摘要:
At a source node, a plurality of packets may be determined for transmission to a destination node in a network comprising a plurality of network nodes. A transmission rate of the plurality of packets from the source node to a neighbor node in the network may be adaptively controlled, based on a determination of a current status of the network by utilizing a plurality of parameters that are estimated via a reinforcement learning routing algorithm. The plurality of parameters include an estimated cost value representing a current cost to transmit the plurality of packets to the destination node via the network. Transmissions from intermediate nodes may also be adaptively deferred based on a determination of a current status of the network by utilizing the plurality of parameters.
摘要:
A reinforcement learning-based method is provided that enables efficient communication for networks having varying numbers and topologies of mobile and stationary nodes. The method provides an autonomous, optimized, routing method that may be implemented in a distributed manner among the nodes that allows the nodes to make intelligent decisions of how to forward data from a source node to a destination node with little or no a priori information about the network. The method involves receiving, at a node within a distributed network, data packets containing position and velocity information from a transmitting node. Position and velocity estimates are determined for the transmitting and receiving nodes using the position and velocity information. State-action pair value estimates are determined in the destination direction for forward packets and the source direction for backward sweeping packets, along with associated destination direction and source direction state value estimates, which determine packet transmittal.
摘要:
At a source node, a plurality of packets may be determined for transmission to a destination node in a network comprising a plurality of network nodes. A transmission rate of the plurality of packets from the source node to a neighbor node in the network may be adaptively controlled, based on a determination of a current status of the network by utilizing a plurality of parameters that are estimated via a reinforcement learning routing algorithm. The plurality of parameters include an estimated cost value representing a current cost to transmit the plurality of packets to the destination node via the network. Transmissions from intermediate nodes may also be adaptively deferred based on a determination of a current status of the network by utilizing the plurality of parameters.