Sobia Arshad
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Sobia Arshad has a total of 16 co-authors. Here are the top co-authors by number of shared publications:
- Muhammad Awais Azam
Technology Innovation Research Group, School of Information Technology, Whitecliffe, Wellington 6145, New Zealand
8 shared publications - Jonathan Loo
School of Computing and Engineering, University of West London, London, U.K.
5 shared publications - Mubashir Husain Rehmani
Department of Computer Science Munster Technological University (MTU) Cork Ireland
2 shared publications - Muhammad Faran Majeed
Department of Computer Science, Kohsar University Murree, Murree 47150, Pakistan
2 shared publications - Sanay Muhammad Umar Saeed
Department of Computer Engineering, University of Engineering and Technology, Taxila 47080, Pakistan
2 shared publications - Syed Hassan Ahmed
Department of Computer Science, California State University, Fullerton, CA, USA
2 shared publications
Latest Publications
- Open Access
- 3 November 2026
- Published by MDPI AG in Applied Sciences
- Vol. 16 (21) , 10897
Multi-Agent Agentic AI and Federated Learning for Attack Detection in Sustainable IoT
The Internet of Things (IoT) comprises a vast number of digitally networked devices. Machine learning (ML), deep learning (DL), and federated learning (FL), all subfields of artificial intelligence (AI), have recently been investigated as cybersecurity solutions for protecting these networked devices from intrusions and cyberattacks. However, existing ML and DL techniques for Sustainable IoT data carry significant privacy implications. Because of its inherent suitability for IoT settings, AI-based cybersecurity, and FL in particular, has become critical, yet traditional FL-based approaches themselves fall short in several respects. We need a framework that prioritizes privacy while enabling effective FL on edge-generated data. This research aims to address three persistent problems in traditional FL-based systems: class imbalance, unequal data distribution across devices, and computational complexity. To this end, we propose a novel Multi-Agentic AI framework in which each device decomposes its local workload across three cooperating: Agent 2 performs data preprocessing, including WGAN-based oversampling to correct class imbalance; Agent 1 performs data distribution, applying a vertical federated learning (VFL) partitioning scheme to address the unequal data distribution that typically affects FL-based IoT systems; and Agent 3 performs classification, training a deep autoencoder and an MLP on the resulting features while keeping the computational load on each device manageable. Experiments on the Modbus ICS network and ToN-IoT datasets show that our framework improves on the performance of comparable ML models while overcoming the limitations of existing FL-based techniques. Our VFL approach delivers consistent data distribution and higher attack detection accuracy, reaching 99.56% and 99.62% on the two datasets, respectively. Overall, this multi-agent agentic AI methodology offers a practical way to address the core challenges of FL and meaningfully improve AI effectiveness in privacy-sensitive sustainable IoT domains.
- 3 November 2026
- Published by MDPI AG in Applied Sciences
- Vol. 16 (21) , 10897
Open AccessMulti-Agent Agentic AI and Federated Learning for Attack Detection in Sustainable IoT
The Internet of Things (IoT) comprises a vast number of digitally networked devices. Machine learning (ML), deep learning (DL), and federated learning (FL), all subfields of artificial intelligence (AI), have recently been investigated as cybersecurity solutions for protecting these networked devices from intrusions and cyberattacks. However, existing ML and DL techniques for Sustainable IoT data carry significant privacy implications. Because of its inherent suitability for IoT settings, AI-based cybersecurity, and FL in particular, has become critical, yet traditional FL-based approaches themselves fall short in several respects. We need a framework that prioritizes privacy while enabling effective FL on edge-generated data. This research aims to address three persistent problems in traditional FL-based systems: class imbalance, unequal data distribution across devices, and computational complexity. To this end, we propose a novel Multi-Agentic AI framework in which each device decomposes its local workload across three cooperating: Agent 2 performs data preprocessing, including WGAN-based oversampling to correct class imbalance; Agent 1 performs data distribution, applying a vertical federated learning (VFL) partitioning scheme to address the unequal data distribution that typically affects FL-based IoT systems; and Agent 3 performs classification, training a deep autoencoder and an MLP on the resulting features while keeping the computational load on each device manageable. Experiments on the Modbus ICS network and ToN-IoT datasets show that our framework improves on the performance of comparable ML models while overcoming the limitations of existing FL-based techniques. Our VFL approach delivers consistent data distribution and higher attack detection accuracy, reaching 99.56% and 99.62% on the two datasets, respectively. Overall, this multi-agent agentic AI methodology offers a practical way to address the core challenges of FL and meaningfully improve AI effectiveness in privacy-sensitive sustainable IoT domains.
