Rafael Ors

Adding instruction cache effect to schedulability analysis of preemptive real-time systems

Cache memories are commonly avoided in real time systems because of their unpredictable behavior. Recently, some research has been done to obtain tighter bounds on the worst case execution time (WCET) of cached programs. These techniques usually assume a non preemptive underlying system. However, some techniques can be applied to allow the use of caches in preemptive systems. The paper describes how to incorporate the effect of instruction cache to the Response Time schedulability Analysis (RTA). RTA is an efficient analysis for preemptive fixed priority schedulers. We also compare through simulations the results of such an approach to both cache partitioning (increase of the cache predictability by assigning private cache partitions to tasks) and CRMA (Cached RMA: cache effect is incorporated in the utilization based rate monotonic schedulability analysis). The results show that the cached version of RTA (CRTA) clearly outperforms CRMA, however the partitioning scheme may be better depending on the system configuration. The obtained results bound the applicability domain for each method for a variety of hardware and workload configurations. The results can be used as design guidelines.

A Wireless Sensor Network approach for distributed in-line chemical analysis of water.

In this work we propose the implementation of a distributed system based on a Wireless Sensor Network for the control of a chemical analysis system for fresh water. This implementation is presented by describing the nodes that form the distributed system, the communication system by wireless networks, control strategies, and so on. Nitrate, ammonium, and chloride are measured in-line using appropriate ion selective electrodes (ISEs), the results obtained being compared with those provided by the corresponding reference methods. Recovery analyses with ISEs and standard methods, study of interferences, and evaluation of major sensor features have also been carried out. The communication among the nodes that form the distributed system is implemented by means of the utilization of proprietary wireless networks, and secondary data transmission services (GSM or GPRS) provided by a mobile telephone operator. The information is processed, integrated and stored in a control center. These data can be retrieved--through the Internet--so as to know the real-time system status and its evolution.

Distributed expert systems as a new tool in analytical chemistry

Abstract This article emphasizes the growing importance of artificial intelligence in the automation of chemical analysis processes. Nevertheless, the authors predict that the use of centralized knowledge-based systems – though not still consolidated – will probably be replaced by distributed expert systems with their multiple advantages (discussed in the article). Distributed expert systems, which are based on the utilization of highly intelligent analytical instrumentation connected by means of communication networks, permit the implementation of highly flexible control systems that are capable of adapting themselves to different situations. At the same time, it is possible to take advantage of the power offered by the new communication technologies in the field of analytical chemistry. Distributed expert systems are thoroughly described and illustrated with a detailed example of an application to the monitoring and control of an alcoholic fermentation process by means of on-line/in-line chemical analysis; this can be extended to any other process whose chemical parameters undergo variations in the course of time.

Using harmonic task-sets to increase the schedulable utilization of cache-based preemptive real-time systems

Cache memories have been traditionally precluded in real-time systems because of their unpredictable behavior. The needs of better performance have motivated the development of tools to obtain tighter bounds on the worst-case execution time (WCET) of cached programs. However, they do not allow preemption, because from the point of view of program analysis, the number of preemptions is unknown. To face this problem, the cache-related preemption cost can be considered in the schedulability analysis, or annulled by the use of private cache partitions. Regarding the first approach, in a previous paper we described how to incorporate the effect of the cache to the Response Time schedulability Analysis (RTA). RTA is an efficient analysis for preemptive fixed-priority schedulers. In this paper, we have improved such a technique by taking into account the harmonic relations amongst task frequencies. The resulting analysis is applied to sample task-sets to illustrate the potential gains in terms of schedulable utilization, particularly for highly demanding task-sets.

An expert system as a tool for the specification and intelligent control of a flow injection analysis system

Abstract An expert system was designed with the intention that it should fully control a fermentation process, whose on-line monitoring is based on the determination of total acidity, reducing sugars, ethanol and pH all along the process. The expert system is then in charge of deciding when and which measurements are to be carried out according to the previously obtained results and a pre-determined set of rules. Additionally, the expert system is able to take decisions concerning possible malfunctions.

Distributed industrial control systems: a fault-tolerant architecture

Abstract Nowadays, distributed architectures are the base of many manufacturing systems. Some aspects like fault-tolerance, system validation and design process are very important in the development of these systems. In this paper we study the dependability of three different architectures of a distributed system, and we show the development of both physical and logical fault injectors and the implementation of local performance monitors. We also study the impact of checkpointing mechanisms on the system performance in a control system based on a CAN network. Finally we propose a distributed system design methodology based on codesign.

Rule nets: application to the advanced automation of a flow-injection analysis system

Abstract An expert system based on rule nets has been developed. This new tool—the rule net—consists of a set of conditional-type propositions which are mutually bound. In this work, a set of rule nets has been implemented by means of a self-developed pack called AUTOMAT. Time has been incorporated into the reasoning of the expert system in such a way that it is able to carry out the on-line monitoring of an alcoholic fermentation process.

A New Energy-Efficient, Scalable and Robust Architecture for Wireless Sensor Networks

The major problem for most of wireless sensor networks applications is the energy consumption. In this paper we propose a new architecture called EDETA (Energy-efficient aDaptive hiErarchical and robusT Architecture) optimized to save nodes power. This architecture is scalable and suitable for heterogeneous and homogeneous wireless sensor networks, supports single or multiple sinks, and also provides features increasely-demanded in the current applications such as fault tolerance and bounded-time response communications. The proposed protocol is able to autoconfigure, and it is based on two-levels hierarchical architecture. The lower level is based on cluster organization, while the upper one is formed as a dynamic tree of clusters heads to send the data to the sink. Our experiments, based on real energy measures, show that EDETA can reduce the energy consumption in typical applications in a factor of 8x compared with most popular power-aware protocols. Furthermore, EDETA multiplies the lifetime of the sensor network while providing, in addition, extra-features such as fault-tolerant mechanisms and bounded time.

SIVEH: Numerical Computing Simulation of Wireless Energy-Harvesting Sensor Nodes

The paper presents a numerical energy harvesting model for sensor nodes, SIVEH (Simulator I–V for EH), based on I–V hardware tracking. I–V tracking is demonstrated to be more accurate than traditional energy modeling techniques when some of the components present different power dissipation at either different operating voltages or drawn currents. SIVEH numerical computing allows fast simulation of long periods of time—days, weeks, months or years—using real solar radiation curves. Moreover, SIVEH modeling has been enhanced with sleep time rate dynamic adjustment, while seeking energy-neutral operation. This paper presents the model description, a functional verification and a critical comparison with the classic energy approach.

Monitoring of a wort fermentation process by means of a distributed expert system

Abstract A distributed expert system has been proposed for the monitoring of a wort fermentation process through the flow injection determination of total acidity, reducing sugars, ethanol and pH. Its configuration is mainly based on the use of distributed nodes connected by means of a network and is capable of adapting itself to different situations. Satisfactory results have been obtained when it has been applied in a brewery plant, important advantages being shown over previously implemented centralized knowledge-based systems.