The programmable supercomputer on wheels

Due to the inherent flexibility provided by programmable logic, automotive suppliers can design systems that will work across multiple platforms and, thus, lower overall cost without having to resort to an Application Specific Integrated Circuit (ASIC) or Application Specific Standard Product (ASSP) design. FPGAs also enable highly distributed parallel processing on many incoming signals, providing higher performance than state-of-the-art DSPs for less cost.

We will look at several of these subsystems in automotive applications and specifically how the FPGA can be architected to implement them. For each block diagram this is simply a conceptual view of one possible implementation. The inherent value of FPGAs is that changing architectures from project to project or during the middle of a project is simple. As different needs become apparent, the system can be expanded to take the new requirements into account without a costly redesign of an ASIC or trolling the data books to find the most optimal ASSP for the project.

Interfacing for efficient controller access
Telematics and entertainment systems in the automobile are often combined into an Infotainment Head Unit (see Figure 1).

Figure 1

Generally, telematics systems are used to control automotive electronics systems and display information about these systems efficiently for the driver and other passengers. They combine wireless communication with a Global Positioning System (GPS) and embedded computing to deliver current traffic conditions, driving maps and directions, and speed and fuel efficiency data. In addition, during an emergency, the system can provide rescue services with the exact location of the vehicle.

The entertainment unit provides access to the automobile’s audio player/CD-ROM, where it can store music files such as MP3 and the system’s GPS navigational data. Entertainment systems will often share the control and display with the telematics system and, thus, have some architectural similarities. Through a variety of bus connections, the interface unit provides the controller access to all of the automobile’s entertainment and driver-information systems, such as the onboard computer.

One major subsystem for both telematics and entertainment controllers is the display. Information coming in from various subsystems, such as GPS or engine control for telematics or MP3 and DVD for entertainment, will be displayed on the head-unit LCD display. One of the major functions that can be implemented in the FPGA is the LCD display control (see Figure 2). The LCD graphics controller displays data from these components, such as navigational information and wheel- and engine-speed information.

Figure 2

A major benefit of using FPGAs for the LCD controller function is the speed with which changes can be made. This is useful for accommodating displays with different resolutions or from different manufacturers, and for allowing the display subsystem to be used for more than a single automobile model. This flexibility also mitigates risk in the event of a shortage from one supplier; another supplier’s LCD screen could be used with minor changes to the FPGA design.

With the plethora of functions and interfaces required in these automotive applications, extreme flexibility is required. FPGAs are ideal for many of the applications found in telematics systems. Programmable logic allows automotive designers to customize the interfaces to suit many sources of data. It provides a complementary, flexible system to work in conjunction with the automotive-based ASSP devices. Most of the functionality of the telematics controller can be implemented within an FPGA.

Automotive electronics designers developing telematics and entertainment systems face the challenge of anticipating which features to add to the systems and how they can be regionally customized and updated over time. Automotive manufacturers also must ensure that their products interface with a wide range of equipment, including after-market equipment installed by the customers themselves.

Using readily available soft-core microprocessors, Intellectual Property (IP) cores (such as CAN or PCI cores), and software drivers on the FPGA can significantly reduce development costs and time to market over designing ASICs or discreet designs using ASSPs for automotive applications.

FPGAs address many of these challenges, allowing automobile manufacturers and their tier-1 suppliers to develop and customize flexible telematics and entertainment systems that meet the specific needs of their customers.

Communicating between control units
As automotive electronic systems increase in complexity, they require processors that can control many functions in the car, from Antilock Braking Systems and fuel injection units to cutting-edge entertainment systems. Usually, automobile networks are divided into body and power train control networks, as well as telematics and multimedia subnetworks. A central controller is the core element of an automotive system, enabling the user to operate the different electronic systems and control units. To communicate with these electronic control units, the central controller must have access to all types of buses through the gateway controller, which acts as a router between the different electrical and optical buses in an automobile.

FPGAs provide upgradeable interconnectivity on a large or small scale for bridging between buses or as a complete bus-interfacing unit, enabling communication between various protocols used by different ASSPs. This connectivity can come from integrating the separate subsystems in an automobile or adding connections to consumer interfaces such as Bluetooth, USB, and Firewire.

