Search

In mobile ad-hoc networks (MANET), nodes are randomly distributed and move freely, and hence the network may face rapid and unexpected topological changes. In this paper, an improved greedy perimeter stateless routing protocol, called “LDAB-GPSR”, is proposed. LDAB-GPSR mainly focuses on maximizing the packet delivery ratio while minimizing the control overhead. In order to accomplish this, two techniques are introduced, the first one is the location prediction technique in which the greedy forwarding strategy is improved by choosing more stable routes for data forwarding. The second one is the adaptive beaconing technique in which the slow start algorithm is employed to adapt the beacon packet interval time based on the mobility of nodes and the data traffic load instead of using the periodic beaconing strategy. These two strategies together improve the overall performance of the GPSR routing protocol. The performance of the new proposed protocol is evaluated by carrying out several NS-2.35 simulation experiments. The simulation results show that LDAB-GPSR protocol outperforms the GPSR+Predict protocol in terms of packet delivery ratio, control traffic overhead, end to end delay, and throughput. The ratios of enhancement approaches 40%.

Background The administration of outpatient parenteral antimicrobial therapy (OPAT) has significantly increased in recent decades. OPAT offers many benefits to patients and the health care system. Best practices suggest patients on OPAT require active monitoring for adverse effects of therapy. Asymptomatic eosinophilia is common finding in patients receiving OPAT and could be benign or herald allergic drug reactions. Little is known about the practice patterns of OPAT providers in their management of asymptomatic eosinophilia. This study aimed to assess the opinions and practice patterns regarding asymptomatic eosinophilia among providers managing patients on OPAT in the United States. Methods This is a cross-sectional study of data collected from a one-time, anonymous, self-administered survey to providers who care for patients receiving OPAT. The survey was posted in the discussion forums of the Infectious Diseases Society of America (IDSA) and shared via emails to twenty ID departments in the US. Data was collected from June to July 2021. Survey responses were analyzed using Python for statistical analysis. Results Sixty-five respondents completed the survey. Thirty-six percent of respondents reported that eosinophilia occurs in up to five percent of their patients, with an average time of 1-2 weeks between antibiotic exposure and the development of eosinophilia Fig (1). Sixty-nine percent of respondents reported that they do not change their laboratory monitoring frequency in patients with asymptomatic eosinophilia. Fifty-two percent of respondents would discontinue/switch antibiotics only if the patient developed complications Fig (2). The risk of subsequent hypersensitivity reaction/end-organ damage was less frequently noted, as seventy-five percent encountered it in < 10% of eosinophilia cases Fig (1). Fig. 1Reported OPAT associated eosinophilia and subsequent hypersensitivity reaction per respondents Fig. 2 Practice patterns among survey respondents to eosinophilia Conclusion Per our survey results, OPAT-related eosinophilia is common though subsequent hypersensitivity/life-threatening reactions are not. Most surveyed ID providers monitor asymptomatic eosinophilia without changing antibiotics. Further prospective studies are needed to understand the course and best practice for asymptomatic eosinophilia among patients on OPAT. Disclosures All Authors: No reported disclosures.

Smart imaging-based medical classification systems help the human diagnose the diseases and make better decisions about patient health. Recently, computer-aided classification of skin diseases has been a popular research area due to its importance in the early detection of skin diseases. This paper presents at its core, a system that exploits convolutional neural networks to classify color images of skin lesions. It relies on a pre-trained deep convolutional neural network to classify between six skin diseases: acne, athlete’s foot, chickenpox, eczema, skin cancer, and vitiligo. Additionally, we constructed a dataset of 3000 colored images from several online datasets and the Internet. Experimental results are encouraging, where the proposed model achieved an accuracy of 81.75%, which is higher than the state of the art researches in this field. This accuracy was calculated using the holdout method, where 90% of the images were used for training, and 10% of the images were used for out-of-sample accuracy testing.

In this paper, we present the results of performance analysis for the Microsoft Azure virtual machines. The tests used NASA’s well known NAS parallel benchmarks for High Performance Computing (HPC) applications. Several tests were executed using different HPC cluster configuration. Both computation and communication times were measured to study the effect of Virtual Machines (VMs) computing and networking features on the execution and communication time of the benchmarks. We concluded that the communication behavior of any benchmark or application has the most influence on the HPC cluster configuration selection.

