Research Of Simulation System For Vehicle Ad Hoc Network Based On ONE

With the progress of science and technology and the improvement of living standards, theproportion of the person with mobile devices and private vehicles is increasingly high. Thereforethe vehicle opportunistic network has become a hot research topic in wireless network andintelligent transportation field. Compared with the traditional ad hoc networks, nodes has its owncharacteristics of high-speed moving in the vehicle opportunistic network, so the network maybe show many problem like as intermittent connectivity, segmentation, frequent topologychanges and so on. Transmission network can only rely on data passed between nodes inopportunistic encounter; seriously affect the performance of data transmission of the networkwith the data center.Vehicular opportunistic network protocols and applications for testing and evaluatingeffective in real world environment, which needs the large-scale deployment of communicationwithin the vehicle node may be great costly. There are many methods of network research,including modeling, experimental test and network simulation. Network simulation with thesmall cycle, cost effective, high accuracy, strong operability and so many advantages are widelyused by the academic research for network. In the research of routing protocol and application ofvehicular opportunistic network environment, the network simulation plays an important role.Especially in the node distribution is sparse, the vehicular network simulation can describe thedetails of wireless link better. So using the simulation of vehicular opportunistic network hasbecome a very effective method for researching. This paper combines delay tolerant network, theopportunistic network, social network theory and ONE for the research of vehicularopportunistic networks simulation system.The ONE simulator is developed by SINDTN and CATDTN project, which from the supportof Nokia Research Center in Finland. ONE is a simulation whose engine is based on discreteevents. Each step in the simulation engine will update many modules, so as to implement thewhole simulation function. It can use different routing protocols to simulate the sending andreceiving of the messages, and the moving track record. The main function of ONE simulatorincludes node mobility modeling, communication, routing and message processing, messagetransmission and so on.The ONE simulator in order to make the simulation is feasible, which needs completeabstract or abandon for some aspects of the real world in some times, it is almost a commonproblem in the simulation software. In recent years, some researchers focus on the problems ofmaking the simulation more like real world. Therefore, according to the current problems ofONE simulation system, this paper in the aspects of simulation platform, mobile model, cachestrategy and routing model is proposed to improve the innovation of theory and technology, thespecific contents are as follows:(1) The vehicle network is highly object-oriented, and compared to other traditionalnetworks, it has some unique characteristics such as huge number of nodes, complex informationforwarding model, big node mobility, etc. The OSI reference model and TCP/IP protocol stackcannot be used simply duo to these features. Learn from the OSI reference model and combinedwith the opportunity network’s own characteristics, the physical layer and link layer model isestablished based on the module classification mechanism of ONE.(2) There is a common assumption during the process of simulation that the mobilecommunication equipment is always opened. But in reality in order to save energy, some devicesmay be switched into idle states according to real needs, and there are also some equipmentperiodically detect neighbor nodes. In conclusion, for there is no interference in ONE simulation environment, the communication time is over optimistic. This paper proposed a wireless channelmodel with various forms of channel. The issue of channel loss is divided to the path loss,random signal interference and the mobile node loss.(3) the GUI is extended to the ONE, which is more convenient for using and makes theresearchers no longer use the lots of command of the situation that better for the operation andcontrol. This can put more focus on results rather than simulation system exploration.(4) The driver of the vehicle is human, so the behavior characteristic of the vehicles oftenreflects the behavior of people. Therefore, this topic provides a mobile model of SVMM (SocialVehicle Movement Model) based on social network theory. In a variety of vehicularopportunistic network environment, this new mobile model is used to more accurately reflect themobile model in real world city environment. In addition, for the network segmentation, unstablecommunication time between nodes and message transmission quantity can not achieve an idealvalue; therefore, this paper also proposes a cache management strategy: ISM (IntelligentSubsection Management), and an improved routing algorithm with copies of the messageradiation distribution characteristics according to the cache management strategy.