To avoid data cache trashing between heap-allocated data and other data areas, a distinct object cache has been proposed for embedded real-time Java processors. This object cache uses high associativity in order to st...
详细信息
To avoid data cache trashing between heap-allocated data and other data areas, a distinct object cache has been proposed for embedded real-time Java processors. This object cache uses high associativity in order to statically track different object pointers for worst-case execution-time analysis. However, before implementing such an object cache, an empirical analysis of different organization forms is needed. We use a cross-profiling technique based on aspect-oriented programming in order to evaluate different object cache organizations with standard Java benchmarks. From the evaluation we conclude that field access exhibits some temporal locality, but almost no spatial locality. Therefore, filling long cache lines on a miss just introduces a high miss penalty without increasing the hit rate enough to make up for the increased miss penalty. For an object cache, it is more efficient to fill individual words within the cache line on a miss.
The paradigm shift from traditional on-premise software to a service based model has gained significant momentum in the past decade. One such concept, Software as a service (SaaS), delivers the functionality of tradit...
详细信息
The paradigm shift from traditional on-premise software to a service based model has gained significant momentum in the past decade. One such concept, Software as a service (SaaS), delivers the functionality of traditional on-premise software as a service over the web. While a defect or a malfunction in a traditional on-premise application may affect a single user, the affected user base in a SaaS application may span the entire group of customers serviced by the provider. The physical disconnect between end users and the SaaS applications puts onus on service providers to deliver highly dependable systems that are available and reliable at all times. In this paper, we explore the general challenges faced in delivering and analyzing highly dependable service based systems. We quantify the challenges of dependability assessment utilizing a commercial case study. Furthermore, we explore one facet of dependability assessment related to log entries not necessarily related to dependability. We provide a novel approach to log filtering and show that the removal of benign log entries leads to more realistic system dependability analysis. We also show the need to merge multiple types of SaaS logs to support effective analysis.
Type-safe high-level languages such as Java have not yet found their way into the domain of deeply embedded systems, even though numerous attempts have been made to make these languages cost attractive. One major chal...
详细信息
Type-safe high-level languages such as Java have not yet found their way into the domain of deeply embedded systems, even though numerous attempts have been made to make these languages cost attractive. One major challenge that remains is the huge existing code base in many industries. Completely reengineering this code base is not viable for cost and time reasons. We present an approach that allows to isolatedly combine legacy software components and safe software components in an embedded system using the two most common communication idioms found in this domain. Our approach allows the developer to freely choose between hardware- and software-based isolation mechanisms. We demonstrate the feasibility of our approach by porting a non-trivial part of a real-world, hard real-time embedded avionics application. Our results show that the cost of this mixed-mode operation is on the same scale as the pure operation.
Multicore processors can deliver higher performance than single-core processors by exploiting thread level parallelism (TLP): applications are split into independent threads, each of which is mapped into a different c...
详细信息
Multicore processors can deliver higher performance than single-core processors by exploiting thread level parallelism (TLP): applications are split into independent threads, each of which is mapped into a different core, reducing the execution time and potentially its worst-case execution time (WCET). Unfortunately, inter-thread interferences generated by simultaneous accesses to shared resources from different threads may completely destroy the performance benefits brought by TLP. This paper proposes a software/hardware cache partitioning approach that reduces the inter-thread memory interferences generated in hard real-time software-pipelined parallel applications. Our results show that our approach effectively reduces memory interferences, while still guaranteeing a predictable timing behaviour, achieving a WCET estimation reduction of 28% for a software pipelined version of the LU decomposition application with respect to the single-threaded version.
One cannot image today's life without mechatronic systems, which have to be developed in a joint effort by teams of mechanical engineers, electrical engineers, control engineers and software engineers. Often these...
详细信息
One cannot image today's life without mechatronic systems, which have to be developed in a joint effort by teams of mechanical engineers, electrical engineers, control engineers and software engineers. Often these systems are applied in safety critical environments like in cars or aircrafts. This requires systems that function correctly and do not cause hazardous situations. However, random errors due to wear or external influences cannot be completely excluded. Consequently, we have to perform a hazard analysis for the system. Further, the union of four disciplines in one system requires the development and analysis of the system as a whole. We present a component-based hazard analysis that considers the entire mechatronic system including hardware, i.e. mechanical and electrical components, and software components. Our approach considers the physical properties of different types of flow in mechatronic systems. We have identified reusable patterns for the failure behavior which can be generated automatically. This reduces the effort for the developer. As cycles, e.g. control cycles, are an internal part of every mechatronic system our approach is able to handle cycles. The presented approach has been applied to a real-life case study.
Smart phones are starting to find use in mission critical applications, such as search-and-rescue operations, wherein the mission capabilities are realized by deploying a collaborating set of services across a group o...
详细信息
Smart phones are starting to find use in mission critical applications, such as search-and-rescue operations, wherein the mission capabilities are realized by deploying a collaborating set of services across a group of smart phones involved in the mission. Since these missions are deployed in environments where replenishing resources, such as smart phone batteries, is hard, it is necessary to maximize the lifespan of the mission while also maintaining its real-time quality of service (QoS) requirements. To address these requirements, this paper presents a deployment framework called Smart Deploy, which integrates bin packing heuristics with evolutionary algorithms to produce near-optimal deployment solutions that are computationally inexpensive to compute for maximizing the lifespan of smart phone-based mission critical applications. The paper evaluates the merits of deployments produced by Smart Deploy for a search-and-rescue mission comprising a heterogeneous mix of smart phones by integrating a worst-fit bin packing heuristic with particle swarm optimization and genetic algorithm. Results of our experiments indicate that the missions deployed using Smart Deploy have a lifespan that is 20% to 162% greater than those deployed using just the bin packing heuristic or evolutionary algorithms. Although Smart Deploy is slightly slower than the other algorithms, the slower speed is acceptable for offline computations of deployment.
