Three trends affecting the way automotive RF engineers test systems
Many varieties of wireless communication technology have made this evolution of the car possible. They include GPS for satellite navigation, cellular (mobile telephone) technologies for communication and access to the internet, Wi-Fi for internet access and car-to-car (C2C) communication, DSRC (Dedicated Short-Range Communication) for automatic payment of tolls and parking fees, Bluetooth for hands-free communication, and several others.
The evolution of the car is far from ended, however, and so now RF engineers in the automotive sector are facing new development challenges which are going to change the scope, duration and complexity of the RF testing carried out on components, modules and complete vehicles.
This article highlights three big trends that automotive engineers in charge of RF testing will need to take account of in the years to come. From both a personal and a corporate point of view, it will be beneficial to address these emerging requirements early: They are:
- The increasing application of rigorous functional safety concepts to RF systems
- The requirement to test highly dynamic wireless networks in which connections and routing are changing second-by-second
- The operation of the car as a piece of mobile telephone user equipment, just like a handset and as a bridge application for e.g. eCall application..
Testing RF systems for functional safety
Today, the wireless interfaces in a car, and their applications, are important and need to function well if they are to give the car’s users a good experience. But none today is actually safety-critical: the driver has full control of the car’s motion.
This is starting to change, as car manufacturers introduce increasingly sophisticated driver-assistance systems. Eventually, it seems inevitable that fully autonomous self-driving vehicles will become a reality.
In the development of autonomous vehicles, car manufacturers will clearly implement exhaustive testing regimes. These will comprise public road testing, virtual testing, fail-safe testing, simulations, traffic scenario testing, safety and crash testing, cyber-threat testing and other categories of tests. Governments and citizens will want to be sure that these vehicles meet exacting regulatory, legal and technical requirements.
And while this has an impact most obviously at the level of the complete vehicle, RF systems will also come under closer scrutiny, since they will become safety-critical when used in autonomous vehicles.
So what is meant by a self-driving car? A fully autonomous vehicle can drive to a specific location in real traffic without the intervention of a human driver. Advanced Driver Assistance Systems (ADAS) are systems which either help the driver to drive safely – for instance by alerting the driver when dangerously close to the vehicle in front – or which actually enable the self- driving process.
Radios used in autonomous cars will include radars for measuring the distance to the vehicle in front, and that vehicle’s speed and direction of travel; and – as will be described below – car-to-car communications for sharing information about road and traffic conditions. In an autonomous vehicle, these radios are safety-critical. For this reason, testing radios according to their technical specification, as automotive suppliers do today, will no longer be adequate. There will also need to be processes which support compliance with the requirements of the ISO 26262 functional safety standard.
ISO 26262 prescribes the requirements for guaranteeing the functional safety of a system, starting from its specification, through design, implementation, integration, verification, validation, and finally to production release. Under the specifications of ISO 26262, equipment is required to achieve a certain ASIL (automotive safety integrity level) grade: this is a risk classification scheme defined by the standard.
The application of ISO 26262 to systems that include a radio is going to dramatically intensify and prolong the testing that RF engineers will need to perform. RF engineers are going to have to develop models which capture and characterise every possible risk which a failing or malfunctioning RF system presents to the operation of the vehicle. They will need to exhaustively catalogue the failure modes of the radio. And they will then need to devise and document testing regimes that verify, with a very high level of confidence, both the risk of failure, and the way in which the radio will handle each failure mode.
In other words, the complexity and duration of RF tests will be far beyond those experienced by most RF engineers working in industry today.
Testing for dynamic network topologies
As has been shown, ADAS and driver information systems of various kinds are going to become increasingly important features in passenger vehicles. Today, these make considerable use of cameras. But in low visibility, for instance because of heavy rain or snow, their operation might be impaired. Even road markings might be impossible to see in snow or ice.
To provide supplementary or more reliable information about road conditions and the operation of the vehicle, car manufacturers are going to implement comprehensive C2I (Car-to-Infrastructure) and C2C communications systems, based on the IEEE 802.11 standards and mobile telephone technology.
