Pejman Ghasemzadeh

Pejman Ghasemzadeh

Auto-generated profile, based on publicly available information

Pejman Ghasemzadeh

Department of Electrical and Computer Engineering, University of Nebraska-Lincoln, USA

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

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

Representative Publications

View All Publications
6 January 2022 in IEEE Internet of Things Journal
Verified
Journal Article
Closed access
PGPejman Ghasemzadeh
Pejman Ghasemzadeh
MHMichael Hempel
Michael Hempel
Hamid SharifHamid Sharif
Hamid Sharif
Cognitive radio (CR) has been introduced into the Internet of Things (IoT) domain as a potential solution to the critical spectrum resource shortage caused by a dramatic increase of IoT solutions within a limited operational spectrum. One technique of use within traditional CR processing for sensing communication activity across a spectrum of interest is automatic modulation classification (AMC). However, utilizing AMC within resource-limited cognitive radio-enabled IoT (CR-IoT) devices poses a significant challenge. In this article, we present a high-efficiency automatic modulation classifier architecture whose core is built upon the implementation of stacking quasirecurrent neural network (S-QRNN) layers acting as a feature extraction stage. This architecture utilizes the low-latency feature extraction capability of convolutional layers, and a minimalist recurrent pooling function that mimics recurrent-layer operations to aggregate the extracted features over time steps for higher classification accuracy. Additionally, implementing dense layers between two consecutive QRNN layers helps keep the network growth rate low. Therefore, the proposed S-QRNN classifier overall exhibits higher efficiency, fitting well within the limitations of CR-IoT devices. In order to demonstrate higher performance efficiency of our classifier, we conduct a comprehensive evaluation of our approach against the state-of-the-art AMC classifiers. Our analysis metrics of the classifier performance focuses on trainability, classification accuracy, execution latency, and resulting overall efficiency. Our results demonstrate that our proposed S-QRNN classifier exhibits higher trainability and, on average, a 75.83% higher efficiency, while it has, on average, a 26.54% lower execution latency. We then expand upon S-QRNN’s foundation by introducing gated recurrent units into our classifier to initially extract temporal features of the received constellations. The resulting GS-QRNN classifier demonstrates an efficiency increase by an average of 191% compared to the gated-enhanced recurrent-neural-network-based AMC classifier, with our GS-QRNN classifier average execution latency being 59.31% lower.

8 September 2020 in IEEE Transactions on Vehicular Technology
Verified
Journal Article
Closed access
PGPejman Ghasemzadeh
Pejman Ghasemzadeh
SBSubharthi Banerjee
Subharthi Banerjee
MHMichael Hempel
Michael Hempel
Hamid SharifHamid Sharif
Hamid Sharif
Automatic Modulation Classification (AMC) is an approach to identify an observed signal's most likely modulation scheme without any a priori knowledge of the intercepted signal. In this research, the authors present a new direction for both stages of feature-based (FB) approach. In the feature extraction stage, the authors design a new architecture that 1) removes the bias issue for the estimator of fourth-order cumulants, and 2) extracts polar-transformed information of the received IQ symbols, and finally 3) forms a unique dataset to be used in the labeling stage. Furthermore, the authors contribute to increasing the classification accuracy in low signal-to-noise ratio (SNR) conditions by employing the deep belief network (DBN) platform in addition to the spiking neural network (SNN) platform to overcome execution latency concerns associated with deep learning architectures. For this research, the authors first study each individual FB AMC classifier to derive their respective upper and lower performance bounds and then propose an adaptive framework that is built and developed with these findings. This framework aims to efficiently classify the modulation scheme by intelligently switching between these different FB classifiers to achieve an optimal balance between accuracy and execution latency for any observed channel conditions derived from the main receiver's equalizer. Subsequently, a performance analysis is conducted using the standard RadioML dataset to achieve a realistic evaluation. Numerical results indicate a notably higher classification accuracy, by 16.02% on average, when DBN is employed, whereas SNN requires significantly lower execution latency to label the modulation scheme when compared against two other modified FB classifiers that are built upon convolutional and recurrent neural networks, shown to be reduced by 34.31%.

30 January 2023 in IEEE Transactions on Wireless Communications
Verified
Journal Article
Closed access
PGPejman Ghasemzadeh
Pejman Ghasemzadeh
MHMichael Hempel
Michael Hempel
HWHonggang Wang
Honggang Wang
Hamid SharifHamid Sharif
Hamid Sharif
Automatic modulation identification (AMI) is a technique to detect the modulation type and order of a received signal, which has the potential to enhance cognitive radio capabilities for future generations of communication devices. However, AMI classifiers traditionally have exhibited low efficiency in low signal-to-noise ratio (SNR) environments. Hence, to address this problem we present our novel Gated Graph Convolutional Neural Network (GGCNN) classifier for feature-based AMI. This architecture includes a robust feature extraction stage to extract deep correlative patterns about the received symbols. Not only does this feature extraction stage use the temporal characteristics of the received symbols, but it also takes advantage of embedded signaling features from the received signal. In the proposed classifier, the received constellations are treated as a graph, allowing it to outperform state-of-the-art classifiers due to its strong performance in graph classification. This is observed clearly in the visualization of the extracted features, even for high-order modulation schemes. In this paper, we present our systematic research conducted for maximizing the efficiency obtainable by our classifier. Extensive simulation results demonstrate a significant accuracy improvement of 18.44 percentage points, and an efficiency increase by 60.78% for our GGCNN-AMI classifier compared to state-of-the-art classifiers in low-SNR environments.

