Three challenging channel measurement campaigns have been undertaken; two in Germany for highways and tunnels in T2T/T2I communication scenarios and one in Greece for T2I communication scenarios in tunnels. The data collected were processed and a multi-domain statistical characterization was performed using the RIMAX processing algorithm, which was specially adapted for this scope. A Geometric–Stochastic (GS) channel model framework was developed to successfully simulate propagation conditions relevant to the ROADART objectives, as well as other vehicular scenarios. Moreover, ray-tracing models were developed for T2T/T2I multi-antenna channels.
Moreover, various impairment parameters for the T2T/T2I wireless communication systems were identified, which result to reduced performance, e.g. channel spatial correlation, co-channel interference, outdated channel state information (CSI), hardware impairments, non-Gaussian noise, position randomness. For these parameters, analytical investigation studies were performed to quantify their impact to T2T/T2I systems. Moreover, several low complexity techniques have been proposed. More specifically, a new reconfigurable antenna pattern selection scheme is proposed, which aims to reduce the complexity of the traditional antenna selection techniques. A new relay selection scheme was also investigated under different T2T communication scenarios to reduce the signal processing complexity. In addition, a novel extended open loop beamforming (eOLB) technique that is realized through pattern selection is designed and demonstrated via a proof-of-concept experiment. The benefits of adopting both space shift keying as well as spatial modulation in IVC scenarios were also studied. The operation of the ROADART modem and the integration of the extended Open Loop Beamforming (eOLB) scheme with the modem, were validated using the ROADART GS channel models.
Various designs of Electronically Switched Parasitic Array Radiator (ESPAR) antenna for T2T and T2I links were investigated. Special concern was given on the antenna mounting on the trucks. ESPARs exhibit low cost, reduced complexity and compact size, and thus, they are considered suitable for easy integration. Their main advantage is the pattern reconfigurability and the fact that they can support various digital communication techniques, e.g. pattern diversity.
Furthermore, novel localization techniques for conditions where GNSS is not applicable, e.g. tunnels, have been developed. A dynamic filter was implemented and integrated into the ROADART system. The filter is based on an Extended Kalman Filter (EKF) that fuses all measurements and produces improved estimates of these measurements. In addition, a novel approach for increasing the robustness and performance of CACC, in the presence of short periods of packet losses, using a Model Predictive Controller (MPC) in combination with a buffer, was successfully developed, implemented and tested. The final system architecture, including the RF modules, the communication unit and the ESPAR antenna, was realized and is shown in Figure 1.
The use of the RTI DDS platform offered a more stable and professional approach. The PHY layer is split between the RF modules and the communication unit. A new wireless message, which is mostly aligned with the ETSI standards, was proposed to increase the robustness from the application side.
ROADART may contribute next year to the definition of a new CACC-message set, which is a current ongoing topic within the ETSI-group. Furthermore, some of the partners involved in other cooperative ITS projects, exploit the outcome of ROADART within the different consortia. The ROADART results have been communicated through 9 journal, 2 magazines and 16 conference papers that already received 29 citations. Aside the project webpage and the social media, participation in European and National Workshops and panels has also contributed to spread the ROADART results to a wider audience.
