Digital Commons Network^™

Control Plane Convergence Measurement In Sd-Wan Fabric, Eugene Khabarov

Defensive Publications Series

In a modern enterprise network, particularly in large-scale software-defined wide area network (SD-WAN) overlays, there is a constant need to measure convergence, also known as route propagation time, to ensure that a network may react quickly to failures and steer traffic over alternative or backup paths in such a manner that business applications and users experience minimal disruption. To address such a need, techniques are presented herein that support a network-wide technique to automatically and granularly measure control plane convergence time across a SD-WAN fabric, without manual command-line interface (CLI) intervention on a router-by-router basis, and report the results using …

Use Of Language Models To Improve Automatic Speech Recognition, Pawel Janus, Agoston Weisz, Aurelien Boffy, Miroslaw Michalski

Defensive Publications Series

Speech recognition enables users to interact with devices via their voice. However, errors in speech recognition during a user’s interaction with such devices can be problematic and lead to a less than satisfactory user experience. This disclosure describes the use of language modeling to recover from automatic speech recognition (ASR) errors by identifying broken queries. The full natural language understanding (NLU) stack is executed to obtain a coherent, alternative, speech recognition. The alternative recognition (or query) runs in parallel to the original, misrecognized query. The potential actions triggered by the misrecognized and the NLU-augmented queries are compared to pick the …

Need-Based Dynamic Reconfiguration Of Network Slice Using A Floating Slice And A Cross-Domain Orchestrator, Soundarya Pendyala, Rajaneesh Shetty, Krushang Shah, John Mullooly

Defensive Publications Series

A need exists for the ability to dynamically reconfigure a network slice from one type to another type based on a business case. To address such a need, techniques are presented herein that support the dynamic and the automatic reconfiguration of a network slice from one type to another type. Aspects of the presented techniques encompass an End-to-End floating network slice with a generic slice template, trigger-based algorithms to change a network slice from one type to another, segment routing (SR) tagging-based dynamic network slice management at the transport network level, the implementation of an SR-aware core network, adaptations to …

Distinguishing Different Media Types On Scanned Images Using Machine Learning To Automate Media Selection, Hp Inc

Defensive Publications Series

When using a printer, one of the most frequent operations is to load media onto which the printing will

occur.

This operation not only consist in physically loading and placing the media in the paper path for the

printer to feed the media, but also consists in providing some media identification to the printer which

then will be used to select the proper parameters for optimal printing.

Making a mistake in the identification of the media, something with can occur for new operators or

these who do not often use the printer, will lead to a non‐optimal printing image quality. …

Load Balancing In Arbitrarily Clustered Microservice Architectures, Anonymous

Defensive Publications Series

The present disclosure relates to load balancing in arbitrarily clustered microservice architectures. In cases where the affinity and load-balancing of the servers cannot be controlled or known in advance due to the hierarchical nature of the clustering architecture, measuring typical response times and sorting so the more expensive work is done first (as described in the section above) is a good way for the client or the cluster service to be able to ensure queues and load balancing is as efficient as possible. the present invention allows for arbitrarily deep hierarchies of clustering running with any combination of processors, cores, …

Interior Gateway Protocol Path Computation Based On The Maximum Transmission Unit For A Link, Geetanjalli Bhalla, Manjari Vishnoi, Navya Kamath, Gayatri Bulla, Mohsin Alam

Defensive Publications Series

Fragmentation or packet drops can occur for data packets at any node in the path of packets based on the maximum transmission unit (MTU) of the links that constitute the path. This path can be a tunneled path, which is formed over multiple nodes based on the constraints advertised by Interior Gateway Protocol (IGP) of those nodes. There are certain flows over a tunneled path where fragmentation and dropped packets can cause multiple issues. Provided herein is a proposal to extend the IGP extended link Type-Length-Value (TLV) object to carry information identifying the MTU of a link, which can be …

Device Failure Data Collection Using Qr Codes, N Shalini Srinivasan, Rashmi Nikhil Patil, Andriani Stylianou, Jetsy Mathew, Joel Obstfeld

Defensive Publications Series

During the lifetime of a device, a customer may encounter a number of errors that require debugging. The process of diagnosing a device can become complicated when network access is not possible. To address such challenges, techniques are presented herein that support obtaining diagnostic information from a device in cases where network access cannot be provided through the use of dynamically generated Uniform Resource Locators (URLs) and quick response (QR) codes. Aspects of the presented techniques employ independent URL and QR code generating logic, in a hardware device, that has its own power source, memory, etc. (to allow for that …