An automotive gateway controller acts as a router between the different electrical and optical buses, integrating standard multimedia interfaces such as USB, Firewire, and Media-Oriented System Transport buses, as well as connecting the CAN systems (refer to Figure 3). The controller can also interface with leading-edge automotive systems using computer-related interfaces such as Ethernet and Bluetooth.

Figure 3

The gateway controller implemented in an FPGA consists of a soft-core microprocessor used to run the software stacks for the Control Area Network (CAN), Ethernet, FlexRay, and so on. With current FPGA technology, a single >100 MIPS soft-core processor can be implemented in equivalent FPGA logic costing less than $0.50. Adding multiple copies of common or new interfaces is relatively easy using the current generation of system tools available from FPGA vendors.

FPGAs offer an ideal bus-bridging solution; designers can use standard low-cost ASSPs and, with the help of an FPGA, interface them to their specific bus systems utilized by various manufacturers. This advantage increases system volume and minimizes development costs by sharing designs across multiple platforms streamlining the development process.

Generating modulation/demodulation
Automotive Software-Defined Radio (SDR) refers to wireless communication in which the transmitter modulation and receiver demodulation are both generated through software. An SDR receiver has an A/D converter right after the antenna or with one interface unit in between. Software-controlled mixing and baseband processing are done digitally. The main advantage of this approach is flexibility; the software runs on one common hardware platform for any type of receiver configuration.

Several receiver configurations may require a different set of channel processing modules, which then can be reloaded into an FPGA under the control of the SDR controller. Various standard IP cores, such as a Numerically Controlled Oscillator, a Finite Impulse Response, an Infinite Impulse Response, Fast Fourier Transforms, and Constellation Mappers, can be used to implement the channel processing, decoding, and waveform modules needed for SDRs. The ability to mix and match these cores quickly and easily to implement different signal processing algorithms is one of the major benefits of using FPGAs to implement SDR.

The need for receiving multiple protocols is increasing as automobiles become a rolling nexus of incoming information. Starting with AM and FM stations, cars will include receivers for satellite radios, television reception, WiFi, 3G WiMax, Orthogonal Frequency Division Multiplexing protocols, and other uses not even thought of today. Having a flexible system that is software-defined and changeable will be necessary. FPGAs are a critical link in the system design because of their inherent flexibility to be changed even after the system has been deployed.

Examining the differences between FPGAs and ASICs

Because ASICs can be mass produced at lower cost, they can be more suitable for large-volume applications. For certain high-performance applications, however, FPGAs may be the wise choice. Generally, FPGAs offer a number of important advantages over ASICs, including:

  • Increased flexibility during the product design cycle, because design adjustments are simply a matter of changing the software programming file
  • Reprogrammability, even after an end product is shipped to a customer
  • Shorter lead times for prototypes or production parts
  • Elimination of large NRE costs for customers

FPGAs allow customers to order only the number of parts they need, when they need them, allowing effective inventory control. Customers using ASICs usually cannot enjoy such control.

FPGAs: A win-win for the automotive market
Under the hood, inside the passenger compartment and in external diagnostic systems, FPGAs offer a flexible, low-risk path to successful automotive electronic system design – reducing manufacturing complexity and offering optimum cost efficiencies.

FPGAs are currently used in powertrain, auto PC, telematics, entertainment, body controller, and other automotive applications. Proprietary solutions combine FPGAs with optimized IP cores, hard and soft microprocessors, powerful design software, and a variety of development kits to create a complete, easy-to-use automotive electronic design platform.

In one integrated FPGA package, automotive designers can meet all their requirements to win in this highly competitive market, including cost reduction by avoiding extensive Non-Recurring Engineering Expenses and time, as well as minimum ordering costs required by ASICs. They can also develop systems while standards are still being set, without having to wait for the availability of ASSPs. Some additional benefits include:

  • Reprogrammability during the design process, including in- vehicle
  • Reusability of one hardware platform design for various systems
  • Risk and cost management through the suppression of multiple silicon iterations
  • Adaptability to changing telematics industry standards and protocols

FPGAs provide flexibility, adaptability, raw performance, and cost benefits that are new to the automotive industry. The car of tomorrow will not be your father’s Oldsmobile.