Joan Vila

Using exact feasibility tests for allocating real-time tasks in multiprocessor systems

The paper introduces improvements in partitioning schemes for multiprocessor real time systems which allow higher processor utilization and enhanced schedulability by using exact feasibility tests to evaluate the schedulability limit of a processor. The paper analyzes how to combine these tests with existing bin packing algorithms for processor allocation and provides new variants which are exhaustively evaluated using two assumptions: variable and fixed number of processors. The problem of evaluating these algorithms is complex, since metrics are usually based on comparing the performance of a given algorithm to an optimal one, but determining optimals is often NP hard on multiprocessors. This problem has been overcome by defining lower or upper bounds on the performance of an optimal algorithm and then defining metrics with respect these bounds. The evaluation has shown that the algorithms exhibit extremely good behavior and they can be considered very close to the maximum achievable utilization. It is also shown that dynamic priority policies produce significantly better results than fixed priority policies when task sets require high processor utilizations.

Building distributed embedded systems with RTLinux-GPL

Linux has revealed in the last few years as an appealing option for developing embedded systems but there are some extra requirements of embedded systems that Linux does not fulfill, such as real-time capabilities, file system size and specific hardware support. Real-time features can be achieved with Real-Time Linux GPL (hereafter RTLinux), which is a small, deterministic, real-time kernel that handles time-critical tasks and runs Linux as its lowest priority execution thread. However RTLinux has also important drawbacks. One of them is that real-time tasks cannot make use of Linux services and, in particular, TCP/IP networking. This paper describes RTL-lwIP, which is a TCP/IP stack for embedded systems based on the lwIP (lightweight) TCP/IP stack that runs on RTLinux and can be used by real-time tasks. RTL-lwIP allows real-time tasks to communicate directly with remote real-time tasks or even with Linux user processes. The importance of introducing TCP/IP on RTLinux is that it enables the possibility of developing real-time distributed embedded systems based on CORBA, thus allowing interoperability with other platforms and Web-integration. This paper describes the porting of the lwIP TCP/IP stack to RTLinux and gives some guidelines in order to implement RTLinux drivers for Ethernet cards using as example the implementation of a RTLinux driver for the 3Com905C-X NIC (Network Interface Card).

A new approach to optimize bandwidth reservation for real-time video transmission with deterministic guarantees

The transmission of video requires new service models for providing quality of service (QoS). Some of these models are based on resource reservation and admission control while others rely on resource provisioning. In both approaches, optimally estimating the resource requirements of a given video is a key issue, since they are always very demanding. This paper introduces a fast and bounded method to optimize the required network resources (bandwidth) to guarantee a maximum deterministic delay for a given workload and network state. The method is based on characterizing the workload of a stored video using a reduced set of points obtained from an off-line analysis of its empirical envelope. The paper also proposes the above workload characterization to perform the admission control test using a WFQ scheduler. The evaluation of this scheme shows that it can achieve an utilization higher than a 50% for a single flow with a 1-s deadline and reach up to an 80% with a 5-s delay. Results are also compared to tests based on EDF schemes, which have been proved to be optimal with one node. It is shown that the proposed scheme is a little less efficient than the optimal EDF scheduler with one node (as expected), but it is practically as efficient (or even better with a moderately high number of nodes) than the best known RC-EDF policies with several nodes. This is an interesting result because it shows that the WFQ schedulers can achieve a similar efficiency to EDF schedulers avoiding the complexity of their admission control tests.The transmission of video requires new service models for providing quality of service (QoS). Some of these models are based on resource reservation and admission control while others rely on resource provisioning. In both approaches, optimally estimating the resource requirements of a given video is a key issue, since they are always very demanding. This paper introduces a fast and bounded method to optimize the required network resources (bandwidth) to guarantee a maximum deterministic delay for a given workload and network state. The method is based on characterizing the workload of a stored video using a reduced set of points obtained from an off-line analysis of its empirical envelope. The paper also proposes the above workload characterization to perform the admission control test using a WFQ scheduler. The evaluation of this scheme shows that it can achieve an utilization higher than a 50% for a single flow with a 1-s deadline and reach up to an 80% with a 5-s delay. Results are also compared to tests based on EDF schemes, which have been proved to be optimal with one node. It is shown that the proposed scheme is a little less efficient than the optimal EDF scheduler with one node (as expected), but it is practically as efficient (or even better with a moderately high number of nodes) than the best known RC-EDF policies with several nodes. This is an interesting result because it shows that the WFQ schedulers can achieve a similar efficiency to EDF schedulers avoiding the complexity of their admission control tests. r 2002 Elsevier Science Ltd. All rights reserved.