Wireless networking is moving toward the adoption of IP protocols and away from the multitude of special-purpose tactical radios traditionally in the hands of emergency personnel, military personnel, and law enforceme...
详细信息
Wireless networking is moving toward the adoption of IP protocols and away from the multitude of special-purpose tactical radios traditionally in the hands of emergency personnel, military personnel, and law enforcement. The adoption of standards, such as IP multicast, has facilitated this. IP multicast also enables recovering some of the advantages of the broadcast medium when using IP in tactical environments. However, the traditional Quality of service (QoS) approaches for IP multicast fall short of satisfying the stringent QoS requirements in tactical environments, which typically have single-hop, line-of-sight connections. The reasons for this are (1) QoS in IP networks, frequently based on Differentiated services, relies on routers to enforce the priorities which typically don't exist in tactical networks, and (2) QoS for tactical users needs to be enforced at the information level, not the packet level where the loss or delay of a single packet can invalidate an entire object of information. We present strategies for QoS management for IP multicast in tactical environments that provides information and user-level QoS and addresses the specific challenges of tactical radios (such as the lack of reliable capacity information). We present our solutions in the context of a tactical information broker that provides beyond line-of-sight information management in a theater of operations.
Recent embedded systems like mobile phones and digital TV appliances have been increasing their functionalities. These systems become very complex, but new products should be developed with short time-to-market. There...
详细信息
Recent embedded systems like mobile phones and digital TV appliances have been increasing their functionalities. These systems become very complex, but new products should be developed with short time-to-market. Therefore, future embedded systems that will be more complex require a new software infrastructure that industries can develop new products faster by reusing existing software as much as possible. In our project, we are developing SPUMONE which is able to compose multiple functionalities with a minimum cost. SPUMONE enables multiple OSes to coexist on a single system. Therefore, most of existing application programs require no modification. For satisfying the requirements of embedded systems, one of the most important goal of SPUMONE is to satisfy real-time constraints of real-time OSes. When real-time OS and general purpose OS like Linux run at the same time, the real-time OS can ensure the real-time responsiveness. It is also important to ensure security in future embedded systems that is always connected to the Internet. Increasing the complexity and functionality in embedded systems makes the OS kernel more vulnerable from malicious programs. This raises the need of the integrity management system for the general purpose OS kernel. The integrity management system that observes the behavior of kernel activities from the outside of the OS kernel is promising approach to increase the security of the general purpose OS kernel. However, traditional integrity management systems require extra resources and do not take into account the monitoring on a multi-core processor. The monitoring service which is an integrity management system running on SPUMONE as a new OS personality checks the integrity of data structures in the general purpose OS kernel. It also takes into account running on a multi-core processor. Therefore, the monitoring service and the general purpose OS can run on different cores to reduce the overhead of the integrity management.
The need of task-adapted and complete information for the management of resources is a well known issue in Grid computing. Globus Toolkit 4 (GT4) includes the Monitoring and Discovery System component (MDS4) to carry ...
详细信息
The need of task-adapted and complete information for the management of resources is a well known issue in Grid computing. Globus Toolkit 4 (GT4) includes the Monitoring and Discovery System component (MDS4) to carry out resource management. The Common Information Model (CIM) provides a standard conceptual view of the managed environment. This work improves the MDS4 functionality through the use of CIM, with the aim of providing a unified, standard representation of the Grid resources. Since a practical CIM model may contain a large volume of information, a new Index service that represents the CIM information through Java instances is presented. In addition, a solution that keeps data in persistent storage has also been implemented. The evaluation of the proposed solutions achieves encouraging results, with an important reduction in memory consumption, a good scalability when the number of instances increases, and with a reasonable response time.
SOA middleware has emerged as a powerful and popular distributedcomputing paradigm because of its high-level abstractions for composing systems and encapsulating platform-level details and complexities. Control of so...
详细信息
SOA middleware has emerged as a powerful and popular distributedcomputing paradigm because of its high-level abstractions for composing systems and encapsulating platform-level details and complexities. Control of some details encapsulated by SOA middleware is necessary, however, to provide managed QoS for SOA systems that require predictable performance and behavior. This paper presents a policy-driven approach for managing QoS in SOA systems called QoS enabled dissemination (QED). QED includes services for: (1) specifying and enforcing the QoS preferences of individual clients;(2) mediating and aggregating QoS management on behalf of competing users;and (3) shaping information exchange to improve real-time performance. We describe QED's QoS services and mechanisms in the context of managing QoS for a set of Publish-Subscribe-Query information management services. These services provide a representative case study in which CPU and network bottlenecks can occur, client QoS preferences can conflict, and system-level QoS requirements are based on higher level, aggregate end-to-end goals. We also discuss the design of several key QoS services and describe how QED's policy-driven approach bridges users to the underlying middleware and enables QoS control based on rich and meaningful context descriptions, including users, data types, client preferences, and information characteristics. In addition, we present experimental results that quantify the improved control, differentiation, and client-level QoS enabled by QED. Copyright (C) 2011 John Wiley & Sons, Ltd.
暂无评论