Early alert system using relative positioning in Vehicular Ad-hoc Network

In a Vehicular Ad Hoc Network (VANET), before reaching a potentially dangerous zone on road, wireless Collision Avoidance (CA) system issues warnings to drivers. Global positioning system can be used to get the vehicle position and which can be shared with other vehicles in the network. In case of emergency not all the vehicle get affected so broadcasting of alert packet is not feasible rather it has to be multicast. Calculating the list of relative vehicles position is depending on the travelling direction, bearing angle and the distance. Recent studies have shown that sparse vehicle traffic leads to network fragmentation which poses crucial research challenge for safety application. Hence in this paper we propose a system which will find out the relative position between multiple vehicles by using Great Circle Algorithm. The improvement in VANET connectivity is made by road side unit which will manage all the vehicle information and detect the failure vehicle and calculate the detail of the vehicles that get affected by the failure vehicle and multicast alert packet to identified vehicles. This will avoid the broadcasting problem.