Abstract: This disclosure relates generally to the field of Electroencephalogram (EEG) classification, and, more particularly, to method and system for EEG motor imagery classification. Existing deep learning works employ the sensor-space for EEG graph representations wherein the channels of the EEG are considered as nodes and connection between the nodes are either predefined or are based on certain heuristics. However, these representations are ineffective and fail to accurately capture the underlying brain's functional networks. Embodiments of present disclosure provide a method of training a weighted adjacency matrix and a Graph Neural Network (GNN) to accurately represent the EEG signals. The method also trains a graph, a node, and an edge classifier to perform graph classification (i.e. motor imagery classification), node and edge classification. Thus, representations generated by the GNN can be additionally used for node and edge classification unlike state of the art methods.

Abstract: A device includes one or more network interfaces to receive layer two (L2) communications from an L2 network having a plurality of L2 devices; and a control unit to forward the L2 communications in accordance with forwarding information defining a plurality of flooding next hops. Each of the flooding next hops stored by the control unit specifies a set of the L2 devices within the L2 network to which to forward L2 communications in accordance with a plurality of trees, where each of the trees has a different one of the plurality of L2 devices as a root node. The control unit of the device computes a corresponding one of flooding next hops for each of the trees using only a subset of the trees without computing all of the trees having all of the different L2 network devices as root nodes.

Abstract: A technique for managing network traffic is disclosed. According to the technique, a set of VLAN IDs is explicitly identified for use with a first service. The rest of the customer traffic is considered as part of a default service. Traffic that is received at a PE from a customer is examined to identify whether or not the traffic belongs to the first service. For example, the VLAN ID and incoming port of a packet is compared to the set of VLAN IDs that were allocated to the first service on the respective port. Traffic that is identified as belonging to the first service is “extracted” from the default service and forwarded on a path that is related to the first service. The remaining traffic is forwarded on a path that is related to the default service. The service extraction technique can be implemented across an MPLS domain.

Abstract: Failure protection is provided between two routers that are actively implementing load sharing using network address translation. Failure protection is achieved by generating state information at the two routers and then exchanging the state information between the two routers so that each router has a combined set of state information that can be used in the event of a failure of the other router.

Abstract: A technique for managing network traffic is disclosed. According to the technique, a set of VLAN IDs is explicitly identified for use with a first service. The rest of the customer traffic is considered as part of a default service. Traffic that is received at a PE from a customer is examined to identify whether or not the traffic belongs to the first service. For example, the VLAN ID and incoming port of a packet is compared to the set of VLAN IDs that were allocated to the first service on the respective port. Traffic that is identified as belonging to the first service is “extracted” from the default service and forwarded on a path that is related to the first service. The remaining traffic is forwarded on a path that is related to the default service. The service extraction technique can be implemented across an MPLS domain.