Abstract: In one embodiment, a learning machine may be used to select observer nodes in a LLN such that the liveness of one or more nodes of interest may be monitored indirectly. In particular, a management device may receive network data on one or more network traffic parameters of a computer network. The management device may then determine, based on the network data, a candidate list of potential observer nodes to monitor activity or inactivity of one or more subject nodes. The management device may then dynamically select, using a machine learning model, a set of optimized observer nodes from the candidate list of potential observer nodes.

Abstract: In one embodiment, periodic round-trip probes are executed in a network, whereby a packet is transmitted along a particular communication path from a source to a destination and back to the source. Statistical information relating to the round-trip probes is gathered, and a transmission delay of the round-trip probes is calculated based on the gathered statistical information. Also, an end-to-end transmission delay along an arbitrary communication path in the network is estimated based on the calculated transmission delay of the round-trip probes.

Abstract: In one embodiment, network metrics are collected and analyzed in a network having nodes interconnected by communication links. Then, it is predicted whether a network element failure is relatively likely to occur based on the collected and analyzed network metrics. In response to predicting that a network element failure is relatively likely to occur, traffic in the network is rerouted in order to avoid the network element failure before it is likely to occur.

Abstract: In one embodiment, a routing topology of a network including nodes interconnected by communication links is determined. Important nodes in the network which are of relative importance are determined based on their location in the determined routing topology. Also, one or more request messages are sent causing the important nodes to gather local network metrics. Then, in response to the one or more request messages, one or more response messages including the network metrics gathered by each important node are received.

Abstract: In one embodiment, a packet to be transmitted along a communication path in a network from a source to a destination is determined, the communication path having one or more hops between the source and the destination. An instruction is sent to one or more tracking nodes along the communication path to track a number of local retransmissions required to successfully transmit the packet from each tracking node to a respective next-hop destination. Then, reports indicating the number of local retransmissions are received from the one or more tracking nodes.

Abstract: In one embodiment, one or more reporting nodes are selected to report network metrics in a network. From a monitoring node in the network, a trigger message is sent to the one or more reporting nodes. The trigger message may trigger the one or more reporting nodes to report one or more network metrics local to the respective reporting node. In response to the trigger message, a report of the one or more network metrics is received at the monitoring node from one of the one or more reporting nodes.

Abstract: In one embodiment, information relating to network metrics in a computer network is collected. A packet delay for a packet to be transmitted along a particular communication path is predicted based on the network metrics. Then, an optimal packet size for optimizing a transmission experience of the packet to be transmitted along the particular communication path is calculated based on the predicted packet delay. Also, a size of the packet to be transmitted along the particular communication path is dynamically adjusted based on the calculated optimal packet size.

Abstract: In one embodiment, a predictive model is constructed by mapping multiple network characteristics to multiple network performance metrics. Then, a network performance metric pertaining to a node in a network is predicted based on the constructed predictive model and one or more network characteristics relevant to the node. Also, a local parameter of the node is optimized based on the predicted network performance metric.

Abstract: In one embodiment, a device (e.g., learning machine) determines a plurality of fate-sharing group (FSG) nodes in a computer network that are prone to simultaneously send an alarm upon detecting an event. As such, the device may elect one or more FSG owner nodes as a subset of the FSG nodes, and instructs the FSG group such that only FSG owner nodes send an alarm upon event detection.

Abstract: In one embodiment, a learning data processor determines a plurality of machine learning features in a computer network to collect. Upon receiving data corresponding to the plurality of features, the learning data processor may aggregate the data, and pushes the aggregated data for select features to interested learning machines associated with the computer network.

Abstract: In one embodiment, a device determines a topological profile of individual nodes in a shared-media communication network, and also determines a respective likelihood of the nodes in the network to become a root of a floating topology based on the topological profiles. Accordingly, the device may provide instructions to particular nodes in the network based on the respective likelihoods.

