Sharif University of Technology

Real-time data processing systems are required to manage large volumes of data and deliver instant feedback. These systems are typically constructed on distributed processing architectures, where addressing the challenges of preventing deadlocks, avoiding divergence, ensuring liveness, and achieving goal reachability is highly complex before the architecture is implemented. This paper presents a framework for verifying formal models of a distributed and real-time stream processing architecture. It can be used to analyze the concurrent behavior of processes in stream data processing architectures. For the case study, a social network stream processing system was modeled. In the proposed method, Communicating Sequential Processes (CSP) and the Process Analysis Toolkit (PAT) were used to properties verification such as deadlock-free, divergence-free, liveness, and goal reachability before architecture implementation. The results indicate that our approach for real-time and distributed processing architecture, enables early detection of design errors in the initial stages, reduces costs, ensures real-time system constraints, identifies performance bottlenecks, and examines the behavior of concurrent system processes under various conditions.

In this paper, we investigate the robust control of a class of bilinear positive systems. The main objective is to minimize the norm of the controlled system while dealing with polytopic uncertainties. In contrast to traditional Lyapunov‐based robust control problems with polytopic uncertainties, we establish that minimizing the objective function over the vertices of the polytopic set, rather than considering its infinite elements, is adequate for achieving the globally optimal solution. Following this, we present a customized algorithm that relies on partial cutting plane to address the min–max problem. The introduced robust control strategy exhibits global convergence to the optimal solution and demonstrates scalability with the number of states and the number of vertices in the uncertain family. Simulation results validate the effectiveness of the proposed strategy in handling uncertainties.

About 150 years ago, two famous Iranian mystic and jurisprudent Najaf seminary people debated the interpretation of one of the poems of Attar Nishabouri. In this chapter, we try to outline the key points of the dispute between these two Hercules of contemporary Islamic mysticism and then try to extract out of this debate a conception of the relation between the unity of God and the manifold and plurality in His creatures to introduce a type of symbiotic perspective of a man–God relationship. To explain the social and even political role of human beings in the journey toward the ethical community, we appeal to the metaphor of the mystical journey from thirty birds [si murgh] toward unified Simurgh. If every person in society does her role perfectly, they will reach the Qaaf of a good and unified society.

This paper investigates a multi-cluster Wireless Powered Communication Network (WPCN) where user clusters cooperate with a Cluster Head (CH) and a Hybrid Access Point (HAP). Employing beamforming, the HAP supplies energy in the downlink phase, while users transmit signals to the HAP and CHs in the uplink phase. We optimize the Energy Beamforming (EB) matrix, user transmit covariance matrices, and time slot allocation to enhance both max-min and sum network throughput. To tackle the non-convexity of the optimization problems, we decompose them into two subproblems and reformulate each as convex Second Order Cone Programming (SOCP) and Quadratic Constraint Quadratic Programming (QCQP) for the max-min and sum throughput problems, respectively. Notably, we account for imperfections in Channel State Information (CSI) and non-linear Energy Harvesting (EH) circuits. Additionally, we incorporate the active Intelligent Reflecting Surfaces (IRS) as a key component in our proposed method. Numerical examples illustrate the significant impact of these contributions across various scenarios.

Network slicing, a key component of 5G, enables simultaneously running incompatible service types on a common infrastructure. Inter-slice isolation, as a key requirement of slicing, ensures that slice activity, i.e., containing a flow under transmission, does not affect the activity of other slices. Isolation in radio access network (RAN) slices is challenging due to interaction between slices. In fact, due to direct or indirect overlap among frequency channels in RAN slices, interference in inevitable, leading to interaction. In this paper, we propose an analytical model to analyze interference-coupled multi-cell RAN slicing where the interaction among slices results in dynamic behavior of slices. To this end, we map our scenario onto a suitable state-dependent queueing network, propose an iterative algorithm to obtain approximately the network steady-state probability distribution, and derive conventional QoS metrics (average delay and throughput). To quantify isolation, we define some new key performance indicators (KPIs) that show how changes in interfering slices affect the QoS metrics. Finally, we propose an interference-aware slice channel allocation policy that significantly reduces overlapping frequency channels. Numerical results demonstrate the accuracy of our analysis and the efficacy of the proposed policy in improving isolation-based KPIs compared to some other allocation policies.

