Evaluation of hybrid positioning scenarios for autonomous vehicle applications

Autonomous vehicle applications, such as assisted driving, collision avoidance or platooning, demand precise, reliable and secure localization in harsh environments such as tunnels and urban areas. In order to solve these positioning challenges, multiple navigation technologies have to be fused. These technologies are typically based on Global Navigation Satellite Systems (GNSS), radar, camera, inertial sensors and signals of opportunity. Indeed, the exploitation of cellular communications for positioning is of special interest due to the advanced physical features of fourth generation (4G) and fifth generation (5G) mobile networks. The objective of this work is to assess the nominal performance of hybrid GNSS and 4G Long Term Evolution (LTE) solutions, based on assisted or opportunistic approaches. For this purpose, a software tool is presented to evaluate different hybrid scenarios, by using field GNSS and simulated LTE observables. The assisted LTE approach is shown to outperform the opportunistic approach, by avoiding network inaccuracies. The proposed hybrid solutions achieve full position availability with a position accuracy around 10 meters at the 95% of the cases in urban and road scenarios. Future work is still necessary to fulfil the stringent location requirements of autonomous vehicle applications, by adopting a hybrid fusion of advanced technologies such as 5G networks, precise GNSS and inertial sensors.