What are the applications of FPGA in the aerospace field

1 Introduction

Field programmable gate arrays (Field programmable gate arrays, FPGA) is a programmable signal processing device. Users can define their functions by changing configuration information to meet design requirements. Compared with traditional digital circuit systems, FPGA has the advantages of programmable, high integration, high speed and high reliability. By configuring the logic functions and input/output ports inside the device, the original board-level design is placed on the chip. , Improve circuit performance, reduce the workload and difficulty of printed circuit board design, and effectively improve the flexibility and efficiency of design. The advantages of designers using FPGA:

(1) Reducing the demand for the required device varieties helps to reduce the volume and weight of the circuit board;

(2) Increase the flexibility of modifying the design after the circuit board is completed;

(3) Flexible design modification helps shorten product delivery time;

(4) After the device is reduced, the solder joint is reduced, which can improve the reliability. In particular, it is worth mentioning that in the case of higher and higher circuit operating frequencies, the complex circuit functions implemented by FPGA reduce the electromagnetic interference problems caused by inappropriate PCB routing on board-level circuits, and help to ensure circuit performance.

FPGA is also the best way to realize the application specific integrated circuit (ASIC, Application Specific Integrated Circuit) at the present stage. Using commercial off-the-shelf FPGAs to design spaceborne electronic systems for spacecraft such as microsatellites can reduce costs. It is a good way to meet the reliability requirements of spaceborne electronic systems by making use of the rich logic resources in FPGA and designing on-chip redundancy and fault tolerance. At present, with the continuous development of satellite technology, the continuous improvement of user technical indicators and the increasingly fierce market competition, functional integration and light and miniaturization have become a mainstream trend in spaceborne electronic equipment. The use of miniaturization technology can reduce the volume, weight and power consumption of spaceborne electronic equipment, and improve the spacecraft's ability to carry payload and efficiency ratio. The use of highly functional integrated miniaturized devices can reduce the size of the printed circuit board, reduce the number of pads, and help to make full use of redundancy technology to improve the fault tolerance of the system. The key to miniaturization of on-board digital circuits is the selection of devices, including the selection of embedded high-integration devices. Among them, the selection of high-density programmable logic device FPGA is an important implementation method.

At present, in the design of aerospace remote sensors, FPGA is widely used in the main control system CPU function expansion CCD image sensor drive timing generation and high-speed data acquisition. This article reviews the development of FPGA, analyzes its main structure, and reviews FPGAs for aerospace applications. It points out the requirements of aerospace applications on FPGA and its design, focuses on the analysis of the effects of space radiation effects on the reliability of FPGA, and summarizes the reliability design methods to improve the immunity of FPGA to radiation. Finally, the development of FPGA for aerospace applications is prospected.

2. FPGA aerospace applications

Programmable logic devices are more and more widely used in the aerospace and space fields due to their convenient design, easy design modification, and easy function expansion. One is a one-time programming anti-fuse FPGA represented by Actel products, and one is SRAM-based reconfigurable FPGA represented by Xilinx products.

2.1 Classification of aerospace applications FPGA

According to its programmability, there are two main types of FPGAs that currently have successful experience in aerospace applications: One is a one-time programming FPGA that can only be programmed once. The other type is reprogrammable FPGAs that can be programmed multiple times, such as SRAM FPGAs and Flash FPGAs. Such FPGAs generally have In System Programming (ISP) capabilities.

2.1.1 One-time programming FPGA

These products use anti-fuse switching elements, which have the characteristics of small size, small layout area, low radiation resistance and interference resistance, and low characteristic impedance of interconnection lines. No external PROM or EPROM is required, and the configuration data of the circuit will not be lost after power-off , It can work after power on, suitable for various fields such as aerospace, military, industry and so on. Among such products, the representative product that has achieved successful experience in aerospace applications is ACTEL's anti-radiation hardened anti-fuse FPGA. Unlike the layout of traditional FPGA planar distributed logic modules, cables, and switch matrixes, anti-fuse FPGAs use a compact, grid-intensive layout of planar logic module structures. The internal connection elements of the metal-to-metal programmable antifuse between the upper and lower logic module layers are used to realize the connection of the device, which reduces the space occupied by the channel and wiring resources. Before programming, the connecting element is in an open circuit state. During programming, the small area of the anti-fuse structure has a sufficiently high current density, instantaneously generates large thermal power consumption, and melts the insulating medium to form a permanent path.

