Sharif Digital Repository

The increasing complexity of embedded systems has led to the integration of tasks with various degrees of criticality on a common hardware platform called a mixed-criticality system. These systems typically exploit the inherent redundancy of multicore systems to employ the fault-tolerant techniques to satisfy the required target reliability. On the other hand, the use of fault-tolerant techniques increases the time that cores are simultaneously active with maximum power, which can violate the thermal design power (TDP) and exceed the safe temperature of the chip. This activates the dynamic thermal manage- ment (DTM) technique. Some of the most well-known methods to reduce the chip surface... 

Due to the increasing development of digital systems based on limited energy sources (i.e. battery), energy efficiency has become one of the most important concerns in the design of these systems. The use of multi-core architecture is an effective solution for the problem of reducing energy consumption. Hence using it in digital systems has become more common. In addition, enabling methods for reducing energy consumption on processor, helps in making energy more efficient. DVFS and DPM are the two major methods used for reducing dynamic and static energy consumption of processors. The using of multi-core architecture due to the higher chip density, results the static and dynamic energy... 

Since multi-core processors have become a primary trend in processor development, new scheduling algorithms are needed to minimize power consumption while achieving the desired timeliness guarantees for multi-core real-time embedded systems. Although various power/energy efficient scheduling algorithms have recently been proposed, existing studies may have degraded run-time performance in terms of power/energy efficiency and real-time guarantees when applied to real-time embedded systems with uncertain execution time. Moreover, these algorithms are only provided for a specific set of tasks, while many industrial applications with real-time demands are composed of mixed sets of tasks with a... 

Graphs are powerful tools for modeling real-world data. Data analysis using graphs allows us to study the samples relations and identify rich patterns. Although graph modeling can result in a better understanding of data, it requires having strong methods. Graph neural network models have attracted more attention in recent years. These networks are able to simultaneously analyze data features and their relationships to each other and find node representation in low-dimensional feature space. However, due to the novelty of this field, many challenges are still unexplored.In this study, we intend to examine the challenges in this field by focusing on improving the representation of nodes with... 

Distributed embedded systems are widely used in safety-critical applications, such as avionics platform and flight control system. Faults in such applications may cause catastrophic effect. Therefore, these systems must be fault-tolerant and must be designed not only to detect faults but also to recover from faults. Fault tolerance is normally realized on multi-processor system via temporal duplication (time-redundancy) or spatial duplication (hardware-redundancy) depending on the availability of the slack time. Slack time is defined as the difference between the deadline and execution time. Duplication is a common technique to achieve fault tolerance in safety-critical applications, but may... 

Hard real-time embedded systems often operate in harsh environmental conditions that necessitate fault tolerant computing techniques. These embedded systems are also energy-constrained, since energy consumption is one of the key objectives. Examples include satellites, surveillance systems, autonomous seaborne and airborne systems. Generally, fault tolerance is obtained by hardware or time redundancy in these systems. Checkpointing with rollback recovery which employs time redundancy is one of the common techniques for tolerating transient faults in real-time embedded systems. Using checkpointing in energy constrained hard real-time systems should be carefully employed. In this thesis, we... 

Recently, the tradeoff between low energy consumption and high fault-tolerance has attracted a lot of attention as a key issue in the design of real-time embedded systems. Dynamic Voltage Scaling (DVS) is known as one of the most effective low energy techniques for real-time systems. Control-theoretic DVS techniques, which have been exploited recently for DVS-enabled systems, are able to adapt well to dynamic workload. In this thesis, we have combined the above two important issues i.e., the control-theoretic DVS and fault-tolerance, and investigated reducing the energy consumption of fault-tolerant hard real-time systems using feedback control theory. Our proposed feedback-based DVS method... 

Energy consumption is one of the most crucial issues in embedded systems and memories are responsible for a considerable portion of it. Furthermore, leakage power is becoming the dominant part of total power consumption as the feature size of transistors is scaled down, therefore, the memory management must be improved in order to mitigate the negative effect of increased static power dissipation. In this study, a new memory architecture and a new data mapping algorithm are presented that reduce the energy consumption of multi-core embedded systems while respecting the timing constraints. We propose a novel two-level scratch-pad memory (SPM) based on STTRAM, which is a type of non-volatile... 

Over the Recent Decade, the embedded systems have expanded to include a wide variety of products, ranging from digital cameras, to medical systems, to Radar and telecommunication systems, to sensor networks. Engineers strive to create ever smaller and faster products, many of which, such as battery operated systems, have stringent power requirements. Coupled with increasing pressure to decrease costs and time-to-market, the design constraints of embedded systems pose a serious challenge to embedded systems designers. Dynamic reconfigurable hardware can provide a flexible and efficient platform for satisfying the area, performance, cost, and power requirements of many embedded systems.... 

