mainstream timing technology
The timing process is realized by means of communication. According to different electromagnetic wave frequencies and transmission methods, there are several mainstream timing methods, mainly including short-wave timing, long-wave timing, low-frequency time code timing, telephone timing, TV timing network timing, satellite timing, etc.
Network time service refers to the time service provided by the time server, and the designated time server is used as the reference time source on the network to provide time service. All networked devices (including computers, servers, network devices, etc.) in the LAN can be synchronized with the time server through the network to achieve the purpose of time service.
As the name implies, satellite timing is to provide timing for ground equipment through satellite systems, collectively referred to as GNSS (Global Navigation Satellite System), mainly including GPS system (US), Beidou system (China), GLONASS (Russia) and Galileo (Europe).
Each satellite can send precise time to devices within the coverage area by transmitting satellite signals. The ground equipment can realize the time synchronization with the satellite by receiving the timing information through the proprietary equipment, so as to achieve the purpose of timing.
Each satellite needs to be equipped with an atomic clock, and the satellite system needs to cover the whole world in order to achieve high precision and full coverage.
Compared with the various types of timing technologies mentioned above, GNSS timing accuracy is relatively higher. In addition to accuracy, GNSS satellite timing has the advantages of global coverage and simple application. It is currently the most widely used and most popular in the world.
NTP time synchronization technology
NTP time synchronization technology (Network Time Protocol) establishes a designated clock source on the network to provide users with unified time service through the network. The accuracy of NTP time synchronization can generally reach tens to hundreds of milliseconds. NTP provides services in the form of client and server. When the client needs clock synchronization, it sends an NTP clock information query packet to the server. After receiving the query packet, the server queries its own local time and returns the NTP information packet. The client receives the time returned by the server. Both packets contain sending time and receiving time. Through these 4 times, the time deviation and network delay between the client and the server can be calculated.
For general applications, the timing accuracy of NTP tens to hundreds of milliseconds is sufficient, but in vehicle-road coordination, the delay requirement from roadside perception of traffic participant information to sending to ICV is generally within 200ms, so that the information received by ICV can be used by the vehicle decision-making system to ensure that vehicles have enough time to deal with traffic conditions. Therefore, for the NTP clock synchronization technology, the timing accuracy error, plus the network delay and computing system operation time, the overall delay is difficult to guarantee in the vehicle-road collaborative application.
Therefore, the IEEE 1588 synchronization protocol with higher timing accuracy is considered.
Application of PTP technology in vehicle-road coordination
Introduction to PTP technology
The PTP protocol is defined by the IEEE 1588 standard. The timing accuracy of the PTP protocol can reach the microsecond level in the measurement and control system, which is much higher than the accuracy of the NTP protocol. In addition to higher precision, the IEEE 1588 protocol is based on Ethernet and TCP/IP protocols with low network protocol overhead, occupies less network resources and computing resources, and has lower deployment costs.
The PTP protocol calculates the network delay and clock deviation between the master clock and the slave clock through the interaction of time messages between the master clock and the slave clock, and achieves the purpose of clock synchronization through the correction of the deviation and delay.
NTP is a standard protocol working at the application layer. This protocol uses round-trip messages to estimate the time spent in network transmission by messages conveying time information, and then estimates the time offset between the slave clock and the master clock, so that the slave clock can calibrate the local time.
The time stamps of the four moments in NTP are all at the application layer of the client/server, and record the time when the message passes through the application layer. The data transmission delay includes the data transmission delay in the network and the processing delay of the computing system for the data. When the transmitted data frames are equal in length, the network transmission delay can be considered equal, but the processing time of the computing system includes not only the processing time of data, but also the time that is difficult to estimate in many aspects such as fault repair and program calling. In reality, the data processing capabilities of the client computer and the server computer are very different, so that the estimation of the message equality between the master and slave clocks is not rigorous. Therefore, NTP will adopt the method of sending and receiving time stamps multiple times, and use statistical methods to reduce errors.
The PTP protocol and the NTP protocol realize the synchronization between the master and slave clocks are not exactly the same, and have certain particularities:
The PTP protocol is divided into "event message" and "general message", a total of 5 message types, namely synchronization message (Sync), delay request message (Follow_Up), follow message (Delay_Req), delay request response message (Delay_Resp), and management message.
The process of PTP protocol to realize time synchronization through message transmission is as follows:
1. The Sync message is sent from the main clock, and the main clock will predict the sending time of the message at the application layer. This time is not accurate, and the accurate time is stamped at the hardware layer. After recording the exact sending time, send the delay request message, which contains the exact time T1, and record the receiving time T2 and the sending time T1 of the Sync message from the clock.
2. Send the following message from the clock, and record the sending time T3 from the clock;
3. The master clock sends a delay request response message, and the slave clock records the receiving time T4 after receiving the message.
4. By recording the above time, through the conversion of a certain formula, calculate the time offset (offset) and network delay (delay) between the master and slave clocks. After obtaining these two parameters, the parameters can be adjusted to achieve the purpose of master and slave clock synchronization.