2.1.2 Reprogrammable FPGA

These products use SRAM or Flash EPROM controlled switching elements, which has the advantage of being programmable repeatedly. The configuration program is stored in a memory outside the FPGA. When the system is powered on, the configuration program is loaded into the FPGA to complete the customization of hardware functions. Among them, SRAM FPGA can also change the configuration during system operation to achieve dynamic reconstruction of system functions. However, the user configuration logic stored in such FPGAs will be lost after power-off, and can only be reloaded from external memory after power-on. FlashEPROM-type FPGA has the advantages of non-volatile and reconfigurable double, but it can not be dynamically configured, and the power consumption is higher than that of SRAM-type FPGA. In this type of FPGA, since the configuration data is stored in the SRAM memory in the FPGA, the programmable logic switch is implemented using a multiplexer, and the internal logic function is implemented using a look-up table based on the SRAM structure. These parts are all single-particle flip-effect sensitive semiconductor structures . Therefore, special attention should be paid to aerospace applications. A representative product with successful experience in aerospace applications is Xilinx's FPGA product based on the SRAM-type Virtex series.

2.2 Current status of FPGA aerospace applications

FPGA has been widely used in the aerospace and space fields at home and abroad, especially commercial satellites. According to statistics, FPGAs are used in a total of 60 projects for deep space exploration, scientific and commercial satellites at home and abroad, and FPGAs are also used in many military satellite projects.

2.2.1 Aerospace applications of Acte FPGA

Actel's radiation-tolerant and radiation-tolerant FPGAs have been used in NASA and ESA's Mars exploration missions since they succeeded in the 1997 Mars Pathfinder and subsequent courage and opportunity missions. Actel's radiation-tolerant and anti-radiation devices are used in the control computer of the Mars rover to perform navigation functions for a six-month flight from Earth to Mars. Actel devices are used in the cameras and wireless communication equipment of the Mars Explorer Rover. In ESA's Mars Express orbit satellite, the solid-state recorder uses more than 20 Actel FPGA devices. Actel's FPGA devices have been used in German Aerospace (DLR) dual-spectrum infrared detection (BIRD) satellites. BIRD is the world’s first satellite that uses infrared sensor technology to detect and study high-temperature events on the earth, such as forest fires, volcanic activity, oil wells, and coal seam combustion. More than 20 high-reliability FPGAs use dry satellite payload data processing, memory management, interface and control, co-processing, and sensor control of infrared cameras.

2.2.2 Aerospace applications of Xilinx FPGA

Compared with ACTEL, Xilinx's products for the aerospace and space fields were developed late. However, the transition of its powerful, high-performance, reconfigurable civilian plastic packaging products to aerospace-grade products, comprehensively improving the ability to resist space radiation, gradually It has become a commonly used FPGA product in the design of space electronic products, and will be more and more widely used.

Xilinx's Virtex radiation-tolerant FPGA was used in the Australian military-civilian mixed communication satellite Optus CL launched in 2003. In the UHF payload of the satellite, XilinxVirtex FPGA (XQVB300) was used to implement the signal processing algorithm for earth data, and Xilinx provided IP core.

Xilinx's ruggedized FPGA XQR4062XL was used in the high-performance computing payload of the Australian scientific satellite Fedsat (a joint satellite used to study the magnetosphere) launched in 2002. HPC-1 is the first example that uses FPGA to implement the configurable computing technology RCT in the standard operation of a spaceborne computer system. The RHC-II currently under development will use Xilinx FPGAs to implement on-board data processing.

