Sebastian Korf

Automatic HDL-Based Generation of Homogeneous Hard Macros for FPGAs

The regularity of resources found in FPGAs is a unique feature, which can be utilized in a number of applications, e.g., in timing critical applications or applications with a demand for homogeneous routing. Current synthesis tools do not support an automatic generation of homogeneous FPGA designs, such that a time-consuming hand-crafted design is required. We present a tool flow, which automatically generates homogeneous hard macros for Xilinx FPGAs starting from a high-level description, such as VHDL. Key functionalities of the tool flow are a homogeneous placer and a suitable routing algorithm, which aim at maintaining the homogeneity of the resulting hard macro. The place and route tools use a resource library that is automatically generated for the target FPGA family by extracting relevant information from the vendor tools. The tool chain is demonstrated for the design of hard macros for a time-to-digital converter and a tiled partially reconfigurable region. The resulting designs are evaluated with respect to resource requirements and timing constraints.

A scalable platform for run-time reconfigurable satellite payload processing

Reconfigurable hardware is gaining a steadily growing interest in the domain of space applications. The ability to reconfigure the information processing infrastructure at runtime together with the high computational power of todays FPGA architectures at relatively low power makes these devices interesting candidates for data processing in space applications. Partial dynamic reconfiguration of FPGAs enables maximum flexibility and can be utilized for performance increase, for improving energy efficiency, and for enhanced fault tolerance. To be able to prove the effectiveness of these novel approaches for satellite payload processing, a highly scalable prototyping environment has been developed, combining dynamically reconfigurable FPGAs with the required interfaces such as SpaceWire, MIL-STD-1553B, and SpaceFibre. Up to 30 SpaceWire interfaces, 5 copper-based SpaceFibre interfaces, and 270 GPIOs can be realized and combined with one to five dynamically reconfigurable Xilinx FPGAs and up to 20 GByte of working memory. The implemented approach for dynamic reconfiguration enables partial reconfiguration at 400 MByte/s. Blind and readback scrubbing is supported and the scrub rate can be adapted individually for different parts of the design.

On-line testing of permanent radiation effects in reconfigurable systems

Partially reconfigurable systems are more and more employed in many application fields, including aerospace. SRAM-based FPGAs represent an extremely interesting hardware platform for this kind of systems, because they offer flexibility as well as processing power. In this paper we report about the ongoing development of a software flow for the generation of hard macros for on-line testing and diagnosing of permanent faults due to radiation in SRAM-FPGAs used in space missions. Once faults have been detected and diagnosed the flow allows to generate fine-grained patch hard macros that can be used to mask out the discovered faulty resources, allowing partially faulty regions of the FPGA to be available for further use.

Methods of Improving the Dependability of Self-optimizing Systems

Various methods have been developed in the Collaborative Research Center 614 which can be used to improve the dependability of self-optimizing systems. These methods are presented in this chapter. They are sorted into two categories with regard to the development process of self-optimizing systems. On one hand, there are methods which can be applied during the Conceptual Design Phase. On the other hand, there are methods that are applicable during Design and Development.

Exploiting dynamic partial reconfiguration for on-line on-demand testing of permanent faults in reconfigurable systems

Reconfigurable systems are increasingly employed in many application fields, including aerospace. The long term exposure to radiation of space electronics can cause permanent faults, that may lead to the failure of the mission. In this paper we present a novel technique for on-line on-demand testing of permanent faults in the routing structure of SRAM-based FPGAs, that are employed in reconfigurable systems. The basic idea is to place testing circuits on the resources of the FPGA which are unused at the moment to test them before using those resources when a functional module of the reconfigurable system has to be placed. The proposed technique has been implemented and the achieved fault coverage has been assessed on a real-world reconfigurable system. This experiment demonstrated that all the faults in the routing resources under test can be detected.