Simulation Of Personal Area Network Devices, Erwin Jansen, Myles Watson, Paris Hsu

Defensive Publications Series

Developers of applications (which may be referred to as “apps”) for various devices (including smartphones, smartwatches, smart glasses, extended reality – XR – headsets, virtual reality – VR – headsets, augmented reality – AR – headsets, gaming systems, streaming gaming systems, desktop computers, laptop computers, tablet computers, televisions, etc.) may often interface with personal area network (PAN) devices via potentially complex PAN protocols. Example PAN protocols include Bluetooth® protocols, ultra wideband (UWB) protocols, Institute of Electrical and Electronics Engineers (IEEE) 802.15 wireless PAN protocols, Z-Wave® protocols, Zigbee® protocols, etc. Rather than develop and test such apps with respect to physical …

Application-Aware Traffic Engineering With Segment Routing, Shakeil Chandler, Rushikesh Jagdale

Defensive Publications Series

Techniques are presented herein that enhance segment routing (SR) traffic engineering (TE) through, among other things, the incorporation of a network protocol facility. Such a facility collects active Internet Protocol (IP) network traffic, as it flows in to or out of an interface, and then analyzes the collected data to develop an understanding of network traffic particulars such as, for example, flow, volume, etc. Aspects of the presented techniques leverage a Path Computation Element (PCE) for path calculations and support the ability to create multiple unique service-level agreement (SLA) tunnels, based on application flows or Layer 7 information within a …

Memory Foam Tool To Measure Powder To Printhead Spacing (Pps) Without Powder, Hp Inc

Defensive Publications Series

Powder to Printhead Spacing (commonly known as PPS) is a key parameter in all 3D Printers based in Jet

Fusion technology, like the Plastics and Metals ones produced by HP. PPS is the distance between the top

surface of the powder in the print bed of the Build Unit and the Printheads that spit the agents to fuse or

agglomerate this powder into printed parts.

This disclosure describes a system to measure PPS without using powder in a fast and reliable way,

requiring just two tools:

 A special Printhead with a laser distance measuring system in the place of …

Graph-Based Feature-Serving For Privacy-Aware Machine Learning, Anonymous

Defensive Publications Series

This disclosure describes graph-based privacy-aware platforms for serving features to machine learning models, optimized to meet constraints of privacy, model quality, multi-granularity, incremental update, overlapping features, scale, etc. Privacy is preserved by grouping features into a collection, subject to lower-bound constraints, such that features corresponding to individual users are rendered unrecoverable. Such feature aggregation does not prevent generation of personalized recommendations, as model quality is maintained by aggregating similar, rather than arbitrary features, subject to upper-bound constraints. Upper and lower bounds can vary to account for differing privacy sensitivities and feature importance with respect to a given model (multi-granularity). Features …

Systems, Methods And Apparatuses For Providing A Personal Carbon Footprint Tracker

Defensive Publications Series

No abstract provided.

Context-Aware Screen Sharing For Online Meetings, Sebastian Jeuk, Gonzalo Salgueiro, David Hanes

Defensive Publications Series

Currently, in most online meeting and collaboration systems, platforms, applications, etc. (which, for convenience, may be referred to herein as a collaboration environment) sharing is typically limited to an application or a whole desktop. The next level of evolution for sharing is for a presenter to be able to intelligently identify contextually-relevant subsections of an application, screen, environment, etc. and share that selection with other attendees. Techniques are presented herein that support an intelligent and context-aware sharing model that take the known feature of selective sharing to that next level. Aspects of the presented techniques support an advanced capability to …

Techniques To Optimize Software-Defined Access Layer 2 Broadcast, Unknown-Unicast And Multicast (Bum) Use-Cases With Bit Index Explicit Replication (Bier), Denis Neogi

Defensive Publications Series

Current implementations involving Broadcast, Unknown-Unicast & Multicast (BUM) traffic in a Software-Defined Access (SDA) fabric typically rely on a traditional multicast model in which each Fabric-Edge device joins a shared Internet Protocol (IP) multicast-group having a pre-defined Multicast Rendezvous Point (RP) node or nodes— effectively establishing one large broadcast domain for all Layer 2 (L2) instances in the fabric. Techniques presented herein provide for the ability to establish a model for assigning shared groups of Virtual Extensible Local Area Network (VxLAN) Virtual Network Instances (VNIs) logically and programmatically on a Fabric-Edge device to a dedicated Bit Index Explicit Replication (BIER) …

Co-Packaged Optics Maximum Rate Allocation Algorithm, Marco Mazzini, Alberto Cervasio

