Abstract: Establishing an expected transmit time at which a network interface controller (NIC) is expected to transmit a next packet. Enqueuing, with the NIC and before the expected transmit time, a packet P1 to be transmitted at the expected transmit time. Upon enqueuing P1, incrementing the expected transmit time by an expected transmit duration of P1. Transmitting at the NIC's line rate and timestamping enqueued P1 with its actual transmit time. Adjusting the expected transmit time by a difference between P1's actual transmit and P1's expected transmit time. Requesting, before completion of transmitting P1, to transmit a P2 at time t(P2). Enqueuing, in sequence, zero or more P0, such that the current expected transmit time plus the duration of the transmission of the P0s at the line rate equals t(P2). Transmitting at the line rate each enqueued P0. Upon enqueuing each P0, incrementing, for each P0, the expected transmit time by the expected transmit duration of the P0.

Abstract: This disclosure describes techniques and mechanisms for enabling a user to run heavy deep learning workloads on standard edge networks without off-loading computation to a cloud, leveraging the available edge computing resources, and efficiently partitioning and distributing a Deep Neural Network (DNN) over a network. The techniques enable the user to split a workload into multiple parts and process the workload on a set of smaller, less capable compute nodes in a distributed manner, without compromising on performance, and while meeting a Service Level Objective (SLO).

Abstract: A method is provided in one example embodiment and includes, for each of a plurality of individual storage units collectively comprising a virtual storage unit, mapping an internal address of the storage unit to a unique IP address, wherein each of the storage units comprises a block of storage on one of a plurality of physical storage devices and wherein the IP address includes a virtual storage unit number identifying the virtual storage unit; receiving from a client a request to perform an operation on at least one of the data storage units, wherein the request identifies the internal address of the at least one of the data storage units; translating the internal address of the at least one of the data storage unit to the unique IP address of the at least one of the data storage units; and performing the requested operation on the at least one of the data storage units.

Abstract: Techniques for provisioning multicast chains in a cloud-based environment are described herein. In an embodiment, an orchestration system sends a particular model of a distributed computer program application comprising one or more sources, destinations, and virtualized appliances for initiation by one or more host computers to a software-defined networking (SDN) controller. The SDN controller determines one or more locations for the virtualized appliances and generates a particular updated model of the distributed computer program application, the updated model comprising the one or more locations for the virtualized appliances. The SDN controller sends the updated model of the distributed computer program application to the orchestration system.

Abstract: A method comprises collecting, by a computing device located at an edge of a network, data items corresponding to information transmitted by endpoints using the network, generating, by the computing device, a probabilistic hierarchy using the data items, generating, by the computing device using the probabilistic hierarchy and natural language data, a similarity metric, generating, by the computing device using the probabilistic hierarchy, the natural language data, and the similarity metric, an ontology, detecting, by the computing device using the ontology, an anomaly, and in response to detecting the anomaly, sending a notification.

Abstract: In one embodiment, an interactivity agent provides display data from a plurality of devices to a touch screen display via a network for simultaneous presentation on the touch screen display. The interactivity agent receives touch information regarding a touch action performed via the touch screen display with respect to the display data from a particular one of the plurality of devices. The interactivity agent translates the received touch information from the touch screen display into translated touch information for the particular device. The interactivity agent sends, to a particular device via the network, the translated touch information. The translated touch information causes the particular device to replicate the touch action on a display of the particular device.

Abstract: According to one aspect, a method includes obtaining a segment routing (SR) packet from an endpoint via a first router at a first server along a path, the SR packet including an SR list and a last address, the last address being an address of a requested service. The method also includes determining, at the first server, whether the requested service is available from the first server, wherein determining whether the requested service is available from the first server includes opening the SR packet, parsing an SR header of the SR packet, and performing a lookup in a service table. Finally, the method includes modifying the SR packet at the first server when it is determined that the requested service is not available from the first server; and forwarding the SR packet along the path.

Abstract: A method is provided in one example embodiment and may include determining at a parent content node that a plurality of recipient content nodes are to receive a same content; generating, based on a determination that the same content is available at the parent content node, a multi-delivery header comprising a plurality of identifiers, wherein each identifier of the plurality of identifiers indicates each recipient content node that is to receive the same content; appending the multi-delivery header to one or more packets of an Internet Protocol (IP) flow associated with the same content; and transmitting packets for the IP flow to each of the plurality of the recipient content nodes.

Abstract: Establishing an expected transmit time at which a network interface controller (NIC) is expected to transmit a next packet. Enqueuing, with the NIC and before the expected transmit time, a packet P1 to be transmitted at the expected transmit time. Upon enqueuing P1, incrementing the expected transmit time by an expected transmit duration of P1. Transmitting at the NIC's line rate and timestamping enqueued P1 with its actual transmit time. Adjusting the expected transmit time by a difference between P1's actual transmit and P1's expected transmit time. Requesting, before completion of transmitting P1, to transmit a P2 at time t(P2). Enqueuing, in sequence, zero or more P0, such that the current expected transmit time plus the duration of the transmission of the P0s at the line rate equals t(P2). Transmitting at the line rate each enqueued P0. Upon enqueuing each P0, incrementing, for each P0, the expected transmit time by the expected transmit duration of the P0.

Abstract: In one embodiment, a device in a network receives control plane packet data indicative of control plane packets for a control plane in the network. The device models the control plane using a machine learning model based on the control plane packet data. The device predicts an instability in the control plane using the machine learning model. The device causes performance of a mitigation action based on the predicted instability in the control plane.