Many embedded systems must be highly reliable and have hard real-time constraints. Technology scaling has enabled integration of multiple fast cores in a single chip. This provides higher computation speed that can achieve low response time in real-time systems. However, shrinking transistor dimensions aggravates reliability threats. Furthermore, hard real-time embedded systems are usually subjected to severe power and energy consumption limitations imposed by battery and cooling units. Therefore, designing hard real-time embedded systems requires careful considerations of reliability and power/energy consumption issues. High reliability can be achieved through exploiting fault tolerance... 

Sparse signal representations and its applications has been a hot topic of research in recent years. It has been demonstrated that sparsity prior can be effectively used as a regularization term to solve many of the inverse problems. One of these problems in which sparse representations have been used is image super-resolution (SR). SR is the problem of finding a high resolution (HR) image from one or several low resolution (LR) images. In this dissertation, we have focused on the problem of finding a HR image from only one LR image which is known as example-based SR. There are two kinds of methods for example-based SR: the methods which use neighborhood embedding and the methods which use... 

Soil-structure interaction may have a considerable effect on the response of the superstructure. One of the main steps in analyzing the soil-structure system is determining the soil impedances which vary with frequency. Although it is possible to analyze a dynamic system with complex frequency-dependent property matrices, this cannot be taken into account in most standard general-purpose softwares. In addition, the nonlinear behaviour of the structure can only be studied in time domain. Therefore, in the past two decades efforts have been made to replace the soil media with discrete models having a small number of degrees of freedom (DOFs) and a few springs, dashpots and masses with... 

In this research, it has been tried to use deep learning methods and embedding vector, in addition to user-item data, from user side information such as age, gender, city, etc., and also for item information such as product name, product category, etc. can be used to better understand customer behavior patterns and provide a relatively rich recommender system. The proposed model in this research has two phases, the first phase tries to identify the user and item feature vector and form the user similarity matrix and the user-item correlation matrix. The outputs of phase one are used as inputs of phase two. In the second phase of the model, using these inputs, Top-N recommendation are... 

Online advertisements have captured a major part of websites and mobile applications nowadays. In this kind of advertisements, as soon as a user interacts with a website or a mobile application, we have to decide which ad to show in a fraction of a second. In online advertising systems, revenue increases only when a user clicks on an ad banner or when they interact with the ad. So in order to select the best (most profitable) ad banner, the probability of clicking on or interacting with the ad banner must be estimated and then the most profitable ad will be selected.So one of the most essential problems in online display advertising, is estimating the probability of the click on a certain... 

Diffuse tensor imaging or tractography is a useful method for tracking the axonal tracts pathways within the brain white matter by monitoring the movements of water molecules along the axons. The higher the level of the tissue anisotropy, the more accurate the pathways can be estimated. But in the swelling regions around an edematous tumor, the excess of watery fluid disrupts the directional movement of water molecules, and consequently, the diffuse tensor imaging is unable to track the pathways. This impairment should be resolved by predicting the axontal tracts arrangement in the blind regions of the images using the numerical modeling. To this end, a finite element model of the human... 

One of the most widely used protocols used on the Internet is the SSL protocol, which is also used in many applications to exchange information between the server and the user. Therefore, the analysis of this traffic can help decision makers in many analyses. In this thesis, we are going to present a mapping for feature vectors extracted from SSL traffic that will lead to improving the performance of machine learning algorithms.In this treatise, three methods for learning mapping are proposed, all of which are based on deep learning. The first method is to use a simple self-encoder for map learning that tries to learn a compact map from the input feature vector.The second method is the... 

The main purpose of this study is to estimate the natural frequency and damping ratio for soil-structure systems with embedded foundation with sufficient accuracy. The results are then compared to guidelines in current seismic codes. This is done through an extensive parametric study using a set of non-dimensional parameters, which define the soil-structure systems. For this purpose a simplified soil-structure model is used. Although more rigorous models and numerical solutions are available, they are time consuming and not suitable for parametric study. Existing formula for seismic provisions for the equivalent period and damping ratio of soil-structure systems are based on studies on... 

Nowdays, embedded systems need high dependability and performance and low cost. Use of multicore processors cause increase performance, and can lead to increase dependability with its inherent redundancy. Also mixed critical leads to decrease cost and increase dependability, because designer just considers dependability techniques for high critical tasks. A lot of work to increase dependability of real-time multicore systems with mixed-criticality has been done. In this study, a comprehensive classification of the approaches that has been done in this area is provided. One of the issues that is expressed in muticore, is shared resources. In multicore platforms some resources like memory and... 

Compile-time optimizations play an important role in the efficient design of real-time embedded systems. Usually, compile-time optimizations are designed to reduce average-case execution time (ACET). While ACET is a main concern in high-performance computing systems, in realtime embedded systems concerns are different and worst-case execution time (WCET) is much more important than ACET. Therefore, WCET reduction is more desirable than ACET reduction in many real-time embedded systems. In this thesis, we propose a compile-time optimization method aimed at reducing WCET in real-time embedded systems. In the proposed method, based on the predicated execution capability of embedded processors,...