ROADART Overview

The ROADART project aimed for the development of a reliable, automated system for truck-to-truck (T2T) or truck-to-infrastructure (T2I) communication is safety, since a reliable T2TI/T2I communication platform can be used to warn professional drivers for immediate dangers and to provide crucial information for upcoming road conditions. In addition efficient and safe automated platooning systems drastically cut down GreenHouse Gas (GHG) and other pollutant emissions, while simultaneously they reduce the required transportation costs through fuel savings.

The most important and complicated objective of the project was the demonstration of the ROADART platform. In this platform, all the designed diversity techniques, ESPAR antenna concepts and the implemented hardware and software components were integrated on two MAN test trucks in order to demonstrate the performance improvements induced by the ROADART approach.

Novelty 

The novelties incorporated in the ROADART platform cover many areas. First of all is the pattern diversity concept, which was made feasible by implementing special reconfigurable antennas, the so called electronically switched parasitic radiator (ESPAR) antennas. Thus the system is adapted dynamically to its environment. The smart RF section cooperates with smart digital algorithms that run on software defined radios to select the best pattern combination and provide an efficient and reliable solution. Finaly a novel localization algorithm that is based on the input from the truck sensors offers high localization accuracy in environments where there is no navigation system available.

The Platform

The designed and implemented antenna in the ROADART project is an electronically switched parasitic radiator (ESPARs). It provides pattern reconfigurability, low manufacturing costs, reduced complexity and smaller size. Thus the produced three radiation patterns from a single ESPAR allow for selecting the most suitable pattern configuration for a corresponding scenario. A pair of this antenna is mounted in each side mirror of the truck. Aside the antennas, the ROADART communication platform of each truck consists also of the following parts. First, two RF Modules (one per truck mirror) responsible for reception and transmission of the Truck-2-Truck Packets and part of the signal processing. And second, one communication unit (per truck), which is responsible for further processing and routing of the detected packets sent from the RF modules via Ethernet. Most of the software runs on the communication unit that among others includes the diversity engine, the ITS-G5 stack, the localization engine etc.

Diversity Engine

In ROADART, a novel Diversity Engine tailored-made for Truck-to-Truck (T2T) and Truck-to-Infrastructure (T2I) Communications was developed. The antenna (or spatial) diversity was used as the base of the diversity engine and additionally, a beampattern selection scheme was implemented with the use of reconfigurable antennas. The beampattern selection scheme was able to produce an omnidirectional pattern for broadcast - multicast transmission/reception as well as a directive pattern that can improve performance for several applications like CACC and platooning. The hybrid spatial-beamspace diversity scheme was implemented with the use of ESPAR Antennas with pattern selection capabilities.

Dynamic Reconfiguration

An important and complicated objective of the project was the implementation of the Diversity Engine and especially the Dynamic Reconfiguration of the antennas for the ROADART platform in order to track the radio channel fluctuations in the complex T2T communications environment. Three dynamic reconfiguration modes were developed with full reconfiguration in all antennas and with omni support from one RF chain per mirror. The engine monitors continuously the SNR and reconfigures the patterns when it reduces below a threshold while reconfiguration is performed in two stages in order to ensure compatibility with the radio standard.

Geonetworking Mode

In ROADART, one of the developed and demonstrated diversity operation modes is the Geo-tracking, Geo-networking scheme. According to that, the patterns are selected based on the geometry of vehicles in a given time instance. This information is achieved through the ROADART localization engine, where each vehicle is aware of its position and additionally through the received ITS messages, which reveals the position of the cooperating vehicles. The engine calculates the relative heading angle between the vehicles and each vehicle decides on the combination of patterns that best fits the geometry of the vehicle. In this video, the behavior of the algorithm during an overtake maneuver that took place in the A55 Highway in Germany is presented.

Localization Technique

A novel localization technique was developed in ROADART attempting to improve positioning performance in challenging conditions like tunnels. The Localization Engine, which is based on an Extended Kalman Filter, is able to operate with variable sampling rates of incoming data from various heterogeneous sources such as the GPS, Truck Sensors as well as positioning information from cooperating vehicles through ITS services. Then, the localization result is distributed to all interested parties with the use of the Data Distribution Service (DDS) protocol. The ROADART engine operated with less than 2m accumulated error for a 2km in-tunnel course during demonstration.