With the ever-increasing demand for higher I/O performance and reliability in data-intensive applications, solid-state drives (SSDs) typically configured as redundant array of independent disks (RAID) are broadly used in enterprise all-flash storage systems . While a mirrored RAID offers higher performance in random access workloads, parity-based RAIDs (e.g., RAID5) provide higher performance in sequential accesses with less cost overhead. Previous studies try to address the poor performance of parity-based RAIDs in small writes (i.e., writes into a single disk) by offering various schemes, including caching or logging small writes. However, such techniques impose a significant performance and/or reliability overheads and are seldom used in the industry. In addition, our empirical analysis shows that partial stripe writes, i.e., writing into a fraction of a full array in parity-based RAIDs, can significantly degrade the I/O performance, which has not been addressed in the previous work. In this paper, we first offer an empirical study which reveals partial stripe writes reduce the performance of parity-based RAIDs by up to 6.85× compared to full stripe writes (i.e., writes into entire disks). Then, we propose a high-performance hyb rid RAID storage architecture, called HybRAID , which is optimized for write-intensive applications. HybRAID exploits the advantages of mirror- and parity-based RAIDs to improve the write performance. HybRAID directs a) aligned full stripe writes to parity-based RAID tier and b) small/partial stripe writes to the RAID1 tier. We propose an online migration scheme, which aims to move small/partial writes from parity-based RAID to RAID1, based on access frequency of updates. As a complement, we further offer offline migration, whose aim is to make room in the fast tier for future references. Experimental results over enterprise SSDs show that HybRAID improves the performance of write-intensive applications by 3.3× and 2.6×, as well as enhancing performance per cost by 3.1× and 3.0× compared to parity-based RAID and RAID10, respectively, at equivalent costs.

In this paper, a robust path and formation tracking controller is presented for cooperative load transportation by aerial robots. The proposed concept is composed of an Equilibrium Load Distributer and a Fast convergence Fuzzy Sliding Mode Controller (ELD-FFSMC). The ELD determines the nominal attitude of each agent for keeping the desired formation shape, which reduces the bound of uncertainty and the gain of switching law in the sliding mode controller. The FFSMC computes the corrective thrust and attitude for path and formation tracking and compensation of external disturbances. The role of fuzzy system is to eliminate the chattering phenomenon. The position of the payload is the only required value for the guidance algorithm leading to a distributed controller system without having to share information. In proposed ELD-FFSMC, the finite time convergence of error is proved by the Lyapunov method for a PID sliding surface, which presents even more robustness. The performance of the proposed ELD-FFSMC shows better precision and faster convergence time with respect to the existing controllers.

Asteroids may contain valuable minerals. A method to exploit asteroid mines is to transfer them closer to the Earth for further mining processes. In this work, we optimally mount a set of fixed-angle spacecraft thrusters on the surface of an asteroid to conduct concurrent detumbling and redirecting to the desired orbit. The optimization objective reconciles the minimum duration of the mission with the minimum required fuel as well as the maximum uniformity of the fuel distribution required for all thrusters. Each thruster can respond to redirection and detumbling commands simultaneously. Redirection and detumbling are performed via the directional adaptive guidance method and PID controllers, respectively, and the weight factors for each orbital element and the gains of the rotational control channels are also optimized in the process. We use the particle swarm optimization algorithm to evaluate the objective function by simulating the entire mission to find the optimal design. The rotational control damps the tumbling of the asteroid without interfering with the simultaneous redirection process and eventually fixes the asteroid in the optimally selected orientation in the inertial reference frame. The rotational velocity and attitude of the asteroid are controlled via separate PID controllers, which are set robustly. We can effectively optimize the mission by collectively tuning both the system’s rotational and redirection behaviors as well as the thrusters’ configuration and optimally selecting the final attitude of the asteroid.

Inrush current refers to the high-magnitude current drawn by a power transformer upon energization. The severity of inrush current is a function of the instantaneous value of voltage at the energization instant and the transformer's residual flux density. This paper proposes an effective energization method for mitigating the inrush current of single-phase power transformers. The method does not rely on the knowledge of transformer design specifications, but the magnitude of the transformer's excitation current. The reference residual flux density is determined with respect to the limitations of the closing operation of the circuit breaker. The method then adjusts the residual flux density of the core to a value deemed appropriate by injecting controlled current into the transformer's winding. This is followed by identifying an appropriate instant for transformer energization that matches the instantaneous value of the steady-state flux density with the adjusted flux density. To validate the efficiency of the proposed method, over 8,000 simulations are conducted in PSCAD/EMTDC. The method is also implemented on a laboratory-scale testbed and extensively tested to demonstrate its effectiveness and superiority over most recent methods under a wide variety of realistic conditions.

