Inspired path computation in a network

An apparatus, comprising:

• one or more network interfaces to communicate with a network;
• a processor coupled to the network interfaces and configured to execute one or more processes;
• and
• a memory configured to store a process executable by the processor, the process when executed configured to:
• train a deep reinforcement learning-based agent to predict traffic flows in the network, wherein the deep reinforcement learning-based agent uses a reward function determined using inverse reinforcement learning that penalizes when a traffic flow is fragmented across multiple paths in the network, and wherein the inverse reinforcement learning is used to determine the reward function based on observed behavior in the network in which flows are transmitted without being fragmented across different network paths;
• determine one or more resource requirements for the predicted traffic flows;
• assign, using the deep reinforcement learning-based agent, paths in the network to the flows based on the determined one or more resource requirements, to avoid fragmentation of a flow during transmission of the flow through the network;
• and
• send, to nodes in the network, assignment instructions that cause the flows to traverse the network via their assigned paths.

A tangible, non-transitory, computer-readable medium storing program instructions that cause a controller in a network to execute a process comprising:

• training, the, a deep reinforcement learning-based agent to predict traffic flows in the network, wherein the deep reinforcement learning-based agent uses a reward function determined using inverse reinforcement learning that penalizes when a traffic flow is fragmented across multiple paths in the network, and wherein the inverse reinforcement learning is used to determine the reward function based on observed behavior in the network in which flows are transmitted without being fragmented across different network paths;
• determining, by the controller, one or more resource requirements for the predicted traffic flows;
• assigning, by the controller and using the deep reinforcement learning-based agent, paths in the network to the flows based on the determined one or more resource requirements, to avoid fragmentation of a flow during transmission of the flow through the network;
• and
• sending, by the controller and to nodes in the network, assignment instructions that cause the flows to traverse the network via their assigned paths.