CACC: Cooperative Adaptive Adaptive Cruise Control

The focus of the time- and safety-critical Cooperative Adaptive Cruise Control (CACC) is to obtain robustness on the application layer against any wireless communication impairments, in particular packet losses and (time-varying) latency, utilizing ROADART communication system characteristics. The development of a model predictive controller (MPC) involves a prediction horizon, which may be used to predict the future output behavior of the leading vehicle. Thus CACC continues its functionality, increases the availability and robustness of the system by sharing look-ahead information and potentially enables shorter inter-vehicle following distances.

ROADART is attending TRA 2018 in Vienna from 16-19 April 2018.

Join us  at the workshop on “Real time ITS services towards a safer and more efficient road transport”
Place of the workshop: TRA conference venue, room Galerie 5+6.
Date: 17/04/2018, from 9:00 to 12:30

You can find us also at the MAN booth in the mall, at level 0, in front of the press lounge / TRA Visions room / stolz  rooms.
This is also the entrance/exit of the Outdoor Area A, where a MAN eTruck prototype will be also exhibited.

ROADART is attending EuCAP 2018 from 8-13 April 2018 and organize the session SW02 on "Multi-antenna concepts and communication techniques in C-ITS systems: From Theory to Application"

Our List of invited speakers and the topics that will address:

• Leonidas Marantis, University of Piraeus, Reconfigurable Antennas for V2V/V2I
• Jacco van de Sluis, TNO Automotive, Wireless Communication for Robust CACC Messages
• Seiko Abreu, InnoTec21, Geo-Temporal Awareness
• Monica Navarro, CTTC, cooperative networks and open data; positioning strategies
• Ioannis Sarris, u-blox, V2X Simulation Platforms
• Gerald Artner, TU Wien, Smart Vehicular Antennas in Chassis

RADIO CHANNEL MEASUREMENTS

The ROADART T2X channel modeling task was based on wideband multidimensional channel measurements at 5.9 GHz ITS frequency band. Three channel measurement campaigns were undertaken; two in Germany and one in Greece.

Measurement campaigne, Peloponnese Greece, 10/2017

The third measurement campaign took place in Panagopoula tunnel, a 3.179 meters long tunnel located in Peloponnese-Greece. The setup of the measurement equipment was designed in a way that one can measure a more generic channel that is free of shadowing in T2I links. The channel sounder used could measure a 2x4 MIMO channel with a signal bandwidth of 25 MHz. The excitation signal was an OFDM-like transmitting signal with the subcarriers uniformly distributed in the measurement bandwidth and it was generated using Frank-Zadoff-Chu sequences. The signal was cyclically extended. The technique that was used for the MIMO measurement procedure is a popular fast switching method with a time division basic principle scheme that alternates between the transmitting antennas for each one of the receiving antennas.

Measurement campaigne, Duesseldorf, 04/2017

The second measurement campaign took place in Kamp-Lintfort, in North Rhine-Westphalia. The measurement equipment setup was designed in order to measure a more generic channel that is free of shadowing in T2T and T2I links. Therefore, to avoid the signal blocking by the vehicle structures, the antennas were placed well-above the roof of two vehicles, at a height similar to the one used in the first measurement campaign, using car roof racks and machined custom made equipment. The channel sounder used could measure a 2x4 MIMO channel with a signal bandwidth of 500 MHz. Due to the unavailability of two baseband generators, a different set of excitation waveforms were used. The measurement routes that were followed included mainly a highway with varying traffic density along its route, two fairly long tunnels and some areas with smaller roads.

Measurement campaigne, Munich, 10/2016

The first campaign took place in the outskirts of Munich, in the district of Dachau, Bavaria.The measurement setup and the placement of the antenna array on the trucks considered the impact of the truck container and the corresponding shadowing of the signal. The channel sounder used could measure a 2x8 MIMO channel with a signal bandwidth of 500 MHz. The excitation signal is an OFDM-like transmitting signal with the subcarriers uniformly distributed in the measurement bandwidth and it is generated using Frank-Zadoff-Chu sequences. The signal is cyclically extended. The technique that was used for the MIMO measurement procedure is a popular fast switching method with a time division basic principle scheme that alternates between the transmitting antennas for each one of the tetrad of the receiving antennas. After the reception of ten frames from the first tetrad, RF switch was taking place, and another ten frames were received from the second tetrad. The overall sequence contains 7000 samples with a sampling frequency of 7 GSamples/sec. The routes followed included mainly a highway with varying traffic density along its path, a fairly long tunnel, and some areas with smaller roads that were used for manoeuvring the trucks in order to get back to the highway.