Distributed embodied evolution over networks

In several network problems the optimal behavior of agents (i.e., nodes network) is not known before deployment. Furthermore, might be required to adapt, i.e. change their based on environment conditions. these scenarios, offline optimization usually costly and inefficient, while online methods more suitable. this work, we use a distributed Embodied Evolution approach optimize spatially distributed, locally interacting by allowing them exchange parameters learn from each other adapt certain task within given environment. Our results test scenarios show that local information, performed means crossover with neighbors, allows conduct process efficiently than cases where interactions are allowed, even when there large differences agent’s neighborhood.