Comparing Control Theory and Deep Reinforcement Learning techniques for decentralized task offloading in the edge-cloud continuum | Publicación