Designing location and mobility aware solutions for message dissemination in vehicular ad-hoc networks

• Main Author: Hosseininezhad, Seyedali
• Language: English
• Published: University of British Columbia 2016
• Online Access: http://hdl.handle.net/2429/59067

Vehicular Ad-hoc Network (VANET) is one of the potential solutions for providing low cost and best efforts connectivity platforms which can coexist with cellular and WiFi technologies. Due to high mobility of nodes and short lifetime of links, server-to-client and server-to-server communications become challenging. Effective context-aware broadcasting of information to the destination areas is also a challenge in VANET’s. It is usually assumed that the information about these areas is known a priori, either by a centralized source of information or by the entire set of vehicles. However, in reality, this information may not be available on demand. Message dissemination using partially known location information can reduce the overall throughput of the network. We propose to enhance the performance of server selection in a heterogeneous VANET by taking link reliability into consideration in the server selection mechanism.Thereby, extra client-to-server hand-offs are avoided and the need of serverto-server synchronization is reduced. We also propose an adaptive broadcasting scheme based on distributed reinforcement learning. In this scheme, vehicles collaboratively tune the rate of their message broadcasts based on the network dynamics without any initial knowledge about the geographical distribution of Areas of Interest (AOI). The proposed approach enables a more practical implementation of distributed context-aware broadcasting, which requires no global information and only partial synchronization. In our third contribution, we propose a message dissemination algorithm for location-aware services in vehicular networks. The objective is to reduce information delivery time in intermittently connected urban vehicular networks by using historical mobility information of vehicles. Roads are divided into dense and sparse paths based on observed traffic density and vehicles share their current knowledge about fastest possible message delivery time to contouring dense roads. We evaluate and compare our methods to well known and related literature by running network and mobility simulations using tools such as network simulators NS2, NS3 and Matlab. Mobility simulations are done using SUMO simulator. === Applied Science, Faculty of === Electrical and Computer Engineering, Department of === Graduate