Research On The Key Technology Of Service Equalization In Vehicle Heterogeneous Network

Science and technology serves life. The development of Internet changes the way of life subtly. People hope to enjoy superior quality of the Internet services all the time. The demand has promoted the popularity of smart devices in various fields. Cellular network works as a facility which covers the largest area and can be accessed in the most flexible way, sustaining the enormous pressure of the IP flow naturally. Vehicular Adhoc Network (VANET) which deploys in a wide range is parts of Intelligent Transportation System (ITS). VANET takes the advantage of IEEE802.11p standard realizing the communication between vehicles and the road facilities. The network is actually a wireless network used in a specific scenario. We hope to exploit the dedicated network to offload the IP flow which is supposed to route to core cellular network when users are in a moving vehicle. This method can reduce the pressure on the cellular network, while providing a better user experience.The research work is composed of four parts. Firstly, we design the integration architecture between LTE and VANET based on 3GPP standards and the application requirements. In this part, we propose the "logically tight coupled" concepts and design the related network elements, interfaces, protocol stacks, and signaling to make the communication between the two independent networks possible. The integration architecture establishes the foundation of this thesis. Secondly, we propose a local traffic offload strategy. The strategy is based on the way of separating the signaling and traffic data. The core cellular network elements can still control user behavior and charging through the received session signalings. Traffic data will bypass the core network based on the the geographical information of the user and route to the local wireless network. This strategy will greatly reduce the amount of traffic in the core network. Thirdly, the core network selects different strategy under different overload conditions. When the core network is in a severely congested state, most of the services that attempt to transmit will be switched to the local wireless network and use "fast recovery" strategy to get rid of the congestion as soon as possible. When the congestion is slight, some core network elements will select a more conservative way which could distinguish service types. By identifying the underlying signaling, the services which are not sensitive to the delay will be offloaded in priority. Fourthly, a mathematical model is proposed to maximize the system throughput in a basic service set. The fourth section models the vehicle moving scenario and the Enhanced Distributed Channel Access (EDCA) mechanism in a non-saturated condition, and constructs a three-dimensional Markov chain to derive a mathematical expression of throughput. According to the explicit expression, we analyze the impact of speed and EDCA behavior to the system throughput and derive the appropriate solutions. Finally, we make use of the network simulation software NS3 to simulate the system performance of the optimized scheme. The simulation results show that when the cellular network is congested, the wireless network can play a good role of bench. For certain types of service, the local wireless network can provide them a superior performance.