Vehicle-Centric Network Association in Heterogeneous Vehicle-to-Vehicle Networks

Safe and reliable operation of future autonomous vehicles requires ultra-low latency vehicle-to-vehicle (V2V) networking. However, due to scalability issues of purely ad hoc network, vehicle-to-infrastructure (V2I) communication is still in demand. To avoid interference between V2V and V2I communications, conventional network association such as handover may generate a large amount of control and signaling overhead, especially in environments with high-speed mobility and small-coverage access points. Taking the additional latency caused by the overhead into consideration, a vehicle-centric approach is proposed to guarantee the low-latency requirements of the V2V networks. Different from solely optimizing the latency performance, the proposed approach focuses on effectively operating the network association and resources re-allocation. This avoids the massive exchange of control information and signaling and reduces its impact on latency. The proposed framework for V2V networking can underlay different virtual cells with a constrained network association rate and we can derive the corresponding upper bound of the latency performance. Analysis and simulation results indicate that our scheme successfully guarantees the upper bound of the latency and simultaneously constrains the network association rate, thus limiting the additional control exchange and signaling overhead.