Performance Analysis of Computation Offloading in Vehicular Networks: Cars vs. Motorcycles

This work explores computation offloading in vehicular networks, emphasizing the distinct challenges and opportunities presented by cars and motorcycles. Computation offloading remains essential and challenging for enabling advanced functionalities and real-time processing in dynamic vehicular environments. In this study, we analyze the different challenges cars and motorcycles face in performing computation offloading. For this, we use different algorithms from the literature to evaluate the performance of these vehicles in various situations. A comprehensive system model was developed to calculate total execution times and offloading failures. Extensive simulations carried out based on a real urban environment revealed intriguing performance results for cars in relation to motorcycles due to the dynamicity of their movement in the scenario. The results show that motorcycles still need new strategies to perform computation offloading with better performance or that reach a level similar to that of cars. These findings guide future efforts to optimize computation offloading mechanisms in motorcycles.