Genetic Algorithm-based Study on Flow Allocation in a Multicommodity Stochastic-flow Network with Unreliable Nodes

Many real-life networks can be abstracted into a stochastic-flow network. In this paper, we assume there are several sorts of resource flows transmitting through a stochastic-flow network with unreliable nodes. We want to find a optimal resource flow allocation and control strategy upon arcs and nodes. Under this strategy, the probability of satisfying sink nodes´ demand is maximized when resource flows transmit from source nodes to sink nodes. We propose a genetic algorithm to seek the optimal strategy. At last, a numerical example is given to test the proposed algorithm.