Evolutionary Routing Strategies for Automotive Networks

The design of efficient communication networks is a challenging task for modern vehicle development. Due to novel technologies and new degrees of freedom in network design, the decision on a bus topology has a severe impact on the overall system cost and performance. In this work, we propose a topology and routing optimization using Evolutionary Algorithms and problem-specific encoding. Our contribution includes a guided topology mutation operator which outperforms standard random mutation. Further, we propose two routing operators for usage during the optimization process and compare their effectiveness on a network application taken from a series vehicle.