- Open Access
ZDBERTa: Advancing Zero-Day Cyberattack Detection in Internet of Vehicle with Zero-Shot Learning
The Internet of Vehicles (IoV) is becoming increasingly vulnerable to zero-day (ZD) cyberattacks, which often bypass conventional intrusion detection systems. To mitigate this challenge, this study proposes Zero-Day Bidirectional Encoder Representations from Transformers approach (ZDBERTa), a zero-shot learning (ZSL)-based framework for ZD attack detection, evaluated on the CICIoV2024 dataset. Unlike conventional AI models, ZSL enables the classification of attack types not previously encountered during the training phase. Two dataset variants are formed: Variant 1, created through synthetic traffic generation using a mixture of pattern-based, crossover, and mutation techniques, and Variant 2, augmented with a Generative Adversarial Network (GAN). To replicate realistic zero-day conditions, denial-of-service (DoS) attacks were omitted during training and introduced only at testing. The proposed ZDBERTa incorporates a Byte-Pair Encoding (BPE) tokenizer, a multi-layer transformer encoder, and a classification head for prediction, enabling the model to capture semantic patterns and identify previously unseen threats. The experimental results demonstrate that ZDBERTa achieves 86.677% accuracy on Variant 1, highlighting the complexity of zero-day detection, while performance significantly improves to 99.315% on Variant 2, underscoring the effectiveness of GAN-based augmentation. To the best of our knowledge, this is the first research to explore ZD detection within CICIoV2024, contributing a novel direction toward resilient IoV cybersecurity.
Open AccessZDBERTa: Advancing Zero-Day Cyberattack Detection in Internet of Vehicle with Zero-Shot Learning
The Internet of Vehicles (IoV) is becoming increasingly vulnerable to zero-day (ZD) cyberattacks, which often bypass conventional intrusion detection systems. To mitigate this challenge, this study proposes Zero-Day Bidirectional Encoder Representations from Transformers approach (ZDBERTa), a zero-shot learning (ZSL)-based framework for ZD attack detection, evaluated on the CICIoV2024 dataset. Unlike conventional AI models, ZSL enables the classification of attack types not previously encountered during the training phase. Two dataset variants are formed: Variant 1, created through synthetic traffic generation using a mixture of pattern-based, crossover, and mutation techniques, and Variant 2, augmented with a Generative Adversarial Network (GAN). To replicate realistic zero-day conditions, denial-of-service (DoS) attacks were omitted during training and introduced only at testing. The proposed ZDBERTa incorporates a Byte-Pair Encoding (BPE) tokenizer, a multi-layer transformer encoder, and a classification head for prediction, enabling the model to capture semantic patterns and identify previously unseen threats. The experimental results demonstrate that ZDBERTa achieves 86.677% accuracy on Variant 1, highlighting the complexity of zero-day detection, while performance significantly improves to 99.315% on Variant 2, underscoring the effectiveness of GAN-based augmentation. To the best of our knowledge, this is the first research to explore ZD detection within CICIoV2024, contributing a novel direction toward resilient IoV cybersecurity.
- Open Access
WGAN-DL-IDS: An Efficient Framework for Intrusion Detection System Using WGAN, Random Forest, and Deep Learning Approaches
The rise in cyber security issues has caused significant harm to tech world and thus society in recent years. Intrusion detection systems (IDSs) are crucial for the detection and the mitigation of the increasing risk of cyber attacks. False and disregarded alarms are a common problem for traditional IDSs in high-bandwidth and large-scale network systems. While applying learning techniques to intrusion detection, researchers are facing challenges mainly due to the imbalanced training sets and the high dimensionality of datasets, resulting from the scarcity of attack data and longer training periods, respectively. Thus, this leads to reduced efficiency. In this research study, we propose a strategy for dealing with the problems of imbalanced datasets and high dimensionality in IDSs. In our efficient and novel framework, we integrate an oversampling strategy that uses Generative Adversarial Networks (GANs) to overcome the difficulties introduced by imbalanced datasets, and we use the Random Forest (RF) importance algorithm to select a subset of features that best represent the dataset to reduce the dimensionality of a training dataset. Then, we use three deep learning techniques, Multi-Layer Perceptron (MLP), Convolutional Neural Network (CNN), and Long Short-Term Memory (LSTM), to classify the attacks. We implement and evaluate this proposed framework on the CICIDS2017 dataset. Experimental results show that our proposed framework outperforms state-of-the-art approaches, vastly improving DL model detection accuracy by 98% using CNN.
Open AccessWGAN-DL-IDS: An Efficient Framework for Intrusion Detection System Using WGAN, Random Forest, and Deep Learning Approaches
The rise in cyber security issues has caused significant harm to tech world and thus society in recent years. Intrusion detection systems (IDSs) are crucial for the detection and the mitigation of the increasing risk of cyber attacks. False and disregarded alarms are a common problem for traditional IDSs in high-bandwidth and large-scale network systems. While applying learning techniques to intrusion detection, researchers are facing challenges mainly due to the imbalanced training sets and the high dimensionality of datasets, resulting from the scarcity of attack data and longer training periods, respectively. Thus, this leads to reduced efficiency. In this research study, we propose a strategy for dealing with the problems of imbalanced datasets and high dimensionality in IDSs. In our efficient and novel framework, we integrate an oversampling strategy that uses Generative Adversarial Networks (GANs) to overcome the difficulties introduced by imbalanced datasets, and we use the Random Forest (RF) importance algorithm to select a subset of features that best represent the dataset to reduce the dimensionality of a training dataset. Then, we use three deep learning techniques, Multi-Layer Perceptron (MLP), Convolutional Neural Network (CNN), and Long Short-Term Memory (LSTM), to classify the attacks. We implement and evaluate this proposed framework on the CICIDS2017 dataset. Experimental results show that our proposed framework outperforms state-of-the-art approaches, vastly improving DL model detection accuracy by 98% using CNN.
