Vehicles nowadays rank the third place where people spend significant amount of time daily after their home and office. However, the high-rate Internet access, which has almost become an integral part of our daily lives, is still not widely available for mobile users in vehicles in most regions of the world. One way to address this is to deploy cost-effective communications infrastructure on the roadside, and provide limited connections to vehicles when they are driving through the radio coverage of the infrastructure, namely drive-thru Internet.
Inspired by the exciting success of WLAN, a prevailing and practical approach is to utilize the widespread WLAN access points (APs) in the city as the infrastructure to enable the drive-thru Internet. Initial field tests demonstrated that when a single vehicle drives through a roadside IEEE 802.11b AP, the vehicle can maintain a connection to the AP for around 500m and transfer 9MB of data at one drive-thru at 80 km/h using either TCP or UDP. However, when many vehicles drive through simultaneously and compete for communications with the roadside AP, the scalability and throughput performance of the network are still unclear. More importantly, to make the drive-thru Internet practical and applicable for different road traffic conditions, an accurate yet efficient theoretical model, which captures the detailed traffic (vehicle mobility and density) and network parameters, is necessary to guide the real-world deployments.
Motivated by above concerns, in this work, the authors have developed a three-dimensional Markov chain model to evaluate the nodal and system MAC throughput performance of the drive-thru Internet, when multiple vehicles contend for transmissions to an roadside IEEE 802.11b AP using the DCF protocol. In specific, by using the vehicle velocity and settings of DCF parameters as inputs to the model, both the average nodal throughput of vehicles and the integrated system throughput after MAC contentions in the drive-thru Internet scenario can be derived. Based on the analysis and observations in simulations, the authors have further proposed three enhancement mechanisms to legacy IEEE 802.11b DCF protocol when deployed in the drive-thru Internet scenario. With the three enhancement mechanisms applied, they have developed an optimization framework to guide the optimal selections of DCF parameters in different road traffic conditions towards the maximal system throughput. Both the accuracy of the analytical model and the efficiency of the optimization framework are then validated through extensive simulations. The results can provide important guideline to the deployment of cost-effective communications infrastructure on the roadside.