Digital Commons Network^™

Intelligent Wireless Network Selection, Jiwoong Lee

Defensive Publications Series

WiFi network selection is typically based on the received signal strength of an access point (AP). However, the strongest signal does not necessarily lead to good user experience. For example, a strong signal or SSID may simultaneously attract many WiFi clients, causing congestion.

This disclosure utilizes machine learning models trained to intelligently select a wireless access point or an SSID based on multiple factors, e.g., neighboring APs, neighboring clients, historical service information, signal and interference levels, ping jitter, time-of-day, day-of-week, etc. Per the techniques, the selected AP is associated with a best overall score as determined by the machine learning …

Mode Selection For Augmented Reality Application To Control Access To Real-World Data, Klaus Weidner

Defensive Publications Series

In situations where virtual, mixed, augmented, and extended reality applications or services are operated by remote providers, images, video, and/or other world-understanding data captured by a camera or other sensors of a user device may be made available to the applications to enable integration of virtual objects with the user’s real world surroundings. It is important to allow users control over how the video generated by the camera is provided to such applications. Current application configuration settings such as permissions to access camera usually offer a binary choice that requires the user to provide the requested access or forego using …

Smart Equipment Kiosk, Justin Von Kennel

Defensive Publications Series

A smart kiosk for the dispensation, return, and swap of customer equipment is described. The smart kiosk includes lockers for storage of equipment and can be deployed to enable a convenient mode for resolution of customer issues related to equipment. The kiosk is deployed at high customer traffic locations with extended availability. Customer service workflows are designed to direct users to the kiosk. The workflows are designed to account for equipment inventory and customer requirements. Customer service agents provide information regarding kiosk locations, identifier information of lockers, and authentication information that enable customers to access the appropriate locker and obtain, …

Voice Interface Responses Based On Prior User Interactions, Minkoo Seo, Brian Spilsbury, Gadi Ben Amram, Joo Hyung Lee, Togan Gencoglu, Youngchan Park

Defensive Publications Series

Voice interfaces, such as those provided by virtual assistant applications on smartphones, smart speakers, and other devices utilize natural language when providing responses to user queries. Per techniques of this disclosure, user interaction patterns are used to generate natural language responses to queries posed over a voice user interface. With user permission and express consent, data that pertain to prior user interaction, e.g., text entry such as search terms entered; other input operations such as clicks, taps, swipes, mouse hovers; viewports for the user, etc. are obtained. For ease of interaction, such permission is obtained, e.g., at initial setup, and …

Identifying Network Faults In Data Centers Using A Covariance Matrix, Raghu Rajendra Arur, Nicholas Basker, Hasmit Grover

Defensive Publications Series

It may be difficult to identify root causes of protocol failures or degradations in application traffic performance due to errors. Embodiments presented herein identify the exact diagnostics needed to perform a root cause analysis. In particular, a covariance clustering methodology is used to generate correlated errors, which, when passed through a partial order set-based rule engine, are used to determine the root cause of a network fault due to protocol errors.

Recognizing A Natural Language Or Language Class In Source Code Files, Armijn Hemel

Defensive Publications Series

When doing analysis of source code archives from an unknown origin it can be helpful to find out where the code originated from geographically. Comments in these files can be helpful, as they are quite often written in the native natural language of the developer. Finding out which language the file is in can help understanding the flow of the code (example: translating comments) and provenance. By analyzing the contents of a file and seeing which character sets the contents belong to a better guess can be made.

Using Build Identifiers To Fingerprint Elf Binaries And Link To Build Information Without Having Access To Source Code, Armijn Hemel

Defensive Publications Series

Finding out where a software program or library comes from and how it was built without having direct access to the source code is not a trivial problem to solve. While versions of programs can be fairly accurately guessed this is a lot more difficult for build configuration. By comparing build identifiers from binaries of which nothing is known with build identifiers extracted from binaries for which source code and build information is available it is in certain cases possible to find out what source code and build information was used for a binary.