1 February 2019
Verified
Conference Paper
Closed access
PGPejman Ghasemzadeh
Pejman Ghasemzadeh
SBSubharthi Banerjee
Subharthi Banerjee
MHMichael Hempel
Michael Hempel
HSHamid Sharif
Hamid Sharif
The process of automatic classification of a detected signal's employed modulation type has gained importance in recent years. The goal of such an approach is to maximize the achievable throughput for intelligent receiver designs in civilian applications as well as jamming malicious signals in military applications. Automatic Modulation Classification (AMC) increases in difficulty since there is no a-priori knowledge of transmitted signal properties, such as signal power, carrier frequency, or bandwidth, nor any associated link properties such as channel state information (CSI), noise characteristics, signal-to-noise ratio (SNR) or any offset in frequency and phase. The most complex, albeit also most realistic, scenarios for AMC are faced when considering Non-Gaussian noise with multipath fading in frequency selective and time-varying channels. Different methods have been proposed in the literature to estimate unknown signals and channel parameters for AMC. However, a key consideration in selecting among them is attaining low computational complexity in order for AMC to become a technique feasible for real-time applications. Predominantly, blind AMC and associated parameter estimation utilizes feature-based approaches, owing to their low-complexity calculations of statistical values. In this work, we have analyzed the accuracy of High-order Statistics-based (HoS) methods utilizing feature extraction approaches, Support Vector Machine classifiers, and estimation techniques to determine an optimized framework for different real-time applications.

1 December 2018
Verified
Conference Paper
Closed access
PGPejman Ghasemzadeh
Pejman Ghasemzadeh
SBSubharthi Banerjee
Subharthi Banerjee
MHMichael Hempel
Michael Hempel
HSHamid Sharif
Hamid Sharif
Automatic Modulation Classification (AMC) is a vital component for intelligent receivers. Due to the rapid growth of wireless communications, modulation classification has become a critical factor for efficient and high-performing receiver designs. Automatic modulation classification impacts critical civilian and military applications. In civilian applications, various forms of quaternary and 16-point modulations (QASK, QFSK, QPSK, 16-QAM, etc.) are deployed. For this group of modulations, the two dominant AMC approaches proposed in the literature are Likelihood- (LB) and Feature-based (FB) classification. Feature-based approaches focus on five main properties: Signal Spectral-based Features, Wavelet Transform-based Features, High-order Statistics-based (HoS) Features, Cyclostationary Analysis-based Features and Graph-based Cyclic-Spectrum Analysis Features, all of which are employed to extract initial information for the classifier. In this work, we analyze the performance of automatic modulation classification applied for civilian modulations with various features under practical scenarios.

15 April 2021 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Open access
PGPejman Ghasemzadeh
Pejman Ghasemzadeh
MHMichael Hempel
Michael Hempel
SBSubharthi Banerjee
Subharthi Banerjee
Hamid SharifHamid Sharif
Hamid Sharif
The procedure of classifying a detected signal’s modulation scheme with no a priori information is known as automatic modulation classification (AMC). AMC has presented numerous contributions toward civilian and military applications. However, in the scientific literature, designing automatic modulation classifiers has been limited to a single particular simulated environment in most cases of study. Hence, the performance of such classifiers lacks: 1) reliability in real-world environments and 2) multivariate environment analysis for AMC operation in dissimilar surroundings. These two reasons represent a significant obstacle to the real-world implementation of such classifiers. Therefore, in this research, we present our contribution to remove these obstacles by generating an emulated signal reference dataset named MIMOSigRef-SD. It includes a wide variety of signal streams modulated by different digital modulation schemes, such as M-QAM, MIL-STD-188-110 B/C standard-specific QAM, M-PSK, M-APSK, DVB-S2/S2X/SH standard-specific APSK, and M-PAM with different modulation orders, each with different multiple-input multiple-output (MIMO) system configurations in order to provide an extensive signal reference. Each signal is also exposed to the realistic impact of different channel models at 2450 MHz, specifically Vehicular-A/B, Pedestrian-A/B, and the Butler, through a channel emulator. This signal reference includes the randomly generated transmit signals, characteristics of the emulated environments, and the recorded received symbols. This enables the dataset to be highly applicable to applications beyond AMC as well. In general, this signal reference can help analyze, evaluate, and design any other radio frequency transceiver tasks under realistic environmental effects. We also validate the elements of the signal reference by comparing the emulation and simulation bit error rate of the designed communication system.