Defensive Publications Series

With the introduction of co-packaged optics (CPO)-based hosts, systems can be designed in a manner that is more flexible than conventional systems. It is possible to view the CPO as an electro-optical matrix rather than a battery of pluggable modules as with conventional hosts. Such an approach provides the opportunity to design CPO-based switches for flexibility, protection, and resilience up to 102.4 Terabits or beyond. Pluggable solutions, in addition to being more expensive, may not have such capabilities. Leveraging such opportunities, this proposal provides new techniques for defining the connectivity between CPO hosts in a manner that allows for maximum …

Preserving Manual Modifications To Auto-Generated Code Across Versions, Patrick Blesi

Defensive Publications Series

When an automatic code generator updates code that has been manually modified, the manual edits get overwritten and effectively lost. This is a problem for test and development engineers. In such situations, the user (test/development engineer) must laboriously re-apply the changes they made manually for each auto-generated update. This disclosure describes techniques that preserve manual edits as auto-generated code undergoes automatic updates. The developer inserts sentinel values into the code, indicating lines that have been manually inserted or deleted. During automatic update, the sentinel values are used to recover the base file. An automatic code generator automatically generates a new …

Optimized Multi-Homing In Inter-As Option B For Ethernet Virtual Private Networks, Satya Mohanty, Mankamana Mishra, Ali Sajassi

Defensive Publications Series

It is becoming increasingly common to stretch Ethernet Virtual Private Network (EVPN) services across Inter-Autonomous System (Inter-AS) boundaries. In such Inter-AS deployments, providers do not like to expose loopbacks of a provider edge (PE) router's loopbacks in one AS to the other AS, since this typically ensures less maintenance overhead and also allays security concerns. Inter-AS Option B for EVPN is ideally suited for such deployments as the next-hops of the EVPN routes for Inter-AS Option B are reset at the AS Border Router (ASBR), so the receiving AS does not know the loopback addresses of an advertising AS. This …

Artificial Intelligence Based Creative Companion For Content Creation, Anonymous

Defensive Publications Series

A creative companion utilizing artificial intelligence (AI) based mechanisms to generate and optimize customized outreach plans is described herein. The creative companion receives an original content item from a client and generates a new content item based on the original content item. The new content item may be based on a template and may be associated with automatically generated descriptive language. Further, a persona associated with the new content item may be generated to provide information relating to the new content item and associate the client with a third party to enable a modification to the new content item.

Data Pipelines As A Web Service, Jan-Simon Pendry

Defensive Publications Series

This disclosure describes techniques that leverage the mechanism of pipeline-designer tools to create web services. Whereas web services are typically written as software code in conventional programming languages, the described techniques deploy dataflow design tooling and workflows to provide web services, in the form of a REST API or as a RPC endpoint, without requiring the user to write code. The techniques herein described are referred to as pipeline-as-a-service.

Random Query Generator For Database Testing, Adam Dickinson, Everett Maus, Shannon Bales

Defensive Publications Series

An efficient test of a database is to subject it to random queries that are likely to succeed when evaluated on the database. Currently, automated query generation is difficult to use to target specific query shapes and difficult to adapt to diverse purposes. This disclosure describes a framework for generating random database queries by generating relational algebra trees and by building a query string from the algebra tree. The techniques enable the generation of useful, syntactically correct queries tuned towards particular depths, operator types, subtree structures, etc. Interesting algebra shapes can be targeted to test specific areas efficiently.

Automatic Contextual Adjustments For Interpretation Of Spoken Queries, Agoston Weisz, Alessandro Agostini

Defensive Publications Series

Automatic speech recognition techniques implemented in a virtual assistant or other application can sometimes fail to correctly transcribe a user query, even while utilizing user-permitted contextual information. Query recognition failures lead to misunderstanding of user intent. While increasing the strength of contextual biasing of speech recognition can fix the problem of misrecognition, too strong a bias can hurt queries that don’t benefit from such adjustments. This disclosure describes techniques that, upon failure to find suitable transcription, intent, or response, increase contextual bias provided to the ASR and re-run speech recognition to obtain a better transcription of user speech and determination …

Multilayer Restoration Over A Converged Ip And Optical Network, Randy Zhang, Gabriele Maria Galimberti, Errol Roberts

Defensive Publications Series

In a converged Internet Protocol (IP) plus optical multilayer network that provides an optical restoration capability, techniques are presented herein that enable differentiated forwarding for different classes of traffic (such as, for example, high-priority traffic, low-priority traffic, and best-effort traffic) through restoration awareness at Layer 3 (L3). Aspects of the presented techniques work without the need for any inter-layer protocols, such as the Generalized Multiprotocol Label Switching (GMPLS) User Network Interface (UNI) (GMPLS-UNI). This is important because a converged network, such as a routed optical networking (RON) environment, does not dictate protocols. Aspects of the presented techniques encompass restoration-enabled differentiated …

