Computational Modeling and Simulation of Dynamic Communication Network Resource Allocations in Excel

A simulation tool has been developed to simulate the operation and performance of demand-assigned communication networks having time-varying data traffic patterns. Such networks use protocols to increase or decrease the resources allocated to a user as the resource needs at that node change. These resource-allocation protocols are typically based on either queue sizes or data message generation rates when additional resources are needed, and based on unused resources when excess resources are available and can be released. The protocols have parameters such as numerical thresholds for requesting or releasing resources. The determination of optimum values for such parameters is often assisted by simulations. Although commercially available simulation tools could be used for these simulations, such tools are often expensive and sometimes complicated to adapt to new and not yet standardized protocols. The simulation tool presented here is a simple tool easily programmed into an Excel spreadsheet. The tool has three sections, one that generates data messages having controllable random message-generation characteristics, a second section that queues and processes the messages in a manner that emulates the processing done within the actual network nodes, and a third section that executes the resource-allocation protocol within the network node. The simulator allows control of the resource-allocation protocol parameter values for optimization. The overall design of the simulation tool is described, as is each of the three sections of the simulator. The simulator allows examination of how responsive the protocols are to changes in data message generation rates, and can track messages that are discarded due to being queued an excessively long time. Examples are included to illustrate the effects on performance metrics when parameter values for the resource-allocation protocol are adjusted.