Ran Giladi

SPEC as a performance evaluation measure

Potential computer system users or buyers usually employ a computer performance evaluation technique only if they believe its results provide valuable information. System Performance Evaluation Cooperative (SPEC) measures are perceived to provide such information and are therefore the ones most commonly used. SPEC measures are designed to evaluate the performance of engineering and scientific workstations, personal vector computers, and even minicomputers and superminicomputers. Along with the Transaction Processing Council (TPC) measures for database I/O performance, they have become de facto industry standards, but do SPECs evaluation outcomes actually provide added information value? In this article, we examine these measures by considering their structure, advantages and disadvantages. We use two criteria in our examination: are the programs used in the SPEC suite properly blended to reflect a representative mix of different applications, and are they properly synthesized so that the aggregate measures correctly rank computers by performance? We conclude that many programs in the SPEC suites are superfluous; the benchmark size can be reduced by more than 50%. The way the measure is calculated may cause distortion. Substituting the harmonic mean for the geometric mean used by SPEC roughly preserves the measure, while giving better consistency. SPEC measures reflect the performance of the CPU rather than the entire system. Therefore, they might be inaccurate in ranking an entire system. To remedy these problems, we propose a revised methodology for obtaining SPEC measures. >

Performance Analysis of Software-Defined Networking (SDN)

Software-Defined Networking (SDN) approaches were introduced as early as the mid-1990s, but just recently became a well-established industry standard. Many network architectures and systems adopted SDN, and vendors are choosing SDN as an alternative to the fixed, predefined, and inflexible protocol stack. SDN offers flexible, dynamic, and programmable functionality of network systems, as well as many other advantages such as centralized control, reduced complexity, better user experience, and a dramatic decrease in network systems and equipment costs. However, SDN characterization and capabilities, as well as workload of the network traffic that the SDN-based systems handle, determine the level of these advantages. Moreover, the enabled flexibility of SDN-based systems comes with a performance penalty. The design and capabilities of the underlying SDN infrastructure influence the performance of common network tasks, compared to a dedicated solution. In this paper we analyze two issues: a) the impact of SDN on raw performance (in terms of throughput and latency) under various workloads, and b) whether there is an inherent performance penalty for a complex, more functional, SDN infrastructure. Our results indicate that SDN does have a performance penalty, however, it is not necessarily related to the complexity level of the underlying SDN infrastructure.

New Error-Resilient Scheme Based on FMO and Dynamic Redundant Slices Allocation for Wireless Video Transmission

Video quality suffers significant degradation when transmitted over error-prone channel, due to packet loss, errors caused by fading in wireless channel and due to the video codec prediction mechanisms. The H.264/AVC standard suggests some new error-resilient features to enable reliable transmission of compressed video signal over lossy packet networks. Two of those new features are the Redundant Slices and the Flexible Macro-Block Ordering (FMO). In this paper we propose a new error-resilient scheme which merges the H.264/AVC FMO feature with a new technique for dynamic allocation of redundant slices depending on the wireless channel fading parameters. We suggest using a unique smart dynamic redundant slices allocation scheme which considers the dynamic wireless channel parameters rather than using the classical standard static allocation. The proposed redundant slice allocation algorithm is based on both Average Fade Duration (AFD), and Level Cross Rate (LCR) channels characteristics. Moreover, we propose a new Explicit Spiral-Interleaved (ESI) flexible macroblocks ordering technique, which outperforms all other FMO types. The new ESI ordering results in effective error scattering which maximize the number of correctly received macroblocks located around corrupted macroblocks, leading to better error concealment. The proposed scheme greatly improves video transmission quality over lossy wireless transmission channels. Simulations results for wireless channel characterized by Rayleigh fading indicate that the proposed method improves the standard static allocation of redundant slices in terms of PSNR by about 2.5dB. Performance evaluations show that our approach is especially suited for applications such as video conferencing and mobile TV, where typically a specific main important Region of Interest should be more carefully protected

Characteristics' prediction in urban and suburban environments

The characteristics of the propagation in the microcellular urban and suburban environments in the 902-928-MHz frequency band in line-of-sight (LOS) conditions is investigated both theoretically and experimentally for the purpose of wireless radio local loop (WRLL) prediction. The path-loss characteristics and the range of a break point, at which the polynomial character of field intensity decay along the street level is changed, are analyzed for various parameters of street widths, for different average building heights, and for the actual electrical impedance properties of building walls. A multislit waveguide with randomly distributed gaps between the sides of buildings is considered as a model of straight streets. Results of experiments for VHF/UHF wave propagation along the straight streets in urban environment in the conditions of direct visibility between receiver and transmitter are compared with theoretical analysis of field intensity decay, path-loss distribution, and dependence of the break point on street topography in LOS conditions.

