Rasha Mostafa

Dr. Rasha Mostafa

orcid

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

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

Representative Publications

View All Publications
27 February 2025 in Scientific Reports
Verified
Journal Article
Open access
PDF file
EAEssam Abdellatef
Essam Abdellatef
RMRasha Mostafa
Rasha Mostafa
WSWafaa A. Shalaby
Wafaa A. Shalaby
Human Activity Recognition (HAR) plays a critical role in fields such as healthcare, sports, and human-computer interaction. However, achieving high accuracy and robustness remains a challenge, particularly when dealing with noisy sensor data from accelerometers and gyroscopes. This paper introduces HARCNN, a novel approach leveraging Convolutional Neural Networks (CNNs) to extract hierarchical spatial and temporal features from raw sensor data, enhancing activity recognition performance. The HARCNN model is designed with 10 convolutional blocks, referred to as “ConvBlk.” Each block integrates a convolutional layer, a ReLU activation function, and a batch normalization layer. The outputs from specific blocks “ConvBlk_3 and ConvBlk_4,” “ConvBlk_6 and ConvBlk_7,” and “ConvBlk_9 and ConvBlk_10” are fused using a depth concatenation approach. The concatenated outputs are subsequently passed through a 2 × 2 max-pooling layer with a stride of 2 for further processing. The proposed HARCNN framework is evaluated using accuracy, precision, sensitivity, and f-score as key metrics, reflecting the model’s ability to correctly classify and differentiate between human activities. The proposed model’s performance is compared to traditional pre-trained Convolutional Neural Networks (CNNs) and other state-of-the-art techniques. By leveraging advanced feature extraction and optimized learning strategies, the proposed model demonstrates its efficacy in achieving accuracy of 97.87%, 99.12%, 96.58%, and 98.51% for various human activities datasets; UCI-HAR, KU-HAR, WISDM, and HMDB51, respectively. This comparison underscores the model’s robustness, highlighting improvements in minimizing false positives and false negatives, which are crucial for real-world applications where reliable predictions are essential. The experiments were conducted with various window sizes (50ms, 100ms, 200ms, 500ms, 1s, and 2s). The results indicate that the proposed method achieves high accuracy and reliability across these different window sizes, highlighting its ability to adapt to varying temporal granularities without significant loss of performance. This demonstrates the method’s effectiveness and robustness, making it well-suited for deployment in diverse HAR scenarios. Notably, the best results were obtained with a window size of 200ms.

7 May 2022 in Journal of Ambient Intelligence and Humanized Computing
Verified
Journal Article
Open access
PDF file
HGHanan S. Ghanem
Hanan S. Ghanem
RMRasha Mostafa
Rasha Mostafa
WEWalid El-Shafai
Walid El-Shafai
MEMaha Elsabrouty
Maha Elsabrouty
HHHesham Hamed
Hesham Hamed
FEFathi Abd El-Samie
Fathi Abd El-Samie
Convolutional Neural Networks (CNNs) are efficient tools for pattern recognition applications. They have found applications in wireless communication systems such as modulation classification from constellation diagrams. Unfortunately, noisy channels may render the constellation points deformed and scattered, which makes the classification a difficult task. This paper presents an efficient modulation classification algorithm based on CNNs. Constellation diagrams are generated for each modulation type and used for training and testing of the CNNs. The proposed work depends on the application of Radon Transform (RT) to generate more representative patterns for the constellation diagrams to be used for training and testing. The RT has a good ability to represent discrete points in the spatial domain as curved lines. Several pre-trained networks including AlexNet, VGG-16, and VGG-19 are used as classifiers for modulation type from the spatial-domain constellation diagrams or their RTs. Several simulation experiments are presented in this paper to compare different scenarios for modulation classification at different Signal-to-Noise Ratios (SNRs) and fading channel conditions.

25 February 2020 in International Journal of Communication Systems
Verified
Journal Article
Closed access
PDF file
RMRasha Mostafa
Rasha Mostafa
AHAlaa A. Hefnawy
Alaa A. Hefnawy
MEMustafa M. Abd Elnaby
Mustafa M. Abd Elnaby
FEFathi E. Abd El‐Samie
Fathi E. Abd El‐Samie

30 September 2021 in International Journal of Communication Systems
Verified
Journal Article
Closed access
PDF file
RMRasha Mostafa
Rasha Mostafa
HGHanan S. Ghanem
Hanan S. Ghanem
HKHossam M. Kassem
Hossam M. Kassem
MEMaha Elsabrouty
Maha Elsabrouty
HHHesham F. A. Hamed
Hesham F. A. Hamed
Gerges MansourGerges Mansour
Gerges Mansour

24 November 2023 in EURASIP Journal on Advances in Signal Processing
Verified
Journal Article
Open access
PDF file
RMRasha Mostafa
Rasha Mostafa
MKMayada Khairy
Mayada Khairy
WEWalid El-Shafai
Walid El-Shafai
FBMC is a pivotal system in 5G, serving as a cornerstone for efficient use of available bandwidth while simultaneously meeting stringent requirements for high spectral efficiency. Notably, FBMC harnesses the power of multicarrier modulation (MC), a good alternative to orthogonal frequency division multiplexing (OFDM) technology that supports fourth-generation (4G) systems. The wireless communications field is full of challenges, the most important of which are channel estimation and interference cancellation, both of which deserve comprehensive study to increase the efficiency of data transmission. In this paper, our investigation takes a deliberate step towards the convergence of two prominent modulation models: OFDM and FBMC. We specifically contrast these modulation techniques with the intricate field of joint channel estimation and interference cancellation (JCEIC). In this research study, we take advantage of recurrent neural networks' (RNNs') efficiency as a vehicular channel to perform precise channel estimation and recovery of uncorrupted transmitted signals, thereby lowering the bit error rate (BER). Our channel estimation for a dual selective channel is based on the thoughtful placement of pilots scattered over the temporal and frequency dimensions, and is further improved by the interference cancellation method of low complexity that was selected. Our JCEIC proposal aims to integrate RNNs carefully, using the output sequences of JCEIC algorithms as useful inputs to this neural architecture. By clearly demonstrating a decrease in BER as compared to traditional approaches, it is evident that the performance of the novel approach is near to that of a perfect channel. Additionally, a comparison of the performance of FBMC and OFDM systems at various signal-to-noise ratios reveals a clear performance divide that favors the former in terms of system efficiency. The BER is restricted by FBMC to a commendable threshold of less than 0.1 at a modest 5 dB, continuing the higher trend started by its improved RNN-based channel estimate. The accuracy of channel estimation is clearly improved by this paradigm shift, and the computing complexity typical of 5G networks is also clearly reduced.

18 December 2023 in EURASIP Journal on Advances in Signal Processing
Verified
Journal Article
Open access
PDF file
RMRasha Mostafa
Rasha Mostafa
MKMayada Khairy
Mayada Khairy
WEWalid El‑Shafai
Walid El‑Shafai