Sharif Digital Repository

While performance and power consumption of processors present a classic trade-off in designing embedded hardware, software can be optimized in favor of both performance and energy. We evaluate the impact of optimizations at different stages of designing embedded software. We show that algorithm choice and compiler optimizations aimed at improving performance can also reduce energy consumption of an embedded processor. We also propose energy-aware compilation guidelines which can further reduce energy consumption without performance penalties. Our experimental results show that up to 85% energy reduction and 89% performance improvement can be achieved by these techniques  

Users often have very specific visual content in mind that they are searching for. The most natural way to communicate this content to an image search engine is to use key-words that specify various properties or attributes of the content. A naive way of dealing with such multi-attribute queries is the following: train a classifier for each attribute independently, and then combine their scores on images to judge their fit to the query. We argue that this may not be the most effective or efficient approach. Conjunctions of attribute often correspond to very characteristic appearances. It would thus be beneficial to train classifiers that detect these conjunctions as a whole. But not all... 

Efficient combination of MIMO (multiple-input multiple-output) and PCL (passive coherent location) ideas is expected to improve performance of localization schemes. While, in general, adding the number of transmit and receive antennas provides spatial diversity, it is possible to obtain similar effects by using multiple transmitters that are already transmitting standard signals (such as digital TV (DTV) transmission) in the environment (also known as illuminators of opportunity). However, in the case of passive schemes, it is not always possible to ensure that signals of different transmitters are orthogonal to each other. In such cases in which nonorthogonal transmitter signals are to be... 

In this paper, random access procedure in terrestrial trunked radio (TETRA) networks is described and modeled. Then, the impact of different random access parameters on performance metrics such as "delay" and "throughput" is evaluated. Moreover, in order to have a fast "call setup" which is one of the most important features of TETRA, it is shown that random access delay can be minimized, while maintaining peak throughput. by adaptively adjusting the random access parameters according to the backlogged traffic. The performance improvement of the proposed adaptive parameter adjustment is proved via extensive simulation results  

The Tor anonymous network has become quite popular with regular users on the Internet. In the Tor network, an anonymous path is created by selecting three relays through which the connection is redirected. Nevertheless, as the number of Tor users has increased substantially in recent years, the algorithm with which the relays are selected affects the performance provided by the Tor network. More importantly as the performance suffers, users will leave the network, resulting in a lower anonymity set and in turn lower security provided by Tor network. In this paper, we proposed an algorithm for improving performance and security of the Tor network, by employing a combination of different... 

Emerging cloud workloads in today's modern data centers have large memory footprints that make the processor's caches to be ineffective. Since L1 data cache is in the critical path, high data cache miss rates degrade the performance. To fix the issue in traditional workloads, data prefetchers predict the needed data to hide the memory latency and ultimately improve performance. In this paper, we focus on the L1 data cache to answer the question on why state-of-the-art prefetching methods are inefficient for modern workloads in terms of performance and energy consumption? This is because L1 cache is the most important player affecting the processor performance. Results show that, on the one... 

Particle Swarm Optimization (PSO) is a well-known swarm intelligence (SI) algorithm employed for solving various optimization problems. This algorithm suffers from premature convergence to local optima. Accordingly, a number of PSO variants have been proposed in the literature. These algorithms exploited different schemes to improve performance. In this paper, we propose a new variant of PSO with an enhanced Learning strategy and Crossover operator (PSOLC). This algorithm applies three strategies, comprising altering the exemplar particles, updating the PSO parameters, and integrating PSO with Genetic Algorithm (GA). In the proposed learning strategy, each particle is guided by the best... 

Multilevel converters are mainly used in high power applications (usually more than a few hundred kW). It is therefore important that they can continue working under faulty condition. In this paper, a new method is proposed to improve performance of CHB inverter under faulty condition. The proposed method will be compared with existing methods in term of voltage availability. The converter is simulated in Simulink and the results are provided. These results confirm the effectiveness of the proposed scheme, and are in accordance with the theoretical results  

Process variation has already emerged as a major concern in design of multi-processor system on chips (MPSoC). In recent years, there have been several attempts to bring variability awareness into the task scheduling process of embedded MPSoCs to improve performance yield. This study attempts to provide a comparative study of the current variation-aware design-time task and communication scheduling techniques that target embedded MPSoCs. To this end, the authors first use a sign-off variability modelling framework to accurately estimate the frequency distribution of MPSoC components. The task scheduling methods are then compared in terms of both the quality of the final solution and the... 

