Deadline-constrained RSU-To-vehicle task offloading scheme for vehicular fog networks

Abstract

In Vehicular Fog Computing (VFC), the RSU-To-Vehicle (R2V) task offloading process is highly affected by undesirable yet sometimes inevitable events (e.g., buffer exhaustion, task HoL blocking and deadline expiry), the occurrence of which will notably alter the RSU's performance in terms of crucial Quality-of-Service (QoS) metrics such as the average system response time, the blocking and deadline expiry probabilities. This article proposes a novel R2V Deadline-Constrained Task Offload (R2V-DCTO) scheduling scheme with the objective of improving the RSU's performance in terms of the above-mentioned metrics; hence, filling an important literature gap. For this purpose, a stochastic modelling framework is established to capture the RSU's functional dynamics and assess its performance as it operates under R2V-DCTO. Extensive simulations are conducted to confirm the model's validity and accuracy and then compare R2V-DCTO's performance to that of the often adopted FIFO scheme. Results indicate that, on average, R2V-DCTO outperforms FIFO by 33.4% in terms of the probability of deadline mismatch and by 83.3% in terms of mean system response time; those being critical QoS performance metrics. © 1967-2012 IEEE.