Pca (Printed Circuit Assembly) And Electronic Devices Cleaning Station, Joel J. Castillo

Defensive Publications Series

As part of the field/customer repair operation, PCA’s (Printed Circuit Assembly) are received as defective material in many cases having large amount of dust and/or contamination particles incrusted in them. This is due mainly to the years of operation where they receive little or no maintenance inside their products. When they are received as defective parts, the process is to thoroughly clean them in order for them to be able to be repaired and sold as spares parts.

Automatic Dating Of Videos Based On Matching Photos, Rainer Vincent Mager

Defensive Publications Series

This disclosure includes techniques to automatically date a video based on a matching photo. With user permission, the video is analyzed to determine an identity of a person and an identifying feature depicted in the video. Based on the identity of the person, a photo associated with the person in the video that includes the identifying feature is identified. Metadata is extracted from the image to determine a date of capture of the image. If the date of capture of the image is different from the date of capture of the video, a suggestion is provided to the user to …

Notification Output Via Watch Bezel, Aleksandr Rozman

Defensive Publications Series

Rather than requiring a user to view the face of a wearable computing device (e.g., a smartwatch), in order to ascertain information about a notification received at the wearable computing device, the wearable computing device may include a bezel with one or more lights, such as light emitting diodes (“LEDs”), which illuminate upon the wearable computing device receiving a notification. That is, a wearable computing device may include one or more LEDs that provide the user with an indication that the computing device has received notification data associated with one or more applications executing on the wearable computing device. In …

Controlling Access By Tagging Data, Juan Vasquez, Jesper Johansson

Defensive Publications Series

A customer (first party) uses an application developed by an application service provider (second party). A third party needs legitimate access to data of the customer stored on a server operated by the application service provider. In such a scenario, the third party requests the customer to permit the third party to access to such data. However, in some instances, the type of access requested by the third party is broad. This disclosure describes techniques to restrict the types of data that are made available to the third party by enabling a customer to tag data stored by the provider. …

Detection Of Counterfeit Silicon, Miguel Osorio Lozano, Sean Keys

Defensive Publications Series

It is important for manufacturers of silicon chips to guard against counterfeit chips, hardware trojans, exploits, royalty-avoiding mechanisms, and other malicious hardware injected into a production stream. The techniques of this disclosure create individualized chip signatures during the manufacturing process. The signatures can be used to detect counterfeit silicon or otherwise suspicious on-board componentry.

Challenge Keyword Generation For Voice-Based Authentication, Shuyang Chen, Nicholas Felker

Defensive Publications Series

This disclosure describes use of challenge keywords for use in voice authentication systems. Verification keywords are generated for a particular user and serve to verify the authenticity of the user. With user permission, keyword generation is performed based on analysis of previous voice-based interactions of the user. A uniqueness score is assigned to words uttered previously by the user in interactions with the voice interface. The uniqueness score for a word is determined based on a rarity of usage of the word in the user’s speech and a distinctiveness of user pronunciation of the word.

Advising Users In A Meeting What Was Last Said Before A Call Drop, Jonathan Rosenberg

Defensive Publications Series

Techniques are presented for informing a speaker what the speaker last said before the user dropped audio connection in a meeting with the speaker. The user may reconnect to the meeting, and the speaker may be provided this information using meeting transcripts. Packet loss indicators may be used to determine at what point the user was no longer able to make out the audio.

Security Event Timeline

Defensive Publications Series

This publication describes a security solution that incorporates security events in a single place and presents them in an easy-to-understand timeline. The intrinsic dynamic essence of the timeline enables the user to see what affects the overall security and offers guidance for the user to take steps to achieve a desired security state. The security solution shows the most-relevant security events at the highest level, but it also allows the user to see all of the security event details, if the user chooses to do so. The user can easily determine the time at which overall security became compromised and …

Service Layer Dependency Mapping Based Heatmap For Fault Isolation, Carlos M. Pignataro, Nagendra Kumar Nainar, Alejandro Eguiarte