This paper introduces an extensive overview of the Infrastructure as a Service offerings that are presented today by the top ranked cloud providers. It surveys their solutions related to Infrastructure as a Service and High Performance Computing, presenting their benefits and drawbacks from a cloud user perspective. Four leading cloud providers were chosen for this paper, Amazon Elastic Compute Cloud, Microsoft Azure Cloud, Google Cloud and Oracle Cloud. The selected service providers are compared to each other in this paper from many perspectives. The comparison was done from the IaaS perspective. Reviews about these providers are presented in addition to pricing and billing details. Each one of these providers has its competitive advantages that already persuade its existing customer to invest more in their cloud and to attract new ones to migrate their resources to. In addition to the IaaS perspective, we present a comparison between the selected cloud providers from the High Performance Computing user perspective. These solutions are discussed; features of these solutions such as virtual machines options, their operating systems, and networking are compared.

This paper reviews the top leading cloud providers, surveys their offering related to High Performance Computing (HPC). Four top cloud provider are selected, Microsoft Windows Azure, Amazon Elastic Compute Cloud (Amazon EC2), Google Compute Engine and Oracle Cloud. Each one of them has its unique value proposition that enables it to survive in the market and make it a real competitor. A comparative analysis of the offerings and relative benefits of each are provided. This study is an introduction to our extensive work in the HPC field.

A novel approach to managing a fully distributed cognitive radio network (CRN) is presented. This approach builds on the concept of history-based spectrum access, in which cognitive base stations (BSs) independently estimate the system load using history records and adaptively swap their occupied spectrum bands to ensure allocation fairness and high overall throughput. In addition, cognitive BSs monitor primary user (PU) behavior in order to avoid interfering with active PUs. In this work, we address two issues that afflict history-based access: the first is the high cost of the sensing devices needed at each cognitive BS to be able to independently draw conclusions about the status of the CRN and the second is the unreliability inherent in practical sensing hardware (such as energy detectors). Simulation results show that the proposed technique manages to solve the two abovementioned issues without any noticeable drop in performance and without sacrificing the distributed nature of the protocol.

Array data flow analysis (ADFA) is a classical method for collecting array section information in sequential programs. When applying ADFA to parallel OpenMP programs, array access information needs to be analyzed in loops whose iteration spaces are partitioned across threads. The analysis involves symbolic expressions that are functions of the original loop iteration spaces, subscript expressions and thread numbers. Adequate representations of, and operations on, these expressions can be critical for the accuracy of the analysis. This paper presents a new ADFA compiler framework for OpenMP programs. We introduce the \(\pi \) operator to abstractly represent the parallelism effects in array section expressions and improve the accuracy of the cross-thread analysis during data flow computation. We also present a novel delayed symbolic evaluation technique that enables all array section operations in the data flow computation to be performed fully accurately. Using four NAS OpenMP benchmarks, we show that the \(\pi \) operator improves array section operations’ accuracy (i.e., reduces conservative operations) during data flow computation by \(66\) %, on average, compared to the best alternative. In addition, it reduces the number of terms, and thus the complexity of computed array sections by \(33\) %, on average. We also show that delayed symbolic evaluation eliminates conservative operations and does so without significant increase in complexity when combined with \(\pi \) operators.

OpenMP is an explicit parallel programming model that offers reasonable productivity. Its memory model assumes a shared address space, and hence the direct translation - as done by common OpenMP compilers - requires an underlying shared-memory architecture. Many lab machines include 10s of processors, built from commodity components and thus include distributed address spaces. Despite many efforts to provide higher productivity for these platforms, the most common programming model uses message passing, which is substantially more tedious to program than shared-address-space models. This paper presents a compiler/runtime system that translates OpenMP programs into message passing variants and executes them on clusters up to 64 processors. We build on previous work that provided a proof of concept of such translation. The present paper describes compiler algorithms and runtime techniques that provide the automatic translation of a first class of OpenMP applications: those that exhibit regular write array subscripts and repetitive communication. We evaluate the translator on representative benchmarks of this class and compare their performance against hand-written MPI variants. In all but one case, our translated versions perform close to the hand-written variants.