Technologies and Evaluation Metrics for On-Board Over the Air Control

Abstract: This project has been carried out at the Electronic Embedded Systems Architecture Department at Volvo Construction Equipment (VCE), Eskilstuna, Sweden. It forms the baseline for a stepwise systematic research initiative to convert wired technologies used for certain in-vehicle control and communication components to wireless technologies. In-vehicle wireless networks are being increasingly improvised and researched to minimize the manufacturing and maintenance cost of the total amount of wiring harnesses within the vehicles. Fault tracking and maintenance becomes convenient within a wireless network. Wireless intra-vehicular communication provides an open architecture that can accommodate new components and applications. One such usability has been studied in this thesis for the Display Control Unit of Volvo paver machines. A newly designed hardware demands new technologies to ensure operator safety, security, comfort, convenience, and information. The research conducted in this thesis takes into account five probable use-cases in terms of control and communication around the new hardware, and studies suitable wireless technologies that could replace the wired technology that will be used. From a detailed literature study and the specifications provided by VCE, WAVE/IEEE802.11p and DMG/IEEE 802.11ad have been selected as optimal candidates. These two have been modelled at the physical layer of the system. After comparing the results of WAVE for 4 different channel models and 8 different coding and modulation schemes, it has been found that 1/2BPSK(3 Mbps) and 1/2 QPSK(6 Mbps) are optimal for the 3 Rician Fading Channels of Rural LOS, Urban LOS, and Highway LOS. For use-cases that involve larger distances and a large exchange of control signals, WAVE is a good choice. DMG has 19 modulation schemes for Single Carrier Modes and some of them are extremely robust at low SNR. Around SNR 20, it shows lesser packet errors than WAVE. The lower error rate is also evident from the BER values. For use-cases that involve smaller distances and a lot of image data, DMG is preferable. The work, however, does not study the safety and security aspects. Thean alysis and modelling are theoretical being based on literature studies and the necessary parameters provided by VCE. The model needs to be evaluated against field studies and practical measurements with a prototype before implementation inside the paver.