Intelligent Distributed Caching In Iot Environment To Avoid Data Loss, Mariusz Kaźmierski, Sanika Joshi, Bansari Trivedi, Paridhi Srivastava, Keerthyraja Rajaganapathi

Defensive Publications Series

We are witnessing a fourth industrial revolution and the Internet of Things (IoT) is a key part of that revolution. In an IoT world, the availability of data drastically changes the way in which information is interpreted. A loss of data from critical IoT devices can have a catastrophic impact. Accordingly, an intelligent distributed caching capability in an IoT infrastructure is imperative for successful IoT business deployments. To address such a need, techniques are presented herein that support an intelligent distributed caching mechanism. Aspects of the presented techniques provide for efficiently calculating the caching capability of connected devices to dynamically …

Low Latency Video Conferencing By Inpainting User Image Onto A Background, D Shin

Defensive Publications Series

Users sometimes opt out of artificial (custom) backgrounds or blur effects in video conferencing (VC) because they want to maintain the naturalness of a real background. However, if the wireless connection is poor, video conferencing quality can suffer if sufficient bandwidth to communicate the full video is not available. This disclosure describes techniques to reduce the bandwidth required for transmission of video with real backgrounds by taking one-time sequential snapshots of a video conference participant’s background and with user permission, storing the snapshot to make it available to other participants in the video conference. During the video conference, image regions …

Multitasking Across Native App And Web Content Using Custom Tabs, Victor Gallet, Chirag Desai, Peter Beverloo, Peter E. Conn, Shimi Zhang, Xiangqi Liu, Aina Rubies Espinalt, Theresa Sullivan, Ted Choc

Defensive Publications Series

Interacting with web links in native apps can disrupt the user's experience of the native app, as the web content is generally opened in full-screen mode in a separate browser window. This disclosure describes techniques that enable users to simultaneously access, or smoothly transition between, native app content and web content. Per the techniques, users are not pushed out of a native app experience when interacting with web links within the app. Rather, interacting with a web link results in web content being shown in a split-screen (referred to as a partial custom tab) alongside native app content. App developers …

Vision Improvement Via Artificial Reality Systems

Defensive Publications Series

No abstract provided.

Multilingual Voice Support For Contact Center Agents, Ananth Hegde, Amit Agnihotri, Nivedha Thangaraju

Defensive Publications Series

In a contact center, when a caller elects to speak to an agent the call will be routed (based on the language that the caller chose) to an agent who can speak the language that was selected by the caller. A global company with operations or offices in many different countries faces the challenge of communicating in multiple languages, something that is a key to providing excellent customer service. In particular, it can be challenging for a company to achieve such a capability when they simply cannot hire a customer support agent for every possible language. To address such challenges, …

A System And Method For Providing Digital Trasaction In Offline Mode To Prevent Double Spending Problem, Gokul Anand Manimaran Mr

Defensive Publications Series

Present disclosure provides a system and method for providing a platform to perform digital transaction in offline mode using virtual coins. The system may include network server (107) which is configured to generate virtual coins when the customer (102) requests for generating virtual coins. Each virtual coin includes unique virtual identification number and Universally Unique Identifier (UUID that ensures each coin cannot be spent more than once, preventing double spending problem. Thus, secured cashless transaction is performed in offline mode using the generated virtual coins. Further, network server (107) is configured to process payment transaction claim request received from the …

Fully Integrated On-Device Search And Web-Based Search Framework Incorporating Privacy Controls, Eugenio Marchiori, Donghui Feng, Trystan Upstill

Defensive Publications Series

This publication describes a framework, implemented on a computing device, for integrating search results provided by a user device system-level search with internet-based search results provided by an internet search engine application, while retaining the privacy of a user of the device. The search results may be ranked based on both the user device system-level search and the internet search.

Determining Fuel Quality, Sean Keys

Defensive Publications Series

The quality of gasoline and other fuel used in vehicles may vary with some gas stations selling higher quality fuels than other gas stations. This disclosure describes techniques for determining the quality of fuel sold by gas stations using existing vehicle sensors to monitor the fuel composition and combustion. The sensors may include harmonic and fuel sensors, such as water and ethanol sensors. The vehicle may determine the fuel quality by at least applying a machine learning algorithm to the collected combustion and fuel compensation information and/or may send the collected combustion and fuel compensation information to a remote computing …