An inter-processor communication interface for data-flow centric heterogeneous embedded multiprocessor systems

Modern high-performance embedded systems are characterized by heterogeneity of the employed processing elements: general-purpose processors working together with embedded processors and with dedicated hardware accelerators or high-speed I/O interfaces. Communication among these processors and interfaces is one of the crucial aspects of the development of such systems. In this paper we present a communication interface for heterogeneous embedded multiprocessor systems. The interface is intended to be used to synchronize the data-flow between hardware accelerators and high-speed I/O interfaces. The interface has been designed with the aim of finding a trade-off between performance, flexibility and usability. In order to assess its performance and usability, the interface has been implemented and applied to a heterogeneous multiprocessor architecture used for payload processing in satellite systems.

The Paradigm of Self-optimization

Machines are ubiquitous. They produce, they transport. Machines facilitate and assist with work. The increasing fusion of mechanical engineering with information technology has brought about considerable benefits. This situation is expressed by the term mechatronics, which means the close interaction of mechanics, electrics/electronics, control engineering and software engineering to improve the behavior of a technical system. The integration of cognitive functions into mechatronic systems enables systems to have inherent partial intelligence. The behavior of these future systems is formed by the communication and cooperation of the intelligent system elements. From an information processing point of view, we consider these distributed systems to be multi-agent-systems. These capabilities open up fascinating prospects regarding the design of future technical systems. The term self-optimization characterizes this perspective: the endogenous adaptation of the system’s objectives due to changing operational conditions. This resuls in an autonomous adjustment of system parameters or system structure and consequently of the system’s behavior. In this chapter self-optimizing systems are described in detail. The long term aim of the Collaborative Research Centre 614 ”Self-Optimizing Concepts and Structures in Mechanical Engineering” is to open up the active paradigm of self-optimization for mechanical engineering and to enable others to develop these systems. For this, developers have to face a number of challenges, e.g. the multidisciplinarity and the complexity of the system. This book povides a design methodology that helps to master these challenges and to enable third parties to develop self-optimizing systems by themselves.

Development of Self-optimizing Systems

In the development of self-optimizing systems, developers have to face different challenges, such as the multidisciplinarity of mechatronics, cross-linking between the subsystems, the lack of current knowledge in the fields of advanced mathematics and artificial intelligence, and increased dependability requirements. To support the developer in this challenging task adequately, the Collaborative Research Center 614 has developed a design methodology consisting of a reference process, methods and tools. The reference process is divided into two phases: the ”Domain-Spanning Conceptual Design” and the ”Domain-Specific Design and Development”. In the first phase, the domain-spanning model-based aproach CONSENS (CONceptual design Specification technique for the ENgineering of complex Systems) is used to create a common understanding basis between the different domains involved. Based on the Principle Solution developed in this phase, the ”Domain-Specific Design and Development” is planned and implemented. To ensure the development of dependable self-optimizing systems, domain-spanning and domain-specific dependability engineering methods can be used. The developer encounters a significant number of these methods, that have to be integrated into the process to obtain an individualized development process for the specific development task.

Methods for the Design and Development

After the domain-spanning conceptual design, engineers from different domains work in parallel and apply their domain-specific methods and modeling languages to design the system. Vital for the successful design, are system optimization methods and the design of the reconfiguration behavior. The former methods enable the parametric adaption of the system’s behavior, e.g. an adaption of controller parameters, according to a current selection of the system’s objectives. The latter realizes structural adaption of the system’s behavior, e.g. the exchange of software or hardware parts. Altogether, this leads to a complex system behavior that is hard to overview. In addition, self-optimizing systems are used in safety-critical environments. Consequently, the system’s safety-critical behavior has to undergo a rigorous verification and testing process. Existing design methods do not address all of these challenges together. Indeed, a combination of established design methods for traditional technical systems with novel methods that focus on these challenges is necessary. In this chapter, we will focus on such new methods. We will introduce new system optimization and design methods to develop reconfigurations of the software and the microelectronics. In order to ensure the correctness of safety-critical functionality, we propose new testing methods and formal methods to ensure safety-properties of the software. We show how to apply virtual prototyping to deal with the complexity of self-optimizing systems and perform an early analysis of the overall system. As each domain applies its own modeling languages, the result of these methods are several overlapping models. In order to keep these domain-specific models consistent among all domains, we will introduce a new semi-automatic model synchronization technique. Each of these design methods are integrated with the reference process for the development of self-optimizing systems.