A task parallel implementation of a scattered node stencil-based solver for the shallow water equations
Predicting the Cost of Lock Contention in Parallel Applications on Multicores using Analytic Modeling
Modern computer architectures, with multicore CPUs and GPUs or other accelerators, make stronger demands than ever on writers of scientific code. As a rule of thumb, the fastest, most efficient program consists of labor-intensive code written by expert programmers for a certain application on a particular computer. This thesis deals with several algorithmic and technical approaches towards effectively satisfying the demand for high-performance parallel programming without incurring such a high cost in expert programmer time. Effective programming is accomplished by writing performance-portable code where performance-critical functionality is provided either by external software or at least a balance between maintainability/generality and efficiency.
This paper addresses the problem of automatically verifying correctness of concurrent algorithms, e.g., as found inconcurrent implementations of common data structures, using model checking. In order to use a model checker to analyze programs in, e.g., C, one must first translate programs to the input language of the model checker. Since our aim is to use SPIN, we present an automated translation from a subset of C to Promela. This translation is able to handle features not covered by previous such translations, notable pointer structures and function calls. We illustrate the application of our translation to a concurrent queue algorithm by Michael and Scott.
In this paper, the optimally rigid topology control problem in wireless sensor networks is considered to improve the algebraic rigidity properties. This problem is first formulated as a constrained optimization problem which can be solved by two stages. A minimally rigid network is constructed in the first stage, then the optimally rigid topology in the second stage. A potential game approach is proposed for solving the optimization problem by choosing a different performance metric as the potential function. It can be seen that the proposed algorithm can significantly improve the network performance, such as reducing communication complexity and transmit power, prolonging network lifetime, and so on. Finally, some simulations demonstrate the effectiveness of the proposed algorithms from multiple perspectives: topology complexity, average degree, consensus convergence speed, average radius, average link length, and network lifetime.
In this paper a summary is given of the ongoing research at the Belgian Royal Military Academy in the field of mobile ad hoc networks in general and wireless sensor networks (WSNs) in particular. In this study, all wireless sensor networks are based on the physical and the medium access layer of the IEEE 802.15.4 low rate wireless personal area networks standard. The paper gives a short overview of the IEEE 802.15.4 standard in the beaconless mode together with a description of the sensor nodes and the software used throughout this work. The paper also reports on the development of a packet sniffer for IEEE 802.15.4 integrated in wireshark. This packet sniffer turns out to be indispensable for debugging purposes. In view of future applications on the wireless network, we made a theoretical study of the effective data capacity and compared this with measurements performed on a real sensor network. The differences between measurements and theory are explained. In case of geograph- ically meaningful sensor data, it is important to have a knowledge of the relative position of each node. In the last part of the paper we present some experimental results of positioning based on the received signal strength indicators (RSSI). As one could expect, the accuracy of such a method is poor, even in a well controlled environment. But the method has some potential.
In this paper we consider adaptive workflow nets, a class of nested nets that allows more comfort and expressive power for modeling adaptability and exception handling in workflow nets. We define two important behavioural properties of adaptive workflow nets: soundness and circumspectness. Soundness means that a proper final marking (state) can be reached from any marking which is reachable from the initial marking, and no garbage will be left. Circumspectness means that the upper layer is always ready to handle any exception that can happen in a lower layer. We define a finite state abstraction for adaptive workflow nets and show that soundness and circumspectness can be verified on this abstraction.
Signaling in communication networks is defined as the exchange of information between nodes to establish, maintain, and remove control state in network nodes. The concept of signaling is not new. The industry recognized the need for a way to create and remove circuits, each associated with an end-toend communications channel, for transporting information over long-haul networks. As a result they developed Signaling System 7 (SS7) for signaling in telephone networks. It nonetheless took until the last decade for network designers to use signaling to improve the ability of packet-switching networks to support emerging services, especially real-time services. With the increasing diversity of services offered across the Internet, there is a new need for signaling over IP-based networks. Examples include reserving resources to provide quality of service (QoS) guarantees, configuring firewall pinholes and network address translator (NAT) bindings, and diagnosing path status. The Internet Engineering Task Force (IETF) developed the Resource Reservation Protocol (RSVP) [1, 2], but RSVP has been designed and applied to resource reservation for both integrated services (IntServ) and later differentiated services (DiffServ), rather than more general signaling services. This led the IETF in 2001 to form a new working group, Next Steps in Signaling (NSIS), to investigate a more flexible IP signaling architecture and protocols. Due to the shortcomings of RSVP and its current extensions, the NSIS working group began to work on a new protocol suite in order to accommodate new signaling needs. As a result, an extensible IP signaling architecture [3] was developed, also referred to as NSIS. It consists of two layers: the lower layer provides a generic transport service for different signaling applications, which reside in the upper layer. The main part of the lower layer is the General Internet Signaling Transport (GIST) protocol [4]. Examples of upper layers (i.e., signaling applications) are QoS signaling, and firewall and NAT control. Furthermore, the working group decided to provide a mechanism to decouple next node discovery from signaling message delivery. This allows more flexibility, such as the ability to use standard transport layer and security protocols. This article provides information about current standardization efforts for the NSIS protocol suite. The rest of the article is organized as follows. In the next section we provide an overview of RSVP and point out why it needs to be updated. We then briefly describe the NSIS signaling approach. There follows a discussion of GIST and application protocols, particularly the QoS signaling protocol [5]. We also present security aspects and the implementation status of NSIS protocols, and compare them to RSVP, in particular with respect to QoS signaling.
Modern computer systems, as well as the Internet, use caching to maximize their efficiency. Nowadays, caching occurs in many different system layers. Analysis of these layers will lead to a deeper understanding of cache performance. ... comes from the uniprocessor environment. Spatial locality implies that the next data item in the address space is most likely to be used next, while temporal locality implies that the last data item used is most likely to be used next. Implementation is typically based on a fast but expensive memory (the price is affordable because, by definition, cache memory is small). Even if we use the same technology for the main memory and cache memory, the cache memory will be faster because smaller memories have a shorter access time. Recent research tries to split the CPU cache into two subcaches: one for spatial locality and one for temporal locality. # SMP On the SMP level, spatial and temporal
Presents the guest editorial for this issue of the publication.
