Research In Wireless Ranging Based On Roadside Unit And Vehicular Integrated Positioning Algorithm

"Internet plus" strategy in the automotive industry includes the core contents and development direction of the Internet of Vehicle (IoV). IoV, the new generation mobile vehicle network, is based on the inner-vehicle network, the inter-vehicle network and the vehicular mobile network which is capable of realizing vehicle active safety, real-time traffic prediction, intelligent traffic management, intelligent dynamic information services and vehicular intelligent control. Some insiders have estimated that the IoV economy will grow as a share of nearly $53 billion market in 2018 and will form as IoV economic circle based on the vehicular nodes. Many new IoV applications, such as the vehicle collision avoidance system, the urban real-time traffic prediction and guide systems, the vehicle real-time monitoring system, ad recommendation system, intelligent vehicular maintenance platform and so on, are expected to grow rapidly. These new services are more and more humanized, intelligent, efficient and user friendly. The vehicular positioning system plays an important role as an elementary function in the IoV application. The existing vehicle positioning solutions do not satisfy the requirement of the IoV application due to low accuracy and limited coverage area which emphasizes the urgent demand for a new vehicle positioning technology for the IoV applications.In this paper, the key technologies of vehicular positioning in the complex urban environment are positioned as a research object. This project, under the background of the dedicated rapid development of the short-range communication technology, abides by the IEEE802.11p physical layer protocols and the IEEE1609.X application layer protocols and realizes the high-accuracy vehicle positioning by utilizing the communication between the vehicle and the Roadside Unit (RSU) where the satellite positioning does not work well. The authors have designed the RSU-based vehicle wireless positioning algorithm by developing the basic framework of the vehicle wireless positioning system and providing the detailed provisions from radio frequency channel, data frame structure, network topology, and so on. Meanwhile, the authors have also investigated the information fusion algorithm for RSU-based positioning system, multimode global navigation satellite system (GNSS) and dead reckoning (DR) system. The fusion algorithm has taken advantage of three systems1 merit to provide high-precision and highly reliable positioning service in a complex urban environment. The main research contents and contributions of this paper are as follows:(1) RSU-based One-Sided Synchronous Two-Way Ranging algorithmIn certain areas, such as the underground parking lots, tunnels and urban center with dense skyscraper, GNSS positioning service is unavailable. The authors explored the RSU-based vehicular wireless positioning system to improve the accuracy and reliability of on-board positioning system in these area. The authors have particularly focused on resolving the following four problems in the vehicular wireless network positioning:the strict synchronization time between the vehicle and all RSUs, the timing system inconsistency between the vehicle and each RSU, the rapidly changing network topology and the low-quality wireless network. Firstly, RSU-based one-sided synchronous two-way ranging (OSS-TWR) algorithm is proposed in ideal network environments and secondly, the authors have improved this ranging algorithm:a dedicated time-interval measurement module is designed to improve the time-interval measured accuracy, and a novel time-recordable back off algorithm is proposed to improve the measured accuracy of the ranging algorithm in network collision situation. Finally, the Extended Kalman Filter (EKF) algorithm has been used to filter out the distance value noise, which has significantly reduced the influence on the ranging. In the end, the vehicular coordinates have been calculated by using the trilateral positioning algorithm. The experimental results have demonstrated that the positioning accuracy of the OSS-TWR algorithm is fully capable of meeting the application requirements and has more adaptive capability than the Symmetrical Double-Sided Two-Way Ranging (SDS-TWR) algorithm proposed by IEEE802.15.4 in area where GNSS positioning service is unavailable.(2) The research on RSU/GNSS/DR integrated positioning information fusion algorithmThe GNSS subsystem has high positioning accuracy if the signal is in the light of sight (LOS) area, but has very poor accuracy in case of none light of sight (NLOS) area because it cannot capture enough satellites. The DR positioning subsystem is an autonomous positioning system that has the advantage of performing well without influence on the external environment but also has the disadvantage of accumulated error with the passage of time. Therefore, the DR subsystem needs to be periodically reset to its initial position. The RSU-based positioning subsystem only works well in RSU deployment area and deployment of many RSUs increases the cost, thus, it is only suitable for small areas, such as underground parking lot. Unfortunately, the single positioning system can not provide positioning services for vehicles in complex urban environments. One of the feasible methods is the information fusion within multiple positioning systems. In order to solve the vehicle positioning problem in complex urban environments, the authors have proposed RSU/GNSS/DR integrated positioning algorithm based on the Federal Kalman Filter (FKF) information fusion algorithm. The integrated positioning algorithm offers more precise and reliable positioning service. Firstly, a) the positioning rationale of three subsystems is analyzed in detail;b) the linear Kalman Filter is used as local filter for GNSS and RSU-based positioning subsystem; c) the EKF is utilized as local filter for the DR positioning subsystem; d) system state and observation equations are built; and e) the information fusion for three subsystems is realized. Secondly, a two-level fault detection method combining the residual chi-square test with redundant hardware test is developed. Residual chi-square test is highly sensitive to act directly on the fault observation, the inspection is very effective for the hard fault. The hardware redundancy based detection method can effectively detect the soft fault in the information fusion system but this method will increase the cost. In the proposed solution, both cost and system reliability have been considered simultaneously to improve the system positioning accuracy by employing low-cost accelerometer to realize partial redundancy rather than full redundancy. Finally, a self-adaptive information sharing coefficient dynamic adjustment method based on the subsystem positioning accuracy is developed to realize the fault isolation for the fault-subsystem and the reconstruction fault-free subsystem. The experiments have verified the positioning accuracy and have demonstrated that the reliability of the positioning system based on the integrated algorithm has met the road-level positioning requirement for IoV applications.(3) The platform designation of integrated positioning systemThe new emerging IoV application has raised new claims in the vehicle positioning field, such as the vehicle positioning in underground parking lot. The traditional vehicle navigation device cannot meet the requirements of new applications in terms of processing speed, interface design and position accuracy. The on-board device design philosophy has been derived from a single positioning requirement to the requirement diversification and intelligential, such as real-time communications, natural voice recognition etc. For these situations, firstly, the functionality and the requirements of the onboard smart devices are analyzed and the overall system framework is proposed after obtaining an excellent balance between performance and cost. In accordance to the integrated positioning requirements, the authors have designed the hardware of the RSU-based positioning subsystem, the multi-mode GNSS positioning subsystem and the DR positioning subsystem and have addressed the correct rules for the DR subsystem. The vehicle ranging and the RSU/GNSS/DR integrated positioning algorithms have been verified using this hardware platform.