Cross-layer cooperative offloading in vehicular edge computing networks

Abstract

Vehicular edge computing (VEC) uses the computing resources of edge devices to complete tasks for complex calculations and time-sensitive requirements in vehicular networks. However, there are some problems with the current task offloading, such as simplifying the communication mechanism at the media access control (MAC) layer and ignoring the traffic characteristics at the application layer. To address the problems, we first model the transmission mechanism of MAC and the packet queuing delay for IEEE 802.11p. Then, we propose a cross-layer cooperative offloading (CLCO) algorithm based on rate matching between the application layer and the MAC layer. The algorithm evaluates the network load by detecting the queue length of the vehicular sending buffer. The application rate adopts the multiplicative-decrease strategy to reduce the end-to-end packet delay when the queue length exceeds a queue threshold. Alternatively, the application rate adopts the additive-increase strategy to use MAC bandwidth adequately when the queue length is lower than the queue threshold. NS simulations show that the proposed CLCO algorithm reduces the delay in task offloading, while the amount of offloading data and the packet delivery rate are satisfactory.