In addition, XQR4O36XL products are used in GRACE (NASA) sensors.

Xilinx FPGA products have been successfully used in both Mars Discovery rover Discovery and Spirit. Two pieces of aerospace FPGA VirtexTM FPGA XQVR100O are used in Mars rover wheel motor control, robot arm control and other instruments, 4 pieces of radiation-resistant 4000 series FPGA XQR4062XL are used to control the key ignition equipment of the Mars lander, to ensure that the lander is in accordance with the regulations The procedure drops and the landing is successful. Europe's first comet orbiter and lander ROSETTA has a total of 45 FPGAs, all of which use ACTEL RT14I00A, which undertakes important functions such as control, data management, and power management, and any FPGA in flight must not be powered off.

The newly released Virtex-5QVFPGA from Xilinx has a very high radiation resistance, TID tolerance is more than 700 kraD, SEU (Sin-gle Event Upset, Latch Up) resistance exceeds 100 MeV·cM2/Mg, mainly It is used for remote sensing processing, image processing, and navigators on artificial satellites and spacecraft. Therefore, based on the FPGA system configuration, it is not necessary to increase redundancy for radiation measures, and the time and cost required for system development can be reduced. Its scale has also reached 130,000 logic units, integrated high-speed transceivers with a maximum speed of 3.125 Gbit/s, and enhanced DSP functions, as the highest level in the industry as an FPGA for the aerospace industry.

3. Radiation effects and their effects

Aerospace and space electronic equipment are affected by different radiation due to their different orbits and usage environments. Generally speaking, the radiation effects that have a greater impact on the FPGA are: total dose effect (TID: Total ionizing Dose), single particle upset (SEU: Single event upset), single particle latch (SEL: Single event latchup), Single particle function interruption (SEFI: Single eventfunc-TIonal interrupt), single particle burnout (SEB: Single eventburnout), single particle transient pulse (SET: Single event tran-radiation effect are produced by different mechanisms, causing the failure mode of FPGA different.

Total dose effect: photons or high-energy ions are ionized in the material of the integrated circuit to generate electron-hole pairs, which eventually form oxide trap charges or interface trap charges at the interface between the oxide layer and the semiconductor material, which degrades the performance of the device or even fails.

Single particle flip: Heavy particles with a certain energy collide with the PN junction of a storage device or logic circuit. The charge formed around the trajectory of the heavy particles is collected by the sensitive electrode and becomes a transient current. If the current exceeds a certain value, the logic circuit will be triggered. , Forming a logical state inversion. The single-particle flip sensitive area refers to the area in the FPGA that is easily affected by the single-particle effect, including the FPGA's configuration memory, DCM, CLB, and block storage area.

Single-particle latch: The PNPN structure of the CMOS device becomes a thyristor structure. The incident of protons or heavy particles can trigger the PNPN junction to turn on, enter the high-current regeneration state, and generate single-particle latch-up. Only by reducing the power supply voltage can the latch state be exited.

Single particle function interruption: When the proton or heavy particles are incident, the control logic of the device will malfunction, and then the normal control function will be interrupted. The sensitive parts of the single particle function interrupt in FPGA are configuration memory, power-on reset circuit, SelectMAP interface and JATAG interface.

Single particle burnout: The transient current generated by the incident particles causes the sensitive parasitic bipolar junction transistor to turn on. The regenerative feedback mechanism of the bipolar junction transistor causes the collecting junction current to increase continuously until a secondary breakdown occurs, causing a permanent short circuit between the drain and the source, which burns the circuit. The probability of single particle burnout of FPGA is relatively small.

Single particle transient pulse: The transient current pulse generated by the incident of charged particles affects the input of the logic circuit of the next stage, causing the output of the logic circuit to be disordered. Single particle transient pulses may cause short-term errors in the logic circuits inside the FPGA. Single-particle transient pulses have a greater impact on FPGAs with <0.25 μM processes.

