Towards understanding the performance of IEEE 802.11 MAC in heterogeneous traffic conditions

Motivated by the need to study the performance of vehicular communication protocols as applicable to heterogeneous traffic conditions, we study the performance of IEEE 802.11p medium access protocol under such a traffic setup. We consider a setup comprising connected vehicles and human-driven Motorised Two Wheelers (MTWs), where the connected vehicles are required to move as platoon with a desired constant headway despite interruptions from the two wheelers. We invoke specific mobility models for the movement of the vehicles--car following models for connected vehicle platoons and gap-acceptance model to capture the movement of the MTWs--and use them to configure (i) the traffic setup and (ii) the rate at which data packets related to safety-critical messages need to be transmitted. A control-theoretic analysis of the car-following models yields a bound on the admissible communication delay to ensure non-oscillatory convergence of the platoon headway. We then use suitable Markov chain models to derive the distribution of the MAC access delay experienced by packets pertaining to safety-critical events as well as routine safety messages. The distribution along with the bound on the admissible delay enables us to derive the reliability of the 802.11p MAC protocol in terms of traffic and EDCA parameters. Our study highlights the need for redesign of MAC protocols for vehicular communications for safety-critical applications in heterogeneous conditions.