Abstract: Establishing an expected transmit time at which a network interface controller (NIC) is expected to transmit a next packet. Enqueuing, with the NIC and before the expected transmit time, a packet P1 to be transmitted at the expected transmit time. Upon enqueuing P1, incrementing the expected transmit time by an expected transmit duration of P1. Transmitting at the NIC's line rate and timestamping enqueued P1 with its actual transmit time. Adjusting the expected transmit time by a difference between P1's actual transmit and P1's expected transmit time. Requesting, before completion of transmitting P1, to transmit a P2 at time t(P2). Enqueuing, in sequence, zero or more P0, such that the current expected transmit time plus the duration of the transmission of the P0s at the line rate equals t(P2). Transmitting at the line rate each enqueued P0. Upon enqueuing each P0, incrementing, for each P0, the expected transmit time by the expected transmit duration of the P0.

Abstract: A method for retrieving content on a network comprising a first device and a second device is described. The method includes receiving in the network a request for content from the first device, the request identifying the content using an IPv6 address for the content, and determining whether the content is stored in a cache of the second device. Upon determining the content is stored in the cache of the second device, a request is sent to the second device for the content using the IPv6 address of the content. The content is forwarded to the first device from the second device, wherein the first and second devices are part of the same layer 2 domain. Methods of injecting content to a home network and packaging content are also described.

Abstract: Systems, methods, and computer program products relating to resilient transmission of a media stream over a communication network. A plurality of data packets are received over a communications network. The plurality of data packets relate to a first source video portion transformed using a geometric transform. The geometric transform is configured to modify a location of pixels in the first source video portion such that a plurality of adjacent pixels in the first source video portion are not adjacent after transformation. A received video portion is assembled based on the plurality of data packets. The received video portion is transformed, using an inverse of the geometric transform, to generate a second source video portion. The second source video portion and the first source video portion include a plurality of matching pixels.

Abstract: Systems, methods, and computer-readable media for load balancing using segment routing and application monitoring. A method can involve receiving a packet including a request from a source device to an application associated with a virtual address in a network, mapping the request to a set of candidate servers hosting the application associated with the virtual address, and encoding the set of candidate servers as a list of segments in a segment routing header associated with the packet. The method can further involve determining that a first candidate server from the set of candidate servers is a next segment in the list of segments, encoding the first candidate server in a destination address field on a header of the packet, and forwarding the packet to the first candidate server.

Abstract: Embodiments provide techniques for delivering a paced stream of video data packets. One embodiment includes receiving a data stream of video data packets formatted according to a Society of Motion Picture and Television Engineers (SMPTE) standard. A desired rate of delivery for the video data packets at a gateway device is determined. Embodiments generate a padded data stream by inserting one or more pause frames in between the video data packets in the received data stream, based on the desired rate of delivery. The padded data stream is transmitted across a link to a network switch, to be transmitted to the gateway device.

Abstract: Establishing an expected transmit time at which a network interface controller (NIC) is expected to transmit a next packet. Enqueuing, with the NIC and before the expected transmit time, a packet P1 to be transmitted at the expected transmit time. Upon enqueuing P1, incrementing the expected transmit time by an expected transmit duration of P1. Transmitting at the NIC's line rate and timestamping enqueued P1 with its actual transmit time. Adjusting the expected transmit time by a difference between P1's actual transmit and P1's expected transmit time. Requesting, before completion of transmitting P1, to transmit a P2 at time t(P2). Enqueuing, in sequence, zero or more P0, such that the current expected transmit time plus the duration of the transmission of the P0s at the line rate equals t(P2). Transmitting at the line rate each enqueued P0. Upon enqueuing each P0, incrementing, for each P0, the expected transmit time by the expected transmit duration of the P0.

Abstract: In one embodiment, a method includes training a deep neural network using a first set of network characteristics corresponding to a first time and a second set of network characteristics corresponding to a second time, generating, using the deep neural network, a predictive set of network characteristics corresponding to a future time, and assigning a task of a distributed application to a processing unit based on the predictive set of network characteristics.

Abstract: Various systems and methods for counting people. For example, one method involves receiving input data at an analytics system that includes a neural network. The input data includes a representation of an environment, including representations of several people. The method also includes identifying the representations of the people in the representation of the environment. The method also includes updating an output value that indicates the number of people identified as being present in the environment.

Abstract: In one embodiment, a method includes receiving current data, the current data including time series data representing a plurality of time instances. The method includes storing at least a recent portion of the current data in a buffer. The method includes reducing the dimensionality of the current data to generate dimensionality-reduced data. The method includes generating a reconstruction error based on the dimensionality-reduced data and a plurality of neural network metrics. At least one of a size of the recent portion of the current data stored in the buffer or an amount of the reducing the dimensionality of the current data is based on the reconstruction error.

Abstract: A method comprises collecting, by a computing device located at an edge of a network, data items corresponding to information transmitted by endpoints using the network, generating, by the computing device, a probabilistic hierarchy using the data items, generating, by the computing device using the probabilistic hierarchy and natural language data, a similarity metric, generating, by the computing device using the probabilistic hierarchy, the natural language data, and the similarity metric, an ontology, detecting, by the computing device using the ontology, an anomaly, and in response to detecting the anomaly, sending a notification.