Displacement damage: Single-particle displacement damage is the permanent damage caused by the displacement of lattice atoms, the formation of defect groups, and the single particle incident.

Some of the above radiation effects on FPGA are permanent, such as total dose effect, single particle burnout, displacement damage; some can be recovered, such as single particle flipping, single particle function interruption, single particle transient pulse. In the above single particle effects, SEL, SEB and SEGR may cause permanent damage to the device. Therefore, the general on-board system will use anti-SEL devices. Although SEU and SET are instantaneous effects, their incidence is much higher than the above three, but they should be paid more attention to. Next, based on the analysis of the above radiation effects, the reliability design method to improve the radiation resistance of FPGA is studied.

As SRAM-type FPGAs increase in process level, increase in scale, and decrease in device core voltage, the total anti-dose effect performance continues to improve, but they are more susceptible to SEU and SET.

In response to the problem of single particle effects, the report submitted by the MAPLD, NSREC, and RADECS conferences believes that the Virtex-II series has a total dose radiation resistance of 200 krad, and the SEL resistance is LET 160 MeV·cm /mg. , Need to consider SEU, SET, SEFL and other single particle effects

4. Aerospace applications

The reliability design of FPGA is used in aerospace and space electronic equipment. FPGA is mainly used to replace standard logic, and is also used in SOC technology to provide embedded microprocessors, memory, controllers, communication interfaces, etc. Among them, reliability is the main requirement of FPGA design.

According to the different functions and their importance, the design of space electronic systems is divided into two categories: critical and non-critical. Spacecraft control is a critical category, and scientific instruments are a non-critical category. The general requirements of the spacecraft control system for FPGA: high reliability, radiation hardening and fail-safe. The design requirements of FPGAs for scientific instruments are generally high-performance, radiation-resistant, and fail-safe. Its reliability is determined by performance requirements. The requirements for FPGAs also vary from system to system, such as measurement resolution, bandwidth, high-speed storage, and fault tolerance. Ability etc.

The reliability design of the aerospace FPGA is mainly realized by the hardware design and software design of the device itself.

4.1 FPGA

Hardware reliability design of FPGA The hardware reliability design of FPGA is mainly aimed at the influence of space radiation effect, and the problem of single particle effect protection is solved completely with the help of manufacturing process and design technology. The design is generally carried out from the following aspects: FPGA overall design reinforcement, internal design self-test module for indirect detection of radiation effects, and introduction of external high reliability monitoring module.

The overall reinforcement design refers to the use of a certain thickness of material on the outside of the electronic device for overall radiation shielding to reduce the radiation effect on the device. The commonly used materials are aluminum, tantalum, and lipid compounds. This method is widely used in aerospace electronic components and is also relatively mature. For example, Honeywell, a major supplier of US military microelectronic products, has a wide range of ruggedized ASIC technology. Aeroflex adopts the "design reinforcement, commercial IC process line tape" method to provide advanced performance reinforced ASIC products, and has the ability to develop digital and analog mixed reinforced ASICs. This kind of technology circuit that uses commercial line tape to produce military and reinforced microelectronic products is not only helpful to get rid of the constraints of process reinforcement on device development, but also help meet the user's demand for advanced reinforced devices, reduce costs, and shorten delivery time.

Atmel provides users with high-performance, small-size, low-power devices of various types of process resources, including high-speed, low power consumption, aerospace-resistant digital-analog hybrid CMOS process for aerospace, and CMOS process with embedded EEPROM. There are many domestic units engaged in the development of military microelectronic devices, including state-owned scientific research units and non-state-owned IC development companies. However, there are not many units that can complete the development of radiation-resistant ICs. The domestically developed reinforced ASIC products have been successfully applied in satellites.