Mobile telephone networks are useful for automotive communications because they already cover large parts of the globe, are highly standardised, and provide a robust communications link for fast-moving user equipment.
Legacy technologies such as 2G and 3G will be needed for in-car telephony, browsing the internet and the eCall (European) emergency communications system, described below. But their relatively low bandwidth and high latency make them unsuitable for real-time applications such as the control of autonomous cars. New mobile phone technologies however – LTE (4G) and 5G – will meet the requirement for real-time C2I communications.
In C2C communication, the technology for co-operative Intelligent Transport Systems (ITS) is derived from the IEEE 802.11 wireless local area networking (WLAN) standard which is also the basis for Wi-Fi. A specific frequency spectrum in the 5.9GHz range has been allocated to it in Europe, in line with similar allocations in the US. As soon as two or more cars or ITS stations are in range, they will connect automatically and set up an ad hoc network in which all ITS stations know the location, speed and direction of the surrounding stations, and will be able to share messages, warnings and information.
As the communication range of a WLAN connection is limited to a few hundred metres, every vehicle also acts as a router for message forwarding. The routing algorithm will be based on the position of the vehicles, and can handle rapid changes to the ad hoc network topology (see Figure 1).
The implementation of C2C wireless communications thus calls not only for robust RF performance, a physical layer function which can readily be verified with the use of an instrument such as a spectrum analyser – the MS2830A from Anritsu is a good example. The system designer must also implement a sophisticated test plan for verifying the protocol layer performance, to show that it can handle rapid changes in network topology without dropping packets or losing connections. As above, standard test specifications might not capture the full depth of testing required under the provisions of ISO 26262, and this looks set to demand a new approach to the design and implementation of test routines from automotive RF engineers.
eCall: another safety-critical technology
The European Union’s eCall system, and the similar ERA-GLONASS (in Russia) combines mobile communications and satellite positioning to provide fast, reliable assistance to motorists in the event of an accident.
Both systems rely on satellite location data, the first on GPS, the latter on GLONASS. When in-vehicle sensors trigger events such as airbag deployment, eCall automatically transmits location information in the form of an MSD (Minimum Set of Data) to a Public Safety Answering Point (PSAP), and opens up a voice and data channel via an in-band modem.
The operation of the GSM chipset, the modem, and the entire eCall system must be exhaustively tested in the development phase, and its performance verified in production, to provide confidence that it will communicate reliably in all specified conditions. Just like a mobile phone handset, the in-band modem must be able to operate in the presence of multiple sources of interference, or with a weak signal, and must correctly implement a complex protocol for establishing and maintaining a voice/data connection.
Ultimately, automotive system suppliers will have to verify the performance of eCall systems (and other in-car mobile phone technologies) on the road, in a live network. But the use of a network simulator before live testing takes place allows the manufacturer to test in the laboratory every aspect of an eCall system’s interaction with any mobile network globally. Isolated from a live network, the simulator enables the tester to perform repeatable tests in which failures can be accurately attributed to a known cause, without interference from the random and uncontrolled events that occur in a live network.
An instrument such as the MD8475A from Anritsu is suitable for this: it operates as a base station simulator, supporting the 3GPP protocols in operation today, and beginning from legacy GSM up to the latest LTE-Advanced standards. Through a user-friendly graphical interface (in the case of the MD8475A, this is called ‘Smart Studio’), the device designer can quickly implement hundreds of pre-defined test routines. It also provides an environment for the creation of abnormal network behaviour. Furthermore, specific software packages allow its enhancement to an eCall test set with an independent PSAP environment.
Three big trends, then, look set to bring upheaval to the testing practices of the automotive RF engineer: the introduction of functional safety requirements, the testing of dynamic network topologies, and the application of handset test routines to the car. By being prepared in advance, the RF engineer will be in the best possible position to benefit from these changes.
About the author:
Franz-Josef Dahmen is Field Applications Engineer for Anritsu GmbH, Munich, Germany.