PROPOSED SYSTEM A.

Initialization In our experimental study, Vehicles are set with a Global Positioning System (GPS) receiver and thus is able to know its position. It reports about the successive locations of Vehicles at a time interval of around 30 seconds to 1 minute. In simulation, we interpose the distinct trace to get a continuous trajectory. We use the digital to manually deploy access point on the map .As a result, Vehicles can frequently encounter the access point .. These access point on roadside units are implicit to be linked via a wire networks. The default time limit of the data packets is 60 minutes. To demonstrate the signicance of undelivered packets in a VANET, we use the insistent routing algorithm. Using this algorithm, a node would actively imitate a data packet as it encounters another node without holding such packet.

B. Problem Formulation We consider an infrastructure-based VANET consisting of three parts: the set V of moving vehicles, the set of access points, and the mid server. We expect all data packets share the same time limit. set with both 3G radio and short-range radio for inter- vehicle and vehicle-to-access point communications each vehicle can straightforwardly deliver a packet via 3G to access point in the VANET. A shortrangecommunication can only occur when two nodes are in the communication range of each other.The mid server in the system not only gathers data from the network, but also it deals with 3G resources allocations. To this end, it needs to gather the existing network information from vehicles via 3G communications.

C. Empirical Results We now demonstrate the outcome from our experimental study. First, we illustrate delivery delays of individual packets in Fig.3. We can nd those delivery delays of data packets differ signicantly, which range from 0 to 3600 seconds. About 80% packets are successfully delivered within the time limit of 60 minutes, and nearly 20% of data packets fail to be delivered.

Fig. 3.Delivery delays.

To examine the impact of time limit, we differ time limit from 20 to 100 minutes. From Fig. 4, we can see that the delivery ratio increases as the time limit becomes larger.
Fig. 4. Delivery ratio vs. time limit of data packets.

It is simple to realize that the longer the time limit, the elevated delivery probability a data packet has. However, some data packets cannot be delivered to one of the access points even if the time limit is extended enough. In Fig. 5,we can nd that the deliverance ratio increases as the number of vehicles increases. It is logical that the increase in the number of vehicles can increase the connectivity of the network.

Fig. 5. Delivery ratio vs. number of vehicles.

In addition, we can also see from both Fig. 4 and Fig. 5 that a larger number of access point helps increase the delivery ratio in the network. In conclusion, we have experimental that a perceptible percentage of data packets failed to be delivered even with a larger number of vehicles, more access points and a longer time limit. This robus tly suggests that the use of 3G would significantly progress the packet delivery performance in a VANET.

D.Evaluation In this part, we evaluate the performance of 3G-assisted information deliverance algorithm. We mainly consider three performance metrics: 1) Deliverance ratio. It refers to the success proportion of the number of successfully delivered packets to the total number of packets at the end of an experiment of certain time. 2) Average delay. It refers to the average delay for a packet to be received to its goal. It can be calculate by accumulating every delay of each hop. We only calculate end-to-end delay for successfully delivered packets. 3) Utility. It combines both delivery proportion and average interruption. To compare the performance of our algorithm, we evaluate the performance of our system with the following routing schemes S-Random: At the start, it arbitrarily select data packets and deliver them via 3G. E-Random: Arbitrarily choose those data packets which fail to be delivered at the last second and deliver them via 3G. M-Average: The 3G transmission budget is equally allocated to all time slots before a given Time limit. Different simulation results shown are:

Fig. 6. Utility vs. 3G budget Fig. 4. Delivery ratio vs. 3G

Fig. 5. Delay vs. 3G . Fig.7. Delivery ratio vs. number of vehicles