The use of bulk epitaxial layers can also prevent SEI. For example, Xilinx's virtex-II radiation-tolerant products are based on military-grade devices and are further designed with epitaxial substrates. The ability to resist total dose ionization effects is evaluated in batches according to MIL-STD-883 Method 1019. The purpose of the self-test module is to predict the normal operation of the entire FPGA through the normal operation of some modules. The self-checking module is realized by a simple logic circuit distributed near the important wiring area of the FPGA. It can also provide output directly from the results of the multi-mode redundant module voting results or the remaining detection method and parity check method.

4.2 FPGA software reliability design for aerospace applications

The software reliability design of FPGA refers to the application software program configuration to shield the malfunction caused by the radiation effect. Among them, the redundant design method is recognized as a more reliable method to deal with the radiation effect. Commonly used redundancy designs include three-module redundancy method (TMR, Triplemoduleredundancy) and partial three-module redundancy method (PTMR, ParTIaltriple module redundancy). Although TMR can improve the reliability of the system, it will also reduce the speed of the module and increase the resource and power consumption. Considering other design indicators comprehensively, the partial three-mode redundancy method can be used for key parts according to the actual situation.

Although the redundant structure can guarantee the reliability of the system, it cannot find and correct errors in time, or introduce too much combinational logic to find errors. When applied to FPGA, it increases the possibility of fault-tolerant circuits themselves. In addition, the unattended operating characteristics of spaceborne systems make system reconstruction and fault recovery very difficult.

Readback verification and reconfiguration (or partial reconfiguration) of the configuration memory is an effective method to resist the radiation effect. The reload of some configurations can repair the impact of the SEU effect, and the frequency should be the worst case. 10 times the incidence of SEU effect. In the reload logic design, the reloading implementation method and the loaded content need to be carefully designed. Not all content can be reloaded, and not all content needs to be reconfigured.

In system design, high-reliability anti-fuse FPGA is used to read Xilinx FPGA configuration data from non-volatile large-capacity memory to configure it. During operation, the configuration memory that is most susceptible to radiation effects is read in columns, and then compared with the standard data, and the columns with errors are partially reconfigured.

FPGA programmable IO is also susceptible to SEU and SEL due to radiation particles. Designing a three-mode redundancy design method for input and output pins is a very effective method, but this method will require 3 times the I/O resources. If SET acts on the clock circuit or other data and control lines, it is easy to produce short pulse jitter, which may cause false triggering of the circuit or data latch error. Synchronous reset can be used in the design to design the internal reset circuit and control line enable The signal line and logic data cooperate with the enable signal as much as possible when latched.

5. FPGA Aerospace Application Development Trend

At present, under the deep micro Yami semiconductor process, the traditional FPGA design technology faces challenges in device yield, power consumption, interconnect delay, signal integrity, and testability design [9]. FPGAs based on traditional technologies are still developing in the direction of high density, high performance, and low power consumption, making FPGAs develop from the first general-purpose semiconductor devices to platformized system-level devices. FPGA design based on asynchronous circuits, 3D integration technology, and the application of new semiconductor structures will be the focus of FPGA technology development.

In terms of aerospace and space applications, the summary and predictive analysis of FPGA space applications by foreign aerospace shows that space applications show the following trends in the selection of FPGAs:

(1) The working voltage of the device changes from 5 V to 3.3 V, 2.5 V or even 1.8 V;

(2) From the use of total dose hardening FPGA to the use of total dose resistant FPGA products;

(3) From the application of SEU sensitive register FPGA to FPGA using built-in register TMR structure;

(4) Develop from an anti-fuse FPGA that only uses one-time programming to a resettable FPGA based on SRAM/EEPROM.

The outstanding problem brought about by this selection trend is: from register sensitive to SEU to FPGA sensitive to SEU; the design complexity of configuration storage FPGA has been equivalent to the complexity of ASIC.

6 Conclusion

This article reviews the use of FPGAs in aerospace applications. The structural characteristics of FPGAs are analyzed, and the failure modes and reliability design methods of FPGAs for aerospace applications are analyzed according to the irradiation conditions of aerospace and space environments. Finally, the development of FPGA and reliability design technology for aerospace applications is prospected.