Adjustable Structure-Free Data Aggregation Approach On Vanets

With the development of microelectronics, computing and wireless communication technology, smart vehicular sensor system gains rapid evolution and sensor networks are being applied widely in all kinds of applications. Data dissemination is a vital and fundamental problem in vehicular ad hoc networks, and has great influence on performance of kinds of applications on the entire network.In-network data aggregation is a useful technique for applications that require periodic dissemination of information to a large region. Usually, different kinds of applications on VANETs require different degrees of data accuracy. Therefore, data aggregation scheme should adjust the aggregation factor to satisfy the requirements of data accuracy from different applications. Till now the researches about adjustable data aggregation on VANETs was not deep enough. We observe some applications like safe driving requires high accuracy data with small amount in adjacent domain while some applications like route scheduling require more data in a large region which inaccuracy is tolerable. Based on this we formalize the adjustable data aggregation on VANETs problem. Furthermore, we propose data aggregation on one node problem, and prove its NP-Hardness. Then we propose an greedy approximate algorithm ESAS. This algorithm uses only the information that the node has to do data aggregation. It can deeply decrease the amount of the transmitted data, meanwhile it makes sure that the aggregators are accurate enough that are useful to different kinds of applications.In the process of the research about data aggregation on VANETs, we suppose that all data that should be disseminated can be sent in one time slot. But because of the limit of wireless network bandwidth, data dissemination strategy is important to the result of data dissemination. This paper proposes a data dissemination strategy based on data weight, and gives the method of the management of sending data cache. It can make the data dissemination ratio higher in the case of limited network bandwidth.Theoretical analyses and results of experiments show that the proposed algorithm can deeply decrease the amount of the transmitted data, at the same time, it can make the disseminated data precision enough for different precision needs from different sorts of applications. A new data dissemination strategy can improve the data dissemination ratio in the case of limited network bandwidth and make the data dissemination ratio be above 90%.