Online intelligent knowledge extraction from real-world nonstationary data streams presents a multiobjective optimization challenge. Here, we characterize the learning process on a trajectory of global optimality to simultaneously satisfy six high-profile objectives: 1) optimum generalization for the best bias-variance tradeoff; 2) compactness of knowledgebase; 3) memory retention and stability-plasticity balance; 4) universality and full autonomy; 5) robustness against outliers, noise, and model uncertainty; and 6) active concept drift detection and adaptation. We propose a flexible Takagi–Sugeno (TS) fuzzy system, named UFAREX, that self-constructs and self-guards from scratch in a non-iterative sample-wise training scheme without storing data. Through quantification of various uncertainties, an adaptive prediction interval (API) is sequentially learned for each local dynamism to automatically capture the most accurate compact representation with a 95% confidence. This leads to the best linear unbiased estimation (BLUE) of local trends. To avoid catastrophic forgetting, API collaborates with trapezoidal membership functions (TMFs) to expand local boundaries with maximum plasticity and without distortive extrapolation. As a robust detection mechanism, API also pinpoints regions in conflict (RIC) where concept drifts and outliers are actively expressed w.r.t. time of occurrence, location, type, and severity. This establishes a single-sample online active concept drift management with zero buffer latency for regression applications. No heuristic forgetting, pruning, splitting, merging, and weighting mechanisms are exercised to prevent human intervention and render universality. UFAREX was comparatively tested on four real-world benchmarks. It stands out as an autonomous system geared for adaptive modeling, time-series forecasting, anomaly monitoring, and robust fault detection and diagnosis.

Decellularized extracellular matrix (dECM) bioinks hold significant potential in the 3D bioprinting of tissue-engineered constructs (TECs). While 3D bioprinting allows for the creation of custom-designed TECs, the development of bioinks based solely on dAM, without the inclusion of supporting agents or chemical modifications, remains underexplored. In this study, we present the concentration-dependent printability and rheological properties of dAM bioinks, along with an analysis of their in vitro cellular responses. Our findings demonstrate that increasing dAM concentrations, within the range of 1 to 3% w/v, enhances the mechanical moduli of the bioinks, enabling the 3D printing of flat structures with superior shape fidelity. In vitro assays reveal high cell viability across all dAM bioink formulations; however, at 3% w/v, the bioink tends to impede fibroblast proliferation, resulting in round cell morphology. We propose that bioinks containing 2% w/v dAM strike an optimal balance, providing fine-resolved features and a supportive microenvironment for fibroblasts, promoting elongated spindle-like morphology and enhanced proliferation. These results underscore the importance of dAM concentration in regulating the properties and performance of bioinks, particularly regarding cell viability and morphology, for the successful 3D bioprinting of soft tissues.

Five metal dithiocarbamate complexes [M(PTHIQDTC)2] [where PTHIQDTC is (S)-1-phenyl-1,2,3,4-tetrahydroisoquinoline dithiocarbamate anion and M is Ni(II) (1), Sn(II) (2), Hg(II) (3), Pb(II) (4) and Zn(II) (5)] were synthesized from the reaction of MX2 (X is Cl– for 1–3 and OAc– for 4–5) with ligand of triethylammonium (S)-1-phenyl-1,2,3,4-tetrahydroisoquinoline dithiocarbamate [Et3NH][PTHIQDTC] in methanolic solution at room temperature. The five complexes were characterized by IR, ¹H and¹³C NMR, mass spectrometry, elemental analysis and TGA analysis. Recrystallization of [Zn(PTHIQDTC)2] (5) in dimethylsulfoxide (DMSO) converts 5 to [Zn(PTHIQDTC)2(DMSO)] (6). The structure of complex 6 has been determined by X-ray crystallography. The X-ray structural analysis of 6 indicated that the zinc(II) is five-coordinated in a distorted trigonal-bipyramidal by four S atoms from two (S)-1-phenyl-1,2,3,4-tetrahydroisoquinoline dithiocarbamate anion ligands (PTHIQDTC) and one O atom from DMSO. Supplementary Information The online version contains supplementary material available at 10.1038/s41598-024-81589-3.

The detection of the occurrence of working conditions in semiconductor manufacturing systems is fundamental to minimize yield losses and enhance production quality. The main challenges are the lack of data labeled with the machine state (normal/abnormal) and the nonlinear time evolution of the monitored signals. This work develops a novel methodology for the detection of abnormal conditions that, differently from the existing approaches, do not require the availability of labeled data. It consists of: (a) an approach based on k-fold cross-validation for automatically building a training set which does not contain data collected during abnormal conditions; (b) a signal reconstruction model based on stacked autoencoders with Long Short-Term Memory (LSTM) cells for reproducing the machine expected behavior in normal conditions; and (c) a decision module based on an abnormality indicator computed using the Mahalanobis distance. The proposed methodology is shown to outperform other state-of-the-art approaches in two case studies based on data taken from plasma etching machines.

