Network anomalies can arise due to various causes such as abnormal behaviors from users, malfunctioning network devices, malicious activities performed by attackers, software or botnets. With the emergence of machine learning and especially deep learning, many works in literature developed models that are able detect anomalies. However, these require massive amounts labeled data for model training may not be unknown anomalous traffic zero-day attacks. Unsupervised techniques autoencoder its...

Detection of malicious programs using hardware-based features has gained prominence recently. The tamper-resistant hardware metrics prove to be a better security feature than the high-level software metrics, which can easily obfuscated. Hardware Performance Counters (HPC), are inbuilt in most recent processors, often choice researchers amongst metrics. However, lack determinism their counts, thereby affecting malware detection rate, minimizes advantages HPCs. To overcome this problem, our...

Internet-of-Things (IoT) based solutions are gaining momentum in delivering efficient health care domain, reducing financial and physical burden on patients improving ease of treatment for physicians.One such smart solution Serious games.Serious games aid rehabilitation various fields.For rehabilitation, personalization is important training results.A scientific approach difficulty level design can facilitate players to get effective rehabilitation.The automation personalized helps the...

The nature of ever evolving anomalies have become more sophisticated and complex in attacking the defense schemes, thereby leading to serious compromises. Existing software based techniques aim protect a vulnerable with another which is also prone compromise like obfuscation-based attacks. On other hand, hardware performance counters offer robust detection mechanism that difficult since it easier tamper components than features. Hence, this paper, we propose hardware-based monitoring method...

Understanding the dynamic behavior of computer programs during normal working conditions is an important task, which has multiple security benefits such as development behavior-based anomaly detection, vulnerability discovery, and patching. Existing works achieved this goal by collecting analyzing various data including network traffic, system calls, instruction traces, etc. In paper, we explore use a new type data, performance counters, to analyze programs. Using existing primitives,...

Substance use disorder (SUD) is a dangerous epidemic that develops out of recurrent alcohol and/or drugs and has the capability to severely damage one's brain behaviour. Stress an established risk factor in SUD's development addiction reinstating drug seeking. Despite this expanding potential for its grave consequences, there are limited options available management treatment, as well pharmacotherapies psychosocial treatments. To end, need new improved devices dedicated detection,...

Security is a critical aspect in many of the latest embedded and IoT systems. Malware one severe threats security for such devices. There have been enormous efforts malware detection analysis; however, occurrences newer varieties malicious codes prove that it an extremely difficult problem given nature these surreptitious codes. In this article, instead addressing general solution, we aim at platforms more than core performance enhancement. We investigate utility multiple cores from point...

à la diffusion de documents scientifiques niveau recherche, publiés ou non, émanant des établissements d'enseignement et recherche français étrangers, laboratoires publics privés.

Inexact design has been recognized as very viable approach to achieve significant gains in the energy, area and speed efficiencies of digital circuits. By deliberately trading error return for such these gains, inexact circuits architectures have shown be especially useful contexts where our senses sight hearing, can compensate loss accuracy. It is therefore important understand, characterize manner which sensorial systems interact Further use this knowledge optimize guide inexactness...

Recently, compressive sensing has attracted a lot of research interest due to its potential for realizing lightweight image compression solutions. Approximate or inexact computing on the other hand been successfully applied lower complexity hardware architectures applications where certain amount performance degradation is acceptable (e.g. lossy compression). In our work, we present novel method using approximate paradigm, in order realize hardware-efficient architecture. We adopt Gaussian...

Malware detection is still one of the difficult problems in computer security because occurrence newer varieties malware programs. There has been an enormous effort developing a generalised solution to this problem, but little done considering resource constraint embedded devices. In paper, we at- tempt develop lightweight tool designed specifically for platforms using micro-architectural side-channel information obtained through Hardware Performance Counters (HPCs). The methodology aims...

AbstractDomino logic is often the choice for designing high speed CMOS circuits. Switching of gates in such a circuit, account much its dynamic power dissipation. Power minimization techniques like clock-gating, energy recovery clock, gating and so on have limitations namely penalty clock delay, overhead, synchronization issues on. In this work, we propose method implementing clock-gating to Domino logic-based using pattern recognition approach. For given first identify list favourable gate...

In recent times, the usage of Domino logic in design high performance circuits is increasing. addition to its advantage, style also offers flexibility designing individual cells. Flexible height and width cells gives designer an advantage realize a large variety functions. This scope for library free mapping. this work, we present approach mapping circuit using on-the-fly technique. First, node algorithm which maps given net list Next, Equivalence Table, re-order along critical path delay...

This Domino logic is often the choice for designing high speed CMOS circuits.Often VLSI designers choose library based approaches to perform technology mapping of large scale circuits involving static style.Cells designed using style have flexibility accommodate wide range functions in them.Hence, there a scope adopt free synthesis approach logic.In this work, we present an domino circuit On-the-fly technique.First, node algorithm which maps given netlist technique.Next, Equivalence Table,...

Modern customer-facing applications need to be easy use, localizable, and scale out serve large customer bases. Microservice architectures have the potential decentralize functionality, improve flexibility, provide faster time market of incremental changes. However, implemented as microservices also a larger surface area, which may make them more prone cyber attacks. operating systems performance counters are tamper-resistant, can used track run-time behavior applications. In this work, we...

Static CMOS logic style is often the choice of designers for synthesizing low power circuits.This robust in terms noise integrity however, it offers less speed.Domino style, as an alternative found critical paths various large scale high performance circuits.Yet, due to switching activity they are not suitable synthesis circuits.To achieve both and speed benefits, we propose a method designing circuit using mixed taking advantages static Domino styles.For given circuit, extract unate binate...

Major challenge in building models for stroke rehabilitation stems from the non stationarity of EMG signals. In this work we present a methodology improved classification hand gestures using Empirical Fourier Decomposition (EFD). First apply EFD technique on set publicly available dataset and later reduce dimensionality to collect most significant components. Finally extract features these components perform gesture classifications different machine learning (ML) models.Clinical Relevance:...

Principal Component Analysis (PCA) has been a major tool in performing characterization of environmental data where in, the is typically hyper spectral image. Using statistical methods, PCA often capable reducing dimensionality data. On other hand Nonlinear Iterative PArtial Least Squares (NIPALS) algorithm provides an efficient alternative for extracting principal components with minimum penalty on processing speed. In this work we provide hardware implementation NIPALS FPGA, given dataset....

Increase in number of embedded systems which are interconnected has created need for mechanisms can detect malicious exploits a lightweight yet speed efficient fashion. Recent approaches that addressed this challenge focused on utilizing either high level or low features along with machine learning algorithms to analyze behavior unknown programs. However, the less tamper resistant and result false positive. Moreover these often consume computational storage resources, suitable systems. In...

Understanding the dynamic behavior of computer programs during normal working conditions is an important task, which has multiple security benefits such as development behavior-based anomaly detection, vulnerability discovery, and patching. Existing works achieved this goal by collecting analyzing various data including network traffic, system calls, instruction traces, etc. In paper, we explore use a new type data, performance counters, to analyze programs. Using existing primitives,...