), a reference signal received quality (RSRQ), and signal-to-interference-plus-noise ratio (SINR), for example, to distinguish between types of interference if a channel is occupied. User equipment (UE) can aggregate data and report such data to a network device further aggregate the data and to generate reports. Network components can be deployed or optimized based at least in part on network metrics provided by individual UEs or aggregated data provided by a plurality of UEs. In some instances, the UE can be a mobile phone of a customer to gather metrics in a distributed manner.

Growing community of inventors

Bellevue, WA, United States of America

Antoine T Tran