Fadia Noori Hummadi Al-Nuaimy

Fadia Noori Hummadi Al-Nuaimy

Auto-generated profile, based on publicly available information

Fadia Noori Hummadi Al-Nuaimy

Al-Khwarizmi Collage of Engineering, University of Baghdad, Baghdad, Iraq

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 May 2024 in Journal of Engineering and Sustainable Development
Verified
Journal Article
Open access
PDF file
MAMahmood Zaki Abdullah
Mahmood Zaki Abdullah
AJAli Jassim
Ali Jassim
FHFadia Noori Hummadi
Fadia Noori Hummadi
MKMohammed Majid M. Al Khalidy
Mohammed Majid M. Al Khalidy
Gradually, since the number of linked computer systems that use networks linked to the Internet is raised the information that is delivered through those systems becomes more vulnerable to cyber threats. This article presents proposed algorithms based on Machine Learning (ML) that ensure early detection of cyber threats that cause network breaking through the use of the Correlation Ranking Filter feature selection method. These proposed algorithms were applied to the Multi-Step Cyber-Attack Dataset (MSCAD) which consists of 66 features. The proposed strategy will apply machine learning algorithms like Adaptive Boosting-Deep Learning (AdaBoost-Deep Learning) or (ABDL), Multi-Layer Perceptron (MLP), Bayesian Networks Model (BNM), and Random Forest (RF), the feature would be decreased to high valuable of 46 features were included with a threshold of 0.1 or higher. The accuracy would be increased when the no. of features decreased to 46 with a threshold of ≥ 0.1 with the ABDL algorithm producing an accuracy of 99.7076%. The obtained results showed that the proposed algorithm delivered a suitable accuracy of 99.6791% with the ABDL algorithm even with a higher number of features.

1 August 2017
Verified
Conference Paper
Closed access
FAFadia Noori Hummadi Al-Nuaimy
Fadia Noori Hummadi Al-Nuaimy
The problems faced by the deaf and dumb people at the present time and the difficulties of their communication with normal persons sparked our interest and led us to try to find a solution to their difficulties and to minimize them as much as possible. Because they represent a significant part of society and they need to deliver their ideas in the simplest way by simple devices. So our project aims to bridge this gap by enabling communication between dumb\deaf people on the one hand and normal people on other hand by introducing an inexpensive electronic device that translates the fingers presses into the text and speech.

30 December 2022 in Eastern-European Journal of Enterprise Technologies
Verified
Journal Article
Open access
PDF file
EHEkhlas Hamza
Ekhlas Hamza
SASameir Aziez
Sameir Aziez
FHFadia Hummadi
Fadia Hummadi
ASAHMAD H. SABRY
AHMAD H. SABRY
Deep learning has recently been used for this issue with superior results in automatic modulation classification. Previous studies state that it is challenging to categorize a variety of modulation formats using traditional approaches; however, modulation classification is a crucial component of non-cooperative communication in wireless communication. The deep learning network was applied to solve the issue and get decent outcomes. This work uses a deep learning convolutional neural network (DLCNN) to classify three analog and eight digital modulation techniques by generating channel-impaired and synthetic waveforms as training data. The obtained DLCNN is tested by over-the-air indicators and a Software Define Radio(SDR) platform. The trained DLCNN estimates the modulation kind of each frame by taking 1024 samples of channel-impaired signals. The method includes generating several frames of 4-arry pulse amplitude modulation (PAM4) that are impaired with sampling time drift, Additive white Gaussian noise (AWGN), center frequency, and Rician multipath fading. The DLCNN predicts real inputs when receiving a signal with complex samples of baseband. Before updating the network coefficients and on all iterations, the data store transforms data from files and records it. This network takes about 50 minutes to train using in-memory data and 110 minutes to train using disk data. The evaluation of the trained DLCNN is carried out by obtaining the classification accuracy for the test frames. The obtained outcome demonstrates that the developed network can achieve an accuracy of about 94.3 % in roughly 12 epochs for such types of waveforms, which elapsed about 26 minutes for training. This will increase the efficiency of spectrum usage and detect the modulation type of the wireless communication receivers