Everyday, people share their digital media on the virtual networks; therefore, protecting them against piracy is worthy of consideration. Digital watermarking is a method to achieve this goal. In this paper, we propose a watermarking method in Discrete Cosine Transform (DCT) domain. For this purpose, DCT coefficients of the whole image are calculated. It helps to present a system without blockiness effects. Then, singular values of the watermark image are added to the low frequency DCT coefficients of cover image that are located on the main diagonal. We use scaling factors for watermark embedding to improve imperceptibility of the watermark. A vector of fixed elements, which is calculated... 

One of the most prominent uses of ground tire rubber (GTR) is in rubber-modified asphalt industry. Besides improving performance properties of binders, GTR induces some drawbacks like weak storage stability and inferior low temperature (LT) performance. This study sought to rectify these shortcomings through using devulcanized rubber (DVR) in place of common GTR. Binder performance grading, multiple stress creep and recovery, and storage stability tests were conducted. More resistance against traffic loads at high temperatures was observed as DVR's advantages over GTR, while its weaker LT characteristics and lower storage stability can prevent it from being a viable alternative. © 2017... 

Design and modeling of a class of scanning near-field optical microscope (SNOM) probe for improving performance are reported. The basic idea is to generate more than one hot spot with well-known spatial distribution. This not only enables one to calibrate out many sorts of errors in conventional SNOM but also results in the increase in the speed of imaging. We utilize plasmonic nanocolor sorter structures to design aperture probes that create two spatially distinct hot spots. The method of moment analysis technique is employed to tune the probe for operation at standard Ar-ion laser wavelengths (457.9 and 514.5 nm). © 2018 Optical Society of America  

Nowadays, data and information are recognized as a precious resource in an organization. Data quality indicators help organizations manage the quality of data quantitatively and improve organizational processes. Organizations manage data and information with the help of information systems and make decisions within the framework of data collected and analyzed. On the other hand, continuous evaluation of the quality of data flow in systems can lead to a preventive program to formulate strategies for improving performance. This can be done effectively and efficiently in the form of control charts. In this paper, control charts are employed to monitor the quality of data flow in information... 

Progressive use of Web based information retrieval systems such as general purpose search engines and dynamic nature of the Web make it necessary to continually maintain Web based information retrieval systems. Crawlers facilitate this process by following hyperlinks in Web pages to automatically download new and updated Web pages. Freshness (recency) is one of the important maintaining factors of Web search engine crawlers that takes weeks to months. Many large Web crawlers start from seed pages, fetch every links from them, and continually repeat this process without any policies that help them to better crawling and improving performance of those. We believe that data mining techniques... 

Heat source addition to the axisymmetric supersonic inlet may improve the performance parameters, which will increase the inlet efficiency. In this investigation the heat has been added to the flow field at some distance ahead of an axisymmetric inlet by adding an imaginary thermal source upstream of cowl lip. The effect of heat addition on the drag coefficient, mass flow rate and the overall efficiency of the inlet have been investigated. The results show that heat addition causes flow separation, hence to prevent this phenomena, roughness has been added on the spike surface. However, heat addition reduces the drag coefficient and the inlet mass flow rate considerably. Furthermore, the... 

Electric vehicles (EVs) usage is becoming ubiquitous nowadays. Widespread integration of electric vehicles into electric energy distribution systems (EEDSs) has a twofold impact: (1) It may impose a burden on EEDSs when EVs are charging, (2) EVs could discharge their battery capacity using V2G technology when required. To mitigate adverse effects of massive integration of EVs in EEDSs, EVs could be used as mobile energy storage devices (MESDs) to transfer electric energy throughout EEDSs using a proper charging/discharging scheme. In this paper, a mixed integer linear programming (MILP) model is proposed to control charging and discharging of EVs to improve EEDS performance. EVs are modeled...