- Open Access
- 28 November 2023
- Published by MDPI AG in Sustainability
- Vol. 15 (23) , 16361
Hierarchical, Attribute and Hash-Based Naming and Forwarding Aided Smart Campus of Things
In order to provide universal ability to access information and communication among Internet-connected devices, the Sustainable Internet of Things (IoT) is on a mission to bring all objects or devices under one roof. Future Internet architecture, especially Information-Centric Networking (ICN), can easily handle the connectivity offered and information created by the massive amount of devices to make it as sustainable IoT applications. Named Data Networking (NDN), one of the several future Internet designs that employ ICN as its foundation, shows promise. NDN integration with IoT-based applications gives solutions to numerous problems. However, this fusion makes accessing the IoT content easier, provided that an effective naming scheme is created to execute this operation. In this work, we build an innovative NDN-based naming scheme (NDN–NS) and put it into practise for consumer, producer, and content routers using our own secure forwarding schemes (NDN–NFS). Due to its scalability, heterogeneity, and security needs, IoT-based Smart Campus (IoT-SC) scenarios are taken into consideration for design and evaluation. We give a complete activity list based on NDN–NS that is split into two communication models (PusH Type Communication (PHTC) and PulL Type Communication (PLTC)) that can be applied to any IoT application. In terms of interest satisfaction rate (ISR), delay, and number of transmissions, we compare the NDN–NFS to legacy NDN. The outcomes demonstrate that NDN–NFS outperforms classic NDN in terms of performance and efficiency.
- 28 November 2023
- Published by MDPI AG in Sustainability
- Vol. 15 (23) , 16361
Open AccessHierarchical, Attribute and Hash-Based Naming and Forwarding Aided Smart Campus of Things
In order to provide universal ability to access information and communication among Internet-connected devices, the Sustainable Internet of Things (IoT) is on a mission to bring all objects or devices under one roof. Future Internet architecture, especially Information-Centric Networking (ICN), can easily handle the connectivity offered and information created by the massive amount of devices to make it as sustainable IoT applications. Named Data Networking (NDN), one of the several future Internet designs that employ ICN as its foundation, shows promise. NDN integration with IoT-based applications gives solutions to numerous problems. However, this fusion makes accessing the IoT content easier, provided that an effective naming scheme is created to execute this operation. In this work, we build an innovative NDN-based naming scheme (NDN–NS) and put it into practise for consumer, producer, and content routers using our own secure forwarding schemes (NDN–NFS). Due to its scalability, heterogeneity, and security needs, IoT-based Smart Campus (IoT-SC) scenarios are taken into consideration for design and evaluation. We give a complete activity list based on NDN–NS that is split into two communication models (PusH Type Communication (PHTC) and PulL Type Communication (PLTC)) that can be applied to any IoT application. In terms of interest satisfaction rate (ISR), delay, and number of transmissions, we compare the NDN–NFS to legacy NDN. The outcomes demonstrate that NDN–NFS outperforms classic NDN in terms of performance and efficiency.
- Open Access
Recent Advances in Information-Centric Networking based Internet of Things (ICN-IoT)
Information-Centric Networking (ICN) is being realized as a promising approach to accomplish the shortcomings of current IP-address based networking. ICN models are based on naming the content to get rid of address-space scarcity, accessing the content via name-based-routing, caching the content at intermediate nodes to provide reliable, efficient data delivery and self-certifying contents to ensure better security. Obvious benefits of ICN in terms of fast and efficient data delivery and improved reliability raises ICN as highly promising networking model for Internet of Things (IoTs) like environments. IoT aims to connect anyone and/or anything at any time by any path on any place. From last decade, IoTs attracts both industry and research communities. IoTs is an emerging research field and still in its infancy. Thus, this paper presents the potential of ICN for IoTs by providing state-of-the-art literature survey. We discuss briefly the feasibility of ICN features and their models (and architectures) in the context of IoT. Subsequently, we present a comprehensive survey on ICN based caching, naming, security and mobility approaches for IoTs with appropriate classification. Furthermore, we present operating systems (OS) and simulation tools for ICNIoT. Finally, we provide important research challenges and issues faced by ICN for IoTs.
Open AccessRecent Advances in Information-Centric Networking based Internet of Things (ICN-IoT)
Information-Centric Networking (ICN) is being realized as a promising approach to accomplish the shortcomings of current IP-address based networking. ICN models are