Distributed Sensing And Transmission Of Traffic Situation

In recent years, city transportation networks in China have been greatly improved thanks to our fast economic growth. However, the fast economic growth also brings about explosive increase of vehicle population, which is far beyond the upgraded traffic networks’ capacity, thus makes present techniques of information collection/processing and traffic management in current ITS(Intelligent Transportation System) difficult to handle traffic congestion/jam. To overcome shortcomings of traditional ITS, we have proposed a distributed ITS framework(Vahicular Ad-Hoc NETworks-ITS, VANET-ITS). In VANET-ITS, traffic information is collected and processed mainly by vehicle sensor network encompassing vehicles capable of sensing and forwarding vehicle status and local traffic information – thus acting as sensors and relay nodes – and processed by the receiving vehicles to help drivers and passengers avoid traffic congestion in a timely fashion thus utilizing the road network more efficiently. To achieving these goals, this dissertation has focused on new technologies of traffic situation sensing and transmission with main contributions as can be listed as.Part-I. METHODS AND TECHNOLOGIES FOR DISTRIBUTED TRAFFIC SITUATION SENSING including vehicle positioning, group moving speed estimation and traffic status sensing.Location and group moving speed are important parameters in the traffic situation sensing process. The present positioning methods(such as GPS) are impossible to locate the lane on which the vehicle is moving. It reduces the accuracy of traffic situation sensing for different road grades. Therefore, an RFID based Active Vehicle Positioning method(RAVP) is proposed for solving this problem. With the help of this accurate position information, a Cluster based Vehicle Group speed Sensing method(CVGS) is proposed. In this method, vehicles cluster as a group and submit their position and speed to the cluster head. The cluster head obtains the cluster speed using fuzzy theory. A Traffic Situation Detection using Vehicular Ad-hoc NETworks(TraSD-VANET) is also proposed to sense the traffic situation dynamically on the basis of RAVP and CVGS. Comparing with other methods under different traffic scenarios, the simulation results show that TraSD-VANET can sense the traffic situation dynamically with higher accuracy.Part-II, ANALYSIS AND MECHANISM OF TRAFFIC INFORMATION TRANSMISSION IN VEHICULAR AD-HOC NETWORK(VANET) with non-uniform distribution and high mobility features resulting transient topology and intermittent links.The high dynamic topology and transient link in VANET brings many problems to communication between vehicles, such as broadcast collision and the network hole problem. In this part, the performance of broadcast in 2-D traffic environment and multi-hop connectivity between vehicles were analyzed.A 2-D broadcast scenario of an intersection is built to evaluate the influence on packet reception rate caused by the hidden terminal problem, collisions caused by concurrent transmissions, coverage, carrier-sensing range, interference range of neighbor nodes etc. The formula of packet reception rate is derived considering the above factors. The simulation result proves the validity of the theoretical analysis. Network connectivity is another important performance metric to indicate the multi-hop communication in V2 V. Network hole is the main cause of disconnected links in VANETs. In this dissertation, the appearing possibility and distribution probability of network hole are studied. The delay caused by network hole is also analyzed. The analysis result offers a reference to the length of store-carry time in multi-hop broadcast method which is proposed in the third part.Part-III. Novel BROADCASTING METHODS TO SOLVE COLLISION, NETWORK OVERLOAD AND NETWORK HOLE PROBLEM FOR TRAFFIC INFORMATION TRANSMISSION.Method-1: A Self-Decision-Making Beaconing method(SDMB) is proposed to solve the problem of the neighbor information getting outdated quickly and the overload of VANETs caused by frequent beaconing in high node density scenario. In this method, the beaconing interval of SDMB is adjusted based on node activity and the rate of its neighbor change. Results show that, using SDMB, the successful beaconing rate is higher and the channel load is less than the method used in WAVE model in IEEE802.11 p.Method-2: A proactive network hole discovery method is proposed and the probability of hole generation and successful link establishment are analyzed. A distributed carry-greedy-forward-broadcast(CGFB) algorithm without infrastructure is also proposed to solve the network hole problem in VANETs with the help of the vehicles’ mobility. Results show that, CGFB has greater probability of successful link establishment than Greedy-Forward algorithm and saves more storage resources than Store-Carry-Forward algorithm.Method-3: A space-time limited infection broadcast(STLIB) method is proposed to alleviate traffic congestion. This method broadcasts traffic congestion information to the target road segments within a message lifetime via epidemic forwarding to the vehicles that will be affected by the congestion. This method has been proven to overcome problems such as single flooding broadcast storms, the inability of one-hop broadcasting to cover the entire area needed and the disconnection of multi-hop broadcasting. The simulation results show that vehicular traffic congestion can be alleviated or eliminated using the proposed method.