Auction based Energy-Efficient Cooperative Relay Scheduling in Bidirectional Highway Scenarios for VANET

In a Vehicular Adhoc Network (VANET), Road Side Units (RSUs) deployed along the highways are generally dispossessed from a permanent connection to grid-power and mostly the RSU operations are sustainable on alternative sources such as solar power, wind power, etc. The power consumption of RSU is dominated by the downlink (RSU-to-vehicle) communication cost and it is a function of RSU to vehicle distance. Moreover, the RSUs cannot cover entire highway segment due to high deployment cost and they leave an outage area in between neighboring RSUs. In this case, a target vehicle entering into an outage area with unserved requests can be served by selecting relay (store-carry-forward) vehicles. This paper proposes RSU assisted relay scheduling algorithm for faster data retrieval to the target vehicle while minimizing the RSU energy consumption. Firstly, a time slot based relay scheduling problem is formulated and then the NP-hardness of the problem is proved. Secondly, a greedy algorithm (GA) and a forward relay scheduler (FRS) are proposed to schedule the relay vehicles in forward direction (i.e. target moving direction). The GA selects the relay vehicles which are near to RSU and takes less time to reach the target vehicle. The FRS apply the concepts of Auction Theory to optimally assign the relay vehicles to time slots, that can jointly minimize RSU energy consumption and end-to-end delay to the target vehicle. Here, the Auction process allows the relay vehicles to bid for time slots by using their local information such as speed, location, cooperative cache size, direction, etc. Furthermore, the proposed Auction based RSU assisted relay scheduling (RRS) triggers the neighboring RSUs which can cooperate by sharing unserved data and schedule the relay vehicles driving in both forward and backward (i.e. opposite to target) directions in a bidirectional highway. Simulation results show that the efficacy of the proposed algorithms over other relay scheduling algorithms with respect to energy consumption, end-to-end delay and residual data delivery.