Defensive Publications Series

Techniques are described herein for a two-fold process where a first Machine Learning (ML) model is engaged in building a Dependency Mapping Table (DMT) based on the data from the network (network observations). An Interior Gateway Protocol (IGP) database, service graph information, and in-band Operations, Administration and Management (iOAM) data is leveraged to feed into the ML model to build the DMT. The DMT is built with a per physical element precision (fiber optical cable in each direction). In other words, the database is built in a way to enable identification of the list of (connected/non-connected) services and protocols that …

Determining Region-Of-Interest Based On Motion Patterns, Ruiduo Yang

Defensive Publications Series

An auto-focus camera determines and focuses on a region-of-interest (RoI), e.g., faces, plates of food, flowers, etc. There is a slight lag, of the order of milliseconds, between the determination of the RoI and the capture of an image. During the time period of the lag, the subject of the photograph can sometimes move out of the just-determined RoI which can cause an off-target focus. This disclosure describes techniques to predict the RoI for a frame based on the RoIs of immediately-previous frames, the camera state, machine-learned patterns of motion, etc. The techniques enable robust and accurate auto-focus on the …

Image Recognition For Consumer Shopping Applications, Jibing Wang, Erik Stauffer, Aamir Akram

Defensive Publications Series

This disclosure utilizes image recognition techniques in consumer retail applications. Objects in an image are recognized and corresponding retail information, e.g., links to purchase an object, is provided to users. Image recognition techniques are integrated into point of sale applications and utilized for product identification based on a barcode, a QR code, based on the image of the product, etc. Application program interfaces are provided to link the image recognition application with payment processing and inventory management systems.

Intelligent Presentation Slide Assistant, Sammy El Ghazzal, Sandro Feuz

Defensive Publications Series

A technique is proposed for assisting a user with creating and polishing a document such as a presentation slide stored on a cloud-content based platform. A slide assistant service provided by the cloud-content based platform receives content to be included in a slide for a presentation. The slide assistant service can determine a level of seriousness of the presentation based on the content. Then, the slide assistant service can predict a theme appropriate for the presentation based on the level of seriousness and the content. Next, the slide assistant service can provide a slide that includes the slide objects arranged …

Process Of Improving Software Engineering Predictability With Business Benefits, Ravi Saxena, Atul Deshpande

Defensive Publications Series

A refined process to make software engineering predictable by associating bidding concept. This process helps in delivering software engineering products and solutions to customers with increased business values viz. customer satisfaction, better control over operating cost, profit margins and time to market.

This process greatly enriches decision making to enhance time to delivery, profit margin, knowledge sharing and keeping healthy collaboration among various geographically distributed teams within an organization.

Automatic Extraction And Provision Of Coupon Codes For Online Shopping, Matthew Sharifi

Defensive Publications Series

The disclosure describes techniques to automatically extract coupon codes from email, chat, SMS messages and suggest the use of available coupon codes when a user visits a website or other platform where the coupon codes are valid. Coupon codes are stored in a user account, e.g., on a server, and are synchronized to the user’s devices. User visits to shopping websites are detected, e.g., by a browser application, and corresponding coupon information is provided as a suggestion when the shopping website matches an available coupon. Expired coupons are automatically removed. User permission is obtained to extract coupon information and to …

Transmit Power Setting Based On Dynamic Sar, Qi Jiang, Paul Husted, Srenik Mehta, Nihar Jindal

Defensive Publications Series

Emitted power from consumer devices is required to meet a specific absorption rate (SAR) limit. Since SAR limits apply to the total transmitted power, mobile devices that have multiple radios limit transmit power on each radio based on the on/off status of other radios. For example, if cellular is enabled, then the maximum power available to WiFi is reduced. Such an approach does not account for the actual instantaneous power transmitted by a radio. For example, the cellular radio may be transmitting well below its maximum power when proximate to a base station. In such a case, the WiFi (or …

Turning Off Or Dimming A Device Screen Based On User Attention, Vidhya Navalpakkam, Tomer Shekel, Dmitry Lagun

Defensive Publications Series

Device screens are often set to turn off and/or dim automatically if no user interaction is detected for a specified amount of time. Turning off or dimming the screen saves power and prolongs the amount of time the device can operate without needing to recharge the battery. However, such timeout-based actions can result in false positives or negatives. With user permission, this disclosure utilizes contextual input of a user’s gaze and attention for management of the automatic turn off or dimming of the device screen. The techniques are applied to reduce the false positives and negatives and ensure that the …

