Stochastic network traffic modeling

A system comprising:

• an input data processor configured to:
• receive network information characterizing a network, the network information comprising a plurality of nodes found in the network and an overall network demand;
• a node flow allocator configured to:
• generate a weight vector comprised of weights for each of the plurality of nodes;
• and
• perform a stochastic simulation assigning overall network demand to the network nodes by:
• dividing the overall network demand into a plurality of demand blocks;
• and
• allocating the plurality of demand blocks by, for each demand block:
• determining a first node, based on the weight vector, from the plurality of nodes;
• generating, based on the determined first node, a second weight vector comprised of weights for each of the plurality of nodes other than the first node;
• determining a second node, based on the second weight vector, from the plurality of nodes;
• and
• allocating the demand block to the combination of the first node and the second node;
• and
• a network flow aggregator configured to:
• generate a stochastic model of network demand between pairs of network nodes based on a plurality of stochastic simulations.

A non-transitory computer-readable storage medium comprising instructions that, when executed by a computing system, cause the computing system to perform operations comprising:

• receiving, at a computing system, network information characterizing a network, the network information comprising a plurality of nodes found in the network and an overall network demand;
• generating a weight vector comprised of weights for each of the plurality of nodes;
• performing a plurality of stochastic simulations assigning overall network demand to the network nodes, at least some of the stochastic simulations performed by:
• dividing the overall network demand into a plurality of demand blocks;
• and
• allocating the plurality of demand blocks by, for each demand block:
• determining a first node, based on the weight vector, from the plurality of nodes;
• generating, based on the determined first node, a second weight vector comprised of weights for each of the plurality of nodes other than the first node;
• determining a second node, based on the second weight vector, from the plurality of nodes;
• and
• allocating the demand block to the combination of the first node and the second node;
• and
• generating a stochastic model of network demand between pairs of network nodes based on the plurality of stochastic simulations.