On efficient admission control utilizing cloud compression in drive-thru vehicular networks

Providing vehicles with various services is an challenging issue for roadside units to support applications in future Intelligent Transportation System (ITS). Since each vehicle has only a limited period of time passing through a roadside unit, how to guarantee service data to be downloaded in time has been brought to the front. In this paper, we focus on the admission control problem, which is especially significant in overloaded conditions where tens of vehicles may share the resource of one roadside unit. To enhance downloading efficiency, we employ vehicular cloud which is extremely promising in future ITS to compute downloading time and compress contents. We propose an admission control scheme that calculates and increases the probability of successful downloads in a drive-thru system, which also promotes users' quality of experience. In addition, we formulate the admission control issue as a linear programming problem, and utilize a branch-and-bound algorithm to solve it. Simulations demonstrate that our proposed scheme reduces tasks failure notably and outperforms two regular downloading schemes in terms of throughput.