Calculating an upper bound on the finishing time of a group of threads executing on a GPU: a preliminary case study

Graphics processor units (GPUs) today can be used for computations that go beyond graphics and such use can attain a performance that is orders of magnitude greater than a normal processor. The software executing on a graphics processor is composed of a set of (often thousands of) threads which operate on different parts of the data and thereby jointly compute a result which is delivered to another thread executing on the main processor. Hence the response time of a thread executing on the main processor is dependent on the finishing time of the execution of threads executing on the GPU. Therefore, we present a simple method for calculating an upper bound on the finishing time of threads executing on a GPU, in particular NVIDIA Fermi. Developing such a method is nontrivial because threads executing on a GPU share hardware resources at very fine granularity.

LLF schedulability analysis on multiprocessor platforms

LLF (Least Laxity First) scheduling, which assigns a higher priority to a task with smaller laxity, has been known as an optimal preemptive scheduling algorithm on a single processor platform. However, its characteristics upon multiprocessor platforms have been little studied until now. Orthogonally, it has remained open how to efficiently schedule general task systems, including constrained deadline task systems, upon multiprocessors. Recent studies have introduced zero laxity (ZL) policy, which assigns a higher priority to a task with zero laxity, as a promising scheduling approach for such systems (e.g., EDZL). Towards understanding the importance of laxity in multiprocessor scheduling, this paper investigates the characteristics of ZL policy and presents the first ZL schedulability test for any work-conserving scheduling algorithm that employs this policy. It then investigates the characteristics of LLF scheduling, which also employs the ZL policy, and derives the first LLF-specific schedulability test on multiprocessors. It is shown that the proposed LLF test dominates the ZL test as well as the state-of-art EDZL test.

Supporting parallelism in server-based multiprocessor systems

Developing an efficient server-based real-time scheduling solution that supports dynamic task-level parallelism is now relevant to even the desktop and embedded domains and no longer only to the high performance computing market niche. This paper proposes a novel approach that combines the constantbandwidth server abstraction with a work-stealing load balancing scheme which, while ensuring isolation among tasks, enables a task to be executed on more than one processor at a given time instant.

A traffic differentiation add-on to the IEEE 802.15.4 protocol: implementation and experimental validation over a real-time operating system

The IEEE 802.15.4 is the most widespread used protocol for Wireless Sensor Networks (WSNs) and it is being used as a baseline for several higher layer protocols such as ZigBee, 6LoWPAN or WirelessHART. Its MAC (Medium Access Control) supports both contention-free (CFP, based on the reservation of guaranteed time-slots GTS) and contention based (CAP, ruled by CSMA/CA) access, when operating in beacon-enabled mode. Thus, it enables the differentiation between real-time and best-effort traffic. However, some WSN applications and higher layer protocols may strongly benefit from the possibility of supporting more traffic classes. This happens, for instance, for dense WSNs used in time-sensitive industrial applications. In this context, we propose to differentiate traffic classes within the CAP, enabling lower transmission delays and higher success probability to timecritical messages, such as for event detection, GTS reservation and network management. Building upon a previously proposed methodology (TRADIF), in this paper we outline its implementation and experimental validation over a real-time operating system. Importantly, TRADIF is fully backward compatible with the IEEE 802.15.4 standard, enabling to create different traffic classes just by tuning some MAC parameters.

Competitive analysis of partitioned scheduling on uniform multiprocessors

Consider the problem of scheduling a set of sporadically arriving tasks on a uniform multiprocessor with the goal of meeting deadlines. A processor p has the speed Sp. Tasks can be preempted but they cannot migrate between processors. We propose an algorithm which can schedule all task sets that any other possible algorithm can schedule assuming that our algorithm is given processors that are three times faster.

Competitive analysis of static-priority partitioned scheduling on uniform multiprocessors

Consider the problem of scheduling a set of sporadically arriving tasks on a uniform multiprocessor with the goal of meeting deadlines. A processor p has the speed Sp. Tasks can be preempted but they cannot migrate between processors. On each processor, tasks are scheduled according to rate-monotonic. We propose an algorithm that can schedule all task sets that any other possible algorithm can schedule assuming that our algorithm is given processors that are √2 / √2−1 ≈ 3.41 times faster. No such guarantees are previously known for partitioned static-priority scheduling on uniform multiprocessors.

Linearizability and schedulability

Consider the problem of scheduling a set of tasks on a single processor such that deadlines are met. Assume that tasks may share data and that linearizability, the most common correctness condition for data sharing, must be satisfied. We find that linearizability can severely penalize schedulability. We identify, however, two special cases where linearizability causes no or not too large penalty on schedulability.

Uniprocessor EDF scheduling with mode change

Consider the problem of scheduling sporadically-arriving tasks with implicit deadlines using Earliest-Deadline-First (EDF) on a single processor. The system may undergo changes in its operational modes and therefore the characteristics of the task set may change at run-time. We consider a well-established previously published mode-change protocol and we show that if every mode utilizes at most 50% of the processing capacity then all deadlines are met. We also show that there exists a task set that misses a deadline although the utilization exceeds 50% by just an arbitrarily small amount. Finally, we present, for a relevant special case, an exact schedulability test for EDF with mode change.