To maintain stable glucose level, diabetic patients should monitor and control their glucose levels several times a day through insulin injections. However, it is important to acknowledge that errors can occur during such a process, which may affect the desired outcomes. The artificial pancreas is designed to overcome such undesirable situations using a closed-loop control approach. The vital impressing challenge to closed-loop control is dealing with the complexity of glucose–insulin model, including model order and nonlinearity. This paper presents a closed-loop reduced multiple model predictive control to address these complexities, aiming to ease the monitoring and controlling glucose level in type 1 diabetic patients. Considering the influence of meal patterns on glucose control, the reduced model predictive control system is capable of effectively maintaining a safe and stable glucose level. Additionally, cost function in model predictive control could be defined to optimize both the insulin dosage and the glucose level, ensuring a well-balanced approach between the two. To assess the effectiveness of the proposed method, simulations using the Hovorka nonlinear glucose–insulin system are conducted in this regard.

This research numerically and analytically examines the dynamics of underwater movable porous functional gradient (FG) microsize beams integrated with a piezoelectric layer. Also, the efficiency and accuracy of support vector machine (SVM) techniques in the vibration prediction of the microbeam are assessed. The dynamical simulation is conducted by considering the assumptions of Rayleigh beam theory, different porosity distribution models, nonlinear and linear stress–temperature relationships, and modified couple stress theory (MCST). The eigenvalues of the system, critical temperature increment, and critical speed of the microbeam are computed. Comparative studies are carried out, frequency analyses are performed, and stability diagrams are drawn. The impacts of microbeam geometry, fluid mass ratio, piezoelectric voltage, and axial force on stability behavior in variable complex environmental conditions are parametrically inspected. The outcomes revealed that the smaller the contribution of voids on the outer surface of the microbeam, the better the microbeam stability. It is understood that among the utilized regression-based SVMs, the quadratic polynomial and medium Gaussian kernel models have the highest performance and speed, respectively. These research outcomes will be advantageous for designing the next generation of targeted drug delivery devices.

Efficiently predicting the paratope holds immense potential for enhancing antibody design, treating cancers and other serious diseases, and advancing personalized medicine. Although traditional methods are highly accurate, they are often time-consuming, labor-intensive, and reliant on 3D structures, restricting their broader use. On the other hand, machine learning-based methods, besides relying on structural data, entail descriptor computation, consideration of diverse physicochemical properties, and feature engineering. Here, we develop a deep learning-assisted prediction method for paratope identification, relying solely on amino acid sequences and being antigen-agnostic. Built on the ProtTrans architecture, and utilizing pre-trained protein and antibody language models, we extract efficient embeddings for predicting paratope. By incorporating positional encoding for Complementarity Determining Regions, our model gains a deeper structural understanding, achieving remarkable performance with a 0.904 ROC AUC, 0.701 F1-score, and 0.585 MCC on benchmark datasets. In addition to yielding accurate antibody paratope predictions, our method exhibits strong performance in predicting nanobody paratope, achieving a ROC AUC of 0.912 and a PR AUC of 0.665 on the nanobody dataset. Notably, our approach outperforms structure-based prediction methods, boasting a PR AUC of 0.731. Various conducted ablation studies, which elaborate on the impact of each part of the model on the prediction task, show that the improvement in prediction performance by applying CDR positional encoding together with CNNs depends on the specific protein and antibody language models used. These results highlight the potential of our method to advance disease understanding and aid in the discovery of new diagnostics and antibody therapies.

The potential of Fe3O4 nanoparticles (NPs) can only be fully realized when highly stable water-dispersible particles with narrow particle size distribution and uniform morphology are produced in a simple sustainable and scalable process. Fatty acids have been successfully utilized as surfactants in the coprecipitation process to produce hydrophobic Fe3O4 NPs with desirable properties, which need additional processing to render them hydrophilic. In this study, an innovative single-step process was introduced for coating Fe3O4 NPs using two surfactant-like peptides, AC-A6K± and NH2-A6K± (±; refers to the coexisting positive and negative charge at the C-terminal of the peptides). This simplified process involves a single-step co-assembly of these peptides with aliphatic chains of palmitic acid (PA) at the surface of the Fe3O4 NPs. The impact of these modifications on the physicochemical properties of Fe3O4 NPs was thoroughly investigated, revealing that the peptide-coated particles exhibited enhanced dispersibility in aqueous environments, a more uniform morphology, nearly neutral zeta potential, and reduced aggregation, resulting in smaller and more uniform particles, when compared to lipid-stabilized NPs (Fe3O4 PA). Furthermore, the Fe3O4 PA NPs that were coated with AC-A6K± peptide (Fe3O4 PA/AC−A6K±) demonstrated superior coating efficiency and a more favorable particle size distribution than those coated with A6K± (Fe3O4 PA/A6K±). Notably, the crystal structure and magnetic saturation of the Fe3O4 PA NPs were preserved even after being assembled with AC-A6K± peptide. Our observations suggest that coating fatty acid-stabilized Fe3O4 NPs with surfactant-like peptides by a one-step co-assembly process is feasible and can produce magnetic nanoparticles (MNPs) with potential clinical application.