APDM: An adaptive multi-priority distributed multichannel MAC protocol for vehicular ad hoc networks in unsaturated conditions

Abstract

Vehicular ad hoc networks (VANETs) employ multichannel to provide a variety of safety and non-safety applications. Safety applications require reliable and timely transmission, while non-safety applications need high network throughput. IEEE 802.11p and IEEE 1609.4 protocol divide the bandwidth into seven channels. One control channel (CCH) is to serve safety applications and the other six service channels (SCHs) to serve non-safety applications. The IEEE 1609.4 protocol specifies an alternating scheme to allow vehicles to switch between two types of applications. However, the IEEE 1609.4 multichannel media access control (MAC) protocol has limitations on its capability of supporting either delay-or throughput-sensitive applications. In this paper, we propose an adaptive multi-priority distributed multichannel (APDM) MAC protocol for VANETs. Considering that in realistic VANETs, the queue of MAC layer is far from saturated. We assume that generated packets with different priorities arrive at the MAC layer in a Poisson manner. A Markov analytical model is conducted to optimize the packet transmission probabilities and adjust the ratio between CCH interval and SCH interval dynamically according to the real-time traffic in a distributed way. An M/M/1 queue model is then adopted to analyze the time performance. Extensive simulation results show that the APDM MAC protocol can ensure prioritized transmission of safety packets, reduce the transmission delay of packets and enhance the unsaturated and saturated throughput of SCHs.

Performance analysis of adaptive MAC protocol in VANETs considering the potential impact on throughput and transmission delays

“…APDM 20 -Reduces the safety packet delay. -Safety and non-safety packets are transmitted in a flexible way.…”

“…1,20 Similarly to the 1-D Markov chain model, we let P i,k|j,l denote the one-step transition probabilities where P i,k|j,l is {Pr{s(t + 1) = i, b(t + 1) = k|s(t) = , b(t) = l}. 1,20 Similarly to the 1-D Markov chain model, we let P i,k|j,l denote the one-step transition probabilities where P i,k|j,l is {Pr{s(t + 1) = i, b(t + 1) = k|s(t) = , b(t) = l}.…”

“…20 The expected time spent per state was T es = (1 − p b ) + p e,s T e,s + p s,s T s,s + p e,c T e,c + p s,c T s,c + p es,c max(T e,c , T s,c )· 20 The expected time spent per state was T es = (1 − p b ) + p e,s T e,s + p s,s T s,s + p e,c T e,c + p s,c T s,c + p es,c max(T e,c , T s,c )·…”

“…20 Consequently, based on Equations (1), (3), (4), (5), and (9), the variables e , s , p busy , p e , and p s could be determined by numerical methods, 38 where 0 < e , s , p busy , p e , p s < 1. 20 Consequently, based on Equations (1), (3), (4), (5), and (9), the variables e , s , p busy , p e , and p s could be determined by numerical methods, 38 where 0 < e , s , p busy , p e , p s < 1.…”

“…According to adaptive MAC protocols, 15,17,20,21 the number of reservations made on the control channel is equal to the number of data packets transmitted on all SCHs, ie, G 1 = G 2 . Based on this, the optimal CCHI is computed.…”