Verified Revenue Sharing Between Distributor Oem And App Developers, John Dukellis

Defensive Publications Series

To increase the distribution of their apps, software developers often enter into agreements with mobile device original equipment manufacturers (OEMs) to distribute their apps, e.g., pre-installed on devices shipped by the OEM. OEMs control device software, including the home screen, and pre-install apps on devices upon upfront payment by the app developer. Smaller app developers may not be able to reach agreements with OEMs due to the cost of the sale, e.g., the upfront payment.

The techniques of this disclosure enable OEMs and app developers to reach agreements for the OEM to distribute an app for low or no upfront …

Smart Auto-Deactivation Of Device Airplane Mode, Gregory B. Roth

Defensive Publications Series

The transmitting and receiving capabilities of a device are turned off when it is put in airplane mode. As a result, a device in airplane mode is unable to transmit information about its location to an external entity, e.g., that tracks its whereabouts to help find the device when it is lost or stolen. This disclosure proposes techniques for smart auto-deactivation of airplane mode of a device without explicit user action. The auto-deactivation is based on one or more user-permitted signals, such as device sensor data, user’s calendar and context, information from airline sources, time, device battery charge, etc.

Pcie Riser Extension Assembly, Rick Salmonson, Troy Oxby, Larry Briski, Robert Normand, Russell Stacy, Jeffrey Glanzman

Defensive Publications Series

The PCIe Riser Extension provides geometric flexibility and the capability to expose all PCIe lanes from four processors on a single four socket motherboard to the chassis ends that may be a significant distance away, yet meet the electrical specifications required. The PCIe Riser Extension utilizes standard commercially available connectors and cabling to achieve this task.

Internet Of Things Construction Monitors, Thomas A. Stern

Defensive Publications Series

The present disclosure relates to a monitoring system and related method for determining the state of a structure, determining when a structural failure is apt to occur, and sending notifications associated with the state of a structure to external computing devices. The monitoring system can include a computing system that receives state data including information associated with the state of a structure. The state data can be used to determine when the structure is apt to fail. Further, alert notifications can be generated and sent to external computing devices and services when such failures are determined to be likely.

Connectivity Assessment Survey For Technology Emergency Response (Caster), Rakesh Bharania, Chandler Johnson, Dustin Li

Defensive Publications Series

Disasters such as hurricanes, earthquakes, and tsunamis often cause widespread communication outages, and there is a need to quickly assess which areas have been impacted so that teams such as Tactical Operations can optimize the delivery and installation of emergency communications infrastructure. By placing a software defined radio (SDR) in low earth orbit on a satellite and processing the data to geolocate radio frequency (RF) signals, heat maps may be generated depicting "ground truth" RF changes, providing timely and accurate data to inform decision-making for emergency communications deployments. This, in turn, will serve as a force multiplier, amplifying the positive …

A Novel Approach To Measure Code Coverage For A Typical Os Deployment Tool, Debdipta Ghosh

Defensive Publications Series

Code Coverage analysis measures how much of the target code executed during a test. This coverage measurement is used to describe the degree to which the source code of a program is tested by a particular test suite. HP-UX use Bullseye, a code coverage tool which is C/C++ code coverage analyzer used to improve the quality of software in vital systems. This paper will discuss why using bullseye is challenging while finding code coverage of HP-UX binaries during cold installation and how those issues are being addressed through this solution.

Item Recommendations For Cache And Synchronization Of Application Stores, Wei Chai, Zhiwei Gu, Sundeep Sancheti, Yilan Liu

Defensive Publications Series

To reduce the latency of rendering the page of an app in an application store (e.g., that provides access to downloads of apps to a mobile device), the application store app periodically synchronizes with its backend and caches on the user mobile device app pages that a user is likely to visit.

This disclosure uses deep learning models to predict the next N apps that a user is likely to visit based on permitted user activity signals, e.g., demographic information, historical visits to deep links associated with the app, etc.