Determination of an implementation order for IS projects

Abstract The process of information systems (IS) planning in an organization occurs when there is a set of IS projects that require implementation, generally within certain time frames and budgetary constraints. We propose a method for determining an optimal order of project implementation. The proposed method utilizes the cost-benefit graph, which considers the expected cost and benefit of each of the projects comprising the set, as well as the weight (importance) of the cost and benefit factors to the decision-makers. The method allows us to analyze the sensitivity of a preferred implementation order to changes in cost and benefits, and the importance of these factors.

An architecture of an intelligent system for routing user requests in a network of heterogeneous databases

We present a general purpose model for routing user requests, e.g. queries, in a network of autonomous heterogeneous databases. The database schemas and other information on the database nodes are used to construct a multi-level knowledge-base (MKB) that resides in various nodes. Access to the databases is not done by creating direct connections between the user and the nodes where the data are presumably located. Rather, the user approaches the network by contents via an intelligent system that utilizes the MKB in order to identify the nodes and databases where the most relevant information resides, and establishes access routes to those nodes.

Multidimensional scaling: An analysis of 1980-1990 computers

Abstract Computer performance is an important issue for engineering and economic aspects of computer usage, planning, design and research. The Co-plot methodology graphically relates attributes, observations and interrelations between attributes, thus enabling the simultaneous study of all the observations and variables of a given data set. An analysis of performance attributes of computers from the 1980s (1981–1991) was carried out according to the Co-plot methodology. The analysis shows that during this decade, computer performance became more dependent on IO rate, cache size and multiprocessing, and less dependent on memory-size and IO addressing capabilities. At the same time, the high correlation between CPU speed and performance was maintained.

Placement of network resources in communication networks

Data communication and telecommunication networks consist of nodes, links and network resources. Many studies have dealt with the design of optimal network topologies. In current methodologies, every node presents the same constraints and behavior, however, in reality, the network resources vary. The research problem at stake here is how to add central network resources (i.e., servers, gateways, network managers, probes, mainframes, etc.) in an optimal way. For small networks or for a small number of resources, it is feasible to solve the problem by checking all the possibilities of locating the resources and choosing the optimal location. However, this method cannot be applied for large networks, where a great many of resources are located. In such cases, we confront a difficult optimization problem that is classified as NPcomplete. In this study we deal with optimal and heuristics techniques in order to solve this kind of problem.

Evaluating cost/utilization of organizations' information systems and users

Abstract The cost/utilization evaluation model of Borovits and Ein-Dor [6] is extended to include the users of the information system, thus enabling all overall organizational viewpoint of cost/utilization. The utilization evaluation process is segmented vertically and horizontally: three levels of utilization evaluation are defined vertically — the computer system, the information system, and the organization — and three major components of the information system are recognized horizontally — hardware, software, and users. The users are further subdivided into “peopleware” and endusers. Adopting this segmentation allows the use of numerical indices and graphical representations that make it easy to identify insufficient use of resources, bottlenecks, and suboptimized resource sharing from an overall organizational viewpoint.

A programmable, generic forwarding element approach for dynamic network functionality

Communication networks are growing exponentially, and new services and applications are being introduced unceasingly. To meet the demands of these services and applications, current network systems have to be modified, replaced or supplemented. Various technologies, such as reconfigurable devices or active networks, have attempted to address this problem. In this paper, we introduce a programmable, generic forwarding element (GFE), which can be used as a platform for a flexible and reconfigurable network system. This platform and the resulting network system enable on-the-fly definition of adaptive and dynamic network functionalities, so that the demands of new services and applications can be met. Additionally, specific service instances or traffic flows can be handled by this platform on a temporary and locality basis, according to traffic patterns, application demands, and provisioning decisions. The proposed GFE complies with todays standards and can easily be adopted for future standards. A network processor is used to implement this platform, so that frame processing is achieved at wire speed, even though each frame is analyzed and processed by a meta-program. An XML-based definition of the forwarding element is used to describe frame processing, based on the frame contents and ingress port, and on various system and network parameters.