Resource Allocation in Mobile Networks: A Decision Model Of Jockeying in Queues

Application specific network slices offered by different vendors is one of those approaches proposed to bridge the gap between the demand and supply of resources in next generation communication networks. We present a decision model as an empirical study that associates different slice configurations to queues in a multi-vendor setting such that consumers continuously assess their preferences and make rational decisions that guide the optimal usage of the resource pools. The model is based on the findings from our developmental tooling which provides measures about the sensitivity to variations in buffer parameters on the impatient customer's decision to jockey and the frequency of this behavior. We analytically express for these sensitivities to formulate for approximations of the number of times a task is switched from one queue to another and validate the decision model against empirical data from the Monte Carlo simulation.