Impact of Geographical Constraints on the Performance of Intent-Based Networks

Paper i proceeding, 2024

In intent-based multilayer networks, one of the possible intents may restrict specific network nodes from routing. In particular, certain parts of the network may be subject to various malicious attacks, or, in disaggregated networks, parts of the network can be equipped with devices not aligned with the enduser business policy. In this work, we study the impact of geographical constraints on the performance of multilayer intent-based networks. To this end, we determine crucial network nodes using the k-shortest path betweenness centrality in two representative network topologies. We investigate nodes' contribution to the network performance, measured as bandwidth blocking probability. In particular, we show that even for light traffic loads, restrictions on crucial network nodes highly impact network performance, while for less important nodes, it is possible to maintain good network performance when the traffic load is light.