T1+delay1+offset=T2；
T3-delay2+offset=T4；
Assume that there is symmetry in the network transmission of the message, that is, delay1=delay2=delay, so the network delay delay and time offset offset can be calculated to realize the time correction between the master and slave devices.
IEEE 1588 High Precision Synchronization Implementation Mechanism
Application of PTP protocol in vehicle-road coordination
In recent years, with the development of intelligent networked vehicles, single-vehicle intelligence and vehicle-road coordination have become the two main lines of exploration and development of intelligent networked vehicles. Each device or system in the communication network has its own clock. Due to manufacturing processes, clock frequency differences, environmental changes, etc., with the operation of the network, the clock value of each device or system will shift, resulting in inconsistent clock values for each clock. In order to ensure the reliability of message transmission, a high-precision and high-reliability clock synchronization mechanism is required in the vehicle-road coordination system.
Vehicle-road coordination technology uses roadside perception devices such as lidar, millimeter-wave radar, and cameras to achieve multi-sensor data fusion through edge computing devices. At the same time, edge computing devices are configured with vehicle-road coordination scene processing algorithms and traffic management algorithms. The roadside intelligent unit RSU sends vehicle-road coordination information to the intelligent networked vehicle OBU. After receiving the roadside information, the intelligent networked vehicle combines the traffic participant information obtained by the vehicle's own sensors to perform unified information processing and judgment. The vehicle driving decision-making and control system makes appropriate driving behaviors to avoid collisions and traffic accidents.
High-precision time synchronization technology is one of the key technologies for vehicle-road coordination. For intelligent connected vehicles, it takes time to synchronize from the synchronous fusion of multiple sensors at the sensing end, to the precise positioning of the vehicle, to the interconnection between vehicles, vehicles and roads, and vehicles and everything.
Intelligent connected vehicles realize holographic perception through the data fusion of their own multi-sensors and roadside multi-sensors. All sensors, controllers, actuators, etc. need to be precisely synchronized and coordinated so that the vehicle can make reasonable and correct decisions and avoid traffic safety accidents.
In vehicle-road coordination, data fusion involving roadside multi-sensors is involved. Generally, the data frequency of sensors is 10-30 Hz, and a data is sent every 30-100 ms. The frequency of various sensors may be different. Therefore, in terms of time, the time of different sensors is first aligned, and the output multi-source target data can be strictly aligned in time, and data fusion is the basis.
The time delay accuracy of vehicle-road communication in vehicle-road coordination is extremely high. The data sent by the roadside needs to be combined with the data of the intelligent networked vehicle itself to make a comprehensive judgment before making a driving decision. Therefore, the time accuracy of the data sent by the roadside needs to be aligned with the time of the vehicle-side data in order to play the role of roadside data and ensure vehicle driving safety.
High-precision positioning is the core technology to realize the safety and stability of intelligent driving vehicles. Time synchronization is the key technology of the positioning system. Whether it is satellite positioning or ground base station enhanced positioning, the basis of orbit determination and positioning is the time synchronization of each observation. If there is a deviation in the observation time, there will inevitably be a deviation in the movement distance. Especially for satellite positioning, a very small time deviation will lead to a huge spatial deviation.
Analysis of Reasons Affecting PTP Timing Accuracy
1. Network asymmetry
During the working process of the PTP protocol, packets are transmitted through the network, and calculations are required to avoid the influence of network delay. In the calculation process, it is considered that the delay of sending and receiving data packets is the same, but in the actual process, due to the influence of network interference, the data processing capabilities of the client and the server are different, and the delay of data sending and receiving is difficult to be the same. Therefore, it has a certain impact on the overall timing accuracy.
2. Network load
When the network is relatively congested, there will be a large packet in front of the school time packet, which will affect the transmission time of the school time packet and affect the calculation of network delay.
3. The location of the timestamp
The PTP protocol supports time stamping at the application layer and also at the hardware layer. After the application layer is time stamped, the message will be sent through the network layer and the MAC layer. There is a deviation between the time of the time stamp and the actual sending time, which will affect the timing accuracy.
4. There are network devices that do not support the PTP protocol in the network, which will affect the operation effect of the protocol and reduce the synchronization accuracy.
Advantages and disadvantages of PTP technology
advantage:
1. The time precision that the PTP protocol can achieve is high, reaching the microsecond level, far exceeding the millisecond level of the NTP protocol, so it can meet the stringent requirements of data fusion in vehicle-road coordination and high-precision, low-latency vehicle-road communication.
2. The PTP protocol has a small overhead on the network and does not affect the data transmission of the existing network.
shortcoming:
1. To obtain ideal accuracy, a professional PTP timing server is required. At the same time, the server can receive satellite signals, which will increase the cost of use.
2. During the use of clock synchronization equipment, the satellite timing signal is unstable. At this time, the local crystal oscillator of the timing server is relied on to maintain the stability of the system time. Different crystal oscillators will affect the timing accuracy.
3. In order for the PTP protocol to be able to realize the operation of the high-precision clock synchronization system, both the network device and the client device need to support the IEEE1588 protocol, which may increase the cost of the overall system.