27 October 2022 in Eastern-European Journal of Enterprise Technologies
Verified
Journal Article
Open access
SASameir Aziez
Sameir Aziez
EHEkhlas Hamza
Ekhlas Hamza
FHFadia Hummadi
Fadia Hummadi
ASAHMAD H. SABRY
AHMAD H. SABRY
Target scattering or reflection is used by radar systems to identify and detect targets. The larger the echo that was returned to the radar receiver, the superior the signal-to-noise ratio and the greater the likelihood of detection. The radar cross-section (RCS) determines the quantity of energy reflected from a target in radar systems. This work shows new modeling for radar targets with growing stages of fidelity. We introduce the RCS concept for straightforward point targets and extend this into additional complex states of targets with several scattering midpoints. In addition, we discuss the modeling of fluctuations in RCS with time and briefly consider the case of the polarized signal. Because the receiver and transmitter are co-located, the effort focuses on narrowband mono-static radar techniques. The RCS value changes between scans. We simulate the replicated power of a sent signal over 10,000 scanning over unit incident signals, assuming that the signal illuminates the target just once each dwell. The target is modeled by four scatterers that are placed at four square vertices. All scatterers are cylindrical-point targets on a 0.5-meter square XY plane without losing generality. The acquired results demonstrated how to generate target echoes while accounting for statistical fluctuations. From the relation between RCS and elevation angle variations for cylindrical targets, the obtained result demonstrated that the first two outputs are the same and confirmed that there is no reliance on azimuth angle. The comparison between wideband and narrowband RCS patterns demonstrated that the RCS profiles of the target-matched shallower nulls for azimuth direction are in the range of (40–50) degrees at zero elevation for 4 scatterers’ extended targets.

1 August 2022 in Indonesian Journal of Electrical Engineering and Computer Science
Verified
Journal Article
Open access
PDF file
ZGZainab Noori Ghanim
Zainab Noori Ghanim
FAFadia Noori Hummadi Al-Nuaimy
Fadia Noori Hummadi Al-Nuaimy
Data transmission in orthogonal frequency division multiplexing (OFDM) system needs source and channel coding, the transmitted data suffers from the bad effect of large peak to average power ratio (PAPR). Source code and channel codes can be joined using different joined codes. Variable length error correcting code (VLEC) is one of these joined codes. VLEC is used in mat lab simulation for image transmission in OFDM system, different VLEC code length is used and compared to find that the PAPR decreased with increasing the code length. Several techniques are used and compared for PAPR reduction. The PAPR of OFDM signal is measured for image coding with VLEC and compared with image coded by Huffman source coding and Bose-Chaudhuri-Hochquenghem (BCH) channel coding, the VLEC code decreases the data transmitted size and keep the same level of PAPR reduction with respect to data coded by Huffman and BCH code when using PAPR reduction methods.

1 October 2020 in TELKOMNIKA (Telecommunication Computing Electronics and Control)
Verified
Journal Article
Open access
PDF file
FAFadia Noori Hummadi Al-Nuaimy
Fadia Noori Hummadi Al-Nuaimy
Normally, the collected EEG signals from the human scalp cortex by using the non-invasive EEG collection methods were contaminated with artifacts, like an eye electrical activity, leading to increases in the challenges in analyzing the electroencephalogram for obtaining useful clinical information. In this paper, we do a comparison of using two decomposing methods (DWT and EMD) with CCA technique or High Pass Filter, for the elimination of eye artifacts from EEG. The eye artifacts (EOG) signals were extracted from the un-cleaned or raw EEG signals by DWT and EMD with CCA approach or H.P.F. The root means square error ratio of the uncontaminated EEG signal to the contaminated EEG signal with eye artifacts were the performance indicators for both elimination methods, which indicate that the combined CCA method outperforms the combined H.P.F method in the elimination of eye blinking contamination artifact from the EEG signal.