The Real-Time Stealer 1

Abstract Slack stealing algorithms are a widely accepted solution for jointly scheduling both hard deadline periodic tasks and soft aperiodic tasks in real-time computations, and they all imply long searches in real-time computations looking for slack gaps. That makes them impractical to implement in software, specially in the dynamic case, since the scheduler overhead would be very important. The advocated solution is to implement them in hardware, but hardware implementations are not straightforward and they often require an algorithm redesign and to face other important design trade-offs to meet time requirements. This paper presents a complete hardware implementation of a dynamic slack stealing algorithm based on a novel algorithm for slack detection that specially fits in a hardware implementation and introduces improvements in response times over previous ones. The algorithm has been implemented in VHDL and built using a FPGA.

A dynamic real-time scheduler for shared memory multiprocessors

This paper presents a dynamic scheduling algorithm for multiprocessor systems which is guarantee-oriented. The algorithm is based on dynamic priority policy where priorities are inversely proportional to the latest start time (LST) of tasks. The paper includes a comparative study of the algorithm with other two well known dynamic algorithms: the earliest deadline first (EDF) and the least laxity first (LLF). The results show that LST is able to schedule some loads where EDF fails and reduces significantly the number of context switches with respect to LLF. The algorithm has been implemented and tested on a tool for multiprocessor analysis and simulation which is also presented in this paper.

A binary-tree architecture for scheduling real-time systems with hard and soft tasks

Complex real-time systems are required to jointly schedule both periodic and aperiodic tasks. To do this, optimal schedulers are dynamic slack stealing algorithms (DSS). However, these software schedulers are impractical due to their large overheads that always result in delays and reduced CPU utilization. One of the proposed solutions to this problem is doing scheduling in hardware. This paper analyzes in depth a hardware design based on binary trees.

Soft aperiodic task scheduling on hard real-time multiprocessor systems

The problem of jointly scheduling both hard deadline periodic tasks and soft aperiodic tasks has been the subject of considerable research in real-time systems. The main goal of such a system is to minimize the response time of soft aperiodic tasks, without jeopardizing the hard deadlines of periodic tasks. Although, several approaches have been developed to schedule critical workloads on multiprocessors and to manage aperiodic tasks on monoprocessors, the proposed problem has not been fully solved for multiprocessors and integrated solutions are still required. This paper addresses the problem of scheduling an aperiodic workload on a multiprocessor system, where every processor has been previously allocated a critical periodic workload. Such pre-allocated workload is the result of executing a partitioning algorithm in order to ensure the timeliness of the critical workload. Two different approaches are used in order to minimize the response time of the aperiodic workload: aperiodic on-line assignment and global aperiodic scheduling. On highly loaded systems, the results show that the global aperiodic scheduling obtains aperiodic average response times between 20% and 50% lower than those obtained by the on-line assignment approach.

RTL-TEP: AN ETHERNET PROTOCOL BASED ON TDMA

Abstract RTLinux-GPL and Ethernet networks can be effectively used for many real-time environments by using some mechanisms or protocols to provide support for real-time and QoS requirements. This paper presents RTL-TEP, a software protocol on top of Ethernet in order to make Ethernet deterministic. The main issues to face implementing TDMA in software are described, as well as the special features of RTL-TEP. Some of the questions that are discussed are clock synchronization, dynamic behaviour and fault-tolerant mechanisms.

Using RT-Linux for Developing Real-time Embedded Systems(*)

ABSTRACT Linux has been effectively used as a reliable operating system in workstations and servers. However, the interest for using it in embedded systems has grown recently, due to the easiness for customizing it and the availability of the source code. This paper describes the experience of customizing RT-Linux for its use in an embedded real-time control. The main issues involved in this task are outlined and the adopted solution is described. These issues include the system configuration, the minimization effort and how to achieve real-time features through the use of RT-Linux. An application for obtaining task executions chronograms is presented as an example of application design under RT-Linux. Finally some measurements about context switching times are presented.

Experiences using Ada in a real-time and distributed laboratory

This paper reviews our experience in the laboratory of two courses in the Control Engineering Curriculum at the Politechnical University of Valencia: Real-time Systems and Distributed Control Systems. These courses, in a Control Engineering curriculum, require the development of industrial prototypes to validate different aspects of the system development: control algorithms, real-time software design, external devices integration, distributed control, etc. To meet all these goals a robot guidance experience has been selected. In this paper, we describe the experience in the design and implementation of a virtual and real environment developed in Ada to complete the student laboratory activities. The system is complex enough to allow the students to analyze, design and implement a complete prototype