Extensible and realistic modeling of resource contention in resource-constrained nodes

In the past, network performance has been strongly determined by the capacity of the wired or wireless links. In future, however, we can expect that with the emergence of resource-constrained nodes (e.g. software routers) and high-speed links (e.g. optical fiber), the network nodes themselves will be the bottleneck. To increase the packet processing of resource-constrained nodes, parallel processing with multi-core processors is required. Intra-node resource contention can have a strong negative impact on the corresponding network node and, therefore, also on the overall performance of the network. However, several commonly used network simulators only offer a rather simplistic node model and do not take into account intra-node resource contention. In particular, latencies and input queueing behavior of resource-constrained nodes are typically not modeled in current simulators. We propose a unified and extensible approach to model intra-node resource contention which enables us to identify and predict performance bottlenecks in current and future networks. We implemented our model as an extension to the network simulator ns-3. The simulation results show that our approach significantly outperforms the original ns-3 in terms of realistic modeling. This claim has been validated successfully by means of real testbed measurements.