Approaches presented herein enable deploying a distributed machine learning framework in an edge computing environment. More specifically, a status of a connection between a computing system and an edge node of a plurality of edge nodes is monitored. At least one server node and a group of worker nodes from the plurality of edge nodes are identified based on the status. A path for distributing the training data to the worker nodes is determined based on the status. The training data from the edge node to the worker nodes is distributed via the path.

The described technology is generally directed towards methods for data latency evaluation. The techniques disclosed herein can provide useful information about when a data consumer can expect to receive data. Methods can create and compare data latency cumulative probability distributions comprising probabilities associated with different latency values, at various different levels of completeness.

The described technology is generally directed towards methods for data latency evaluation. The techniques disclosed herein can provide useful information about when a data consumer can expect to receive data. Methods can create and compare data latency cumulative probability distributions comprising probabilities associated with different latency values, at various different levels of completeness.

A cloud network is a complex environment in which hundreds and thousands of users or entities can each host, create, modify, and develop multiple virtual machines. Each virtual machine can have complex behavior unknown to the provider or maintainer of the cloud. Technologies disclosed include methods, systems, and apparatuses to monitor the complex environment to detect network anomalies using machine learning techniques. In addition, techniques to modify and adapt to user feedback are provided allowing the developed models to be tuned for specific use cases, virtual machine types, and users.

Described is a method and system for connectivity diagnostics in communication systems. The method comprises: querying a first communication device at a first time and a second time to determine whether a second communication device is connected to the first communication device and to determine a value of an operational parameter at the first and second times; and determining the second communication device disconnected from the first communication device based on detecting the second communication device was connected to the first communication device at both the first time and the second time, and detecting the value of the operational parameter at the second time is inside a range of threshold values. In one embodiment, the method comprises determining a link is unstable for connectivity based on connection duration, number and/or pattern of connection and/or disconnection events, and/or traffic activity during connection and/or disconnection events.

Described is a method and system for connectivity diagnostics in communication systems. The method comprises: querying a first communication device at a first time and a second time to determine whether a second communication device is connected to the first communication device and to determine a value of an operational parameter at the first and second times; and determining the second communication device disconnected from the first communication device based on detecting the second communication device was connected to the first communication device at both the first time and the second time, and detecting the value of the operational parameter at the second time is inside a range of threshold values. In one embodiment, the method comprises determining a link is unstable for connectivity based on connection duration, number and/or pattern of connection and/or disconnection events, and/or traffic activity during connection and/or disconnection events.

Disclosed herein is a method, a computer program product, and a carrier for indicating one-way latency in a data network (N) between a first node (A) and a second node (B), wherein the data network (N) lacks continuous clock synchronization, comprising: a pre-synchronisation step, a measuring step, a post-synchronisation step, an interpolation step, and generating a latency profile. The present invention also relates to a computer program product incorporating the method, a carrier comprising the computer program product, and a method for indicating server functionality based on the first aspect.

The present disclosure generally discloses capabilities for supporting new network zones and associated services. The network zones and associated services may include a near-real-time (NRT) zone and associated NRT services, a real-time (RT) zone and associated RT services, or the like. The resilient network zones and associated resilient and non-resilient services may be configured to provide bounded latency guarantees for reliably supporting various types of applications (e.g., mobile fronthaul, cloud computing, Internet-of-Things (IoT), or the like). The network zones and associated services may be provided using a distance-constrained fiber and wavelength switching fabric design comprised of various network devices and using associated controllers, which may be configured to support service provisioning functions, service testing functions, wavelength switching functions, and so forth.

Disclosed are various embodiments for on-demand application-driven network slicing. In one embodiment, it is determined that an application implemented in a particular computing device has an increased quality-of-service requirement in order to send or receive data via a communications network. The increased quality-of-service requirement is greater than an existing quality-of-service provided to the application by the communications network. The application sends a request that causes capacity in a network slice having the increased quality-of-service requirement in the communications network to be allocated for the application. The data is transmitted to or from the application using the network slice.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for receiving from a plurality of publishers messages of a first channel of a plurality of distinct channels wherein each channel comprises an ordered plurality of messages, storing messages of the first channel in one or more first buffers according to the order, each first buffer having a respective time-to-live, for one or more connections, determining a respective sampling rate based on a data type of the first channel and a determined latency of the connection, receiving from a subscriber through a first connection a request for messages of the first channel, selecting messages in the first buffers according to the order and the sampling rate, and sending the selected messages using the first connection to the subscriber according to the order.