Schedulability analysis of generalized multiframe traffic on multihop-networks comprising software-implemented ethernet switches

Consider a multihop network comprising Ethernet switches. The traffic is described with flows and each flow is characterized by its source node, its destination node, its route and parameters in the generalized multiframe model. Output queues on Ethernet switches are scheduled by static-priority scheduling and tasks executing on the processor in an Ethernet switch are scheduled by stride scheduling. We present schedulability analysis for this setting.

Object transportation by multiple mobile robots controlled by attractor dynamics: theory and implementation

Dynamical systems theory is used as a theoretical language and tool to design a distributed control architecture for teams of mobile robots, that must transport a large object and simultaneously avoid collisions with (either static or dynamic) obstacles. Here we demonstrate in simulations and implementations in real robots that it is possible to simplify the architectures presented in previous work and to extend the approach to teams of n robots. The robots have no prior knowledge of the environment. The motion of each robot is controlled by a time series of asymptotical stable states. The attractor dynamics permits the integration of information from various sources in a graded manner. As a result, the robots show a strikingly smooth an stable team behaviour.

Using attractor dynamics to generate decentralized motion control of two mobile robots transporting a long object in coordination

Dynamical systems theory is used here as a theoretical language and tool to design a distributed control architecture for a team of two mobile robots that must transport a long object and simultaneously avoid obstacles. In this approach the level of modeling is at the level of behaviors. A “dynamics” of behavior is defined over a state space of behavioral variables (heading direction and path velocity). The environment is also modeled in these terms by representing task constraints as attractors (i.e. asymptotically stable states) or reppelers (i.e. unstable states) of behavioral dynamics. For each robot attractors and repellers are combined into a vector field that governs the behavior. The resulting dynamical systems that generate the behavior of the robots may be nonlinear. By design the systems are tuned so that the behavioral variables are always very close to one attractor. Thus the behavior of each robot is controled by a time series of asymptotically stable states. Computer simulations support the validity of our dynamic model architectures.

On the development of a test-bed application for the ART-WiSe architecture

The ART-WiSe (Architecture for Real-Time communications in Wireless Sensor Networks) framework aims at the design of new communication architectures and mechanisms for time-sensitive Wireless Sensor Networks (WSNs). We adopted a two-tiered architecture where an overlay Wireless Local Area Network (Tier 2) serves as a backbone for a WSN (Tier 1), relying on existing standard communication protocols and commercial-off-the-shell (COTS) technologies – IEEE 802.15.4/ZigBee for Tier 1 and IEEE 802.11 for Tier 2. In this line, a test-bed application is being developed for assessing, validating and demonstrating the ART-WiSe architecture. A pursuit-evasion application was chosen since it fulfils a number of requirements, namely it is feasible and appealing and imposes some stress to the architecture in terms of timeliness. To develop the testbed based on the previously referred technologies, an implementation of the IEEE 8021.5.4/ZigBee protocols is being carried out, since there is no open source available to the community. This paper highlights some relevant aspects of the ART-WiSe architecture, provides some intuition on the protocol stack implementation and presents a general view over the envisaged test-bed application.

Repeater vs. Bridge-based hybrid wired/wireless PROFIBUS networks: a comparative performance analysis

In the last years, several solutions have been proposed to extend PROFIBUS in order to support wired and wireless network stations in the same network. In this paper we compare two of those solutions, one in which the interconnection between wired and wireless stations is made by repeaters and another in which the interconnection is made by bridges. The comparison is both qualitative and numerical, based on simulation models of both architectures.

Dpws middleware to support agent-based manufacturing control and simulation

Dissertação apresentada na Faculdade de Ciências e Tecnologia da Universidade Nova de Lisboa para obtenção do grau de Mestre em Engenharia Electrotécnica e de Computadores

On the use of the ZigBee protocol for wireless sensor networks

This project was developed within the ART-WiSe framework of the IPP-HURRAY group (http://www.hurray.isep.ipp.pt), at the Polytechnic Institute of Porto (http://www.ipp.pt). The ART-WiSe – Architecture for Real-Time communications in Wireless Sensor networks – framework (http://www.hurray.isep.ipp.pt/art-wise) aims at providing new communication architectures and mechanisms to improve the timing performance of Wireless Sensor Networks (WSNs). The architecture is based on a two-tiered protocol structure, relying on existing standard communication protocols, namely IEEE 802.15.4 (Physical and Data Link Layers) and ZigBee (Network and Application Layers) for Tier 1 and IEEE 802.11 for Tier 2, which serves as a high-speed backbone for Tier 1 without energy consumption restrictions. Within this trend, an application test-bed is being developed with the objectives of implementing, assessing and validating the ART-WiSe architecture. Particularly for the ZigBee protocol case; even though there is a strong commercial lobby from the ZigBee Alliance (http://www.zigbee.org), there is neither an open source available to the community for this moment nor publications on its adequateness for larger-scale WSN applications. This project aims at fulfilling these gaps by providing: a deep analysis of the ZigBee Specification, mainly addressing the Network Layer and particularly its routing mechanisms; an identification of the ambiguities and open issues existent in the ZigBee protocol standard; the proposal of solutions to the previously referred problems; an implementation of a subset of the ZigBee Network Layer, namely the association procedure and the tree routing on our technological platform (MICAz motes, TinyOS operating system and nesC programming language) and an experimental evaluation of that routing mechanism for WSNs.