Calculation and prediction of link duration between two nodes connected, as to create networks for communication system has always been a main area of high degree of importance. With the introduction of Ad Hoc networking concept along with its emerging subsets like MANETs, VANETs and FLYNETs. The issue regarding the calculation and prediction of the node''s link with reference to time is becoming more challenging because of the requirements of different strategies in different scenarios. As well as the changing topology and high speed of node had a great impact of link duration analysis. Geographic routing algorithms have the potency to work with a large network without maintaining the routing table of a node with complete route information, thus decreased the routing overhead. These routing strategies are incredibly suitable for the networks created by the VANETs with their high speeds and changing topology characterization. A number of researchers are involved with the development of routing strategies for Vehicular Ad Hoc Networks (VANETs), which can cope up with the uncertainties of the scenarios associated with the variable speed and high dynamics change of network topologies. With the advent of IPv6 and cheap availability of GPS, system''s geographic routing is becoming a solution for the problem associated with the possibility of the successful routing in VANETs. Disadvantage associated with the geographic routing is its reliability associated with the precise requirement of position information of the node. These networks are quite hard to handle in terms of the presence of such hindrances or may prove to be much more cost effective. Furthermore, time lagging factor’s presence in the routing table of the neighboring node because of the movement of the node position information is a reality which cannot be denied. Hence, the pattern of movement of a node is another issue which should be taken into account before the selection of routing nodes. Opportunely, according to our analysis vehicle movement is not complicated in highway scenarios. By using this, the significant observation, we had designed a movement pattern which can simulate the nodes moving in the highway scenario. Data set of a project group had been used. Project group had gathered data on a high way for bi directional traffic. Roads are divided into 1000 meter segments as to observe the movement pattern and then analysis had been performed. Our highway scenarios are able to capture different events and also connect the vehicles on the road from each other. We have suggested a fresh approach for geographic routing, which exploits the predicted locations of vehicles. We have proposed a stable link duration technique based on the prediction method by using the heading direction and present speed of a vehicle. This can effectively reduce the requested frequency of position update. Special tool in MATLAB has been designed for the testing different scenarios of VANETs by using different entity based mobility models.