Abstract: In one embodiment, techniques are shown and described relating to a point-to-multipoint communication infrastructure for expert-based knowledge feed-back using learning machines. A learning machine may communicate an expert discovery request into a network to discover one or more experts, and then receive from the one or more experts, one or more expert discovery responses. Based on the one or more received expert discovery responses, the learning machine may then build a dynamic multicast tree of experts to assist the learning machine in a computer network.

Abstract: In one embodiment, a capable node in a low power and lossy network (LLN) may monitor the authentication time for one or more nodes in the LLN. The capable node may dynamically correlate the authentication time with the location of the one or more nodes in the LLN in order to identify one or more authentication-delayed nodes. The node may then select, based on the location of the one or more authentication-delayed nodes, one or more key-delegation nodes to receive one or more network keys so that the key-delegation nodes may perform localized authentication of one or more of the authentication-delayed nodes. The capable node may then distribute the one or more network keys to the one or more key-delegation nodes.

Abstract: In one embodiment, a plurality of communication paths in a second direction in a communication network is determined, based on reversing communication paths established in a first direction in the communication network. Then, a path quality of the communication paths in the second direction is monitored. Based on the monitored path quality, it is then determined whether the communication paths in the second direction satisfy a communication requirement. Finally, a particular communication path of unacceptable quality in the second direction is detected when the particular communication path in the second direction fails to satisfy the communication requirement.

Abstract: In one embodiment, a message is received at a caching node in a network including an indication of the message's urgency. The message is transmitted to child nodes of the caching node, and upon transmitting the message, a retransmission timer is initiated when the message is urgent, based on the indication of the message's urgency. Then, one or more acknowledgements of receipt of the transmitted message are received from one or more of the child nodes, respectively. Upon expiration of the retransmission timer, when it is determined that one or more of the child nodes did not receive the transmitted message based on the received acknowledgements, the message is retransmitted to the child nodes.

Abstract: In one embodiment, a capable node in a low power and lossy network (LLN) may monitor the authentication time for one or more nodes in the LLN. The capable node may dynamically correlate the authentication time with the location of the one or more nodes in the LLN in order to identify one or more authentication-delayed nodes. The node may then select, based on the location of the one or more authentication-delayed nodes, one or more key-delegation nodes to receive one or more network keys so that the key-delegation nodes may perform localized authentication of one or more of the authentication-delayed nodes. The capable node may then distribute the one or more network keys to the one or more key-delegation nodes.

Abstract: Statistical and historical values of performance metrics are actively used to influence routing decisions for optimum topologies in a constrained network. Traffic service level is constantly monitored and compared with a service level agreement. If deviation exists between the monitored traffic service level and the terms of the service level agreement, stability metrics are used to maintain paths through the network that meet the terms of the traffic service level agreement or that improve the traffic flow through the network. Backup parent selection for a node in the network is performed based on previous performance of backup parents for the node.

Abstract: In one embodiment, a device in a computer network determines one or more tunnels affected by a downstream fault in the computer network, and determines one or more common Ethernet segments of the device used by the affected tunnels. As such, the device generates, for each of the one or more common Ethernet segments, a respective fault message aggregating tunnel information of each of one or more particular affected tunnels on the corresponding common Ethernet segment, and sends each respective fault message with aggregated tunnel information over a selected tunnel of the one or more particular affected tunnels on the corresponding common Ethernet segment.

Abstract: In one embodiment, a message instructing a particular node to act as a heartbeat relay agent is received at the particular node in a network. The particular node is selected to receive the message based on a centrality of the particular node. Heartbeat messages are then collected from child nodes of the particular node in the network. Based on the collected heartbeat messages, a heartbeat report is generated, and the report is transmitted to a collecting node in the network.

Abstract: In one embodiment, a particular node in a shared-media communication network determines a resource level and in response to determining a trigger condition (e.g., that the resource level is below a threshold), the particular node enters a selective forwarding mode. In the selective forwarding mode, the particular node does not forward non-critical messages. The particular node also notifies one or more neighboring nodes in the shared-media communication network of the entered selective forwarding mode. In another embodiment, a node may receive from a neighboring node, an indication of having entered a selective forwarding mode, and in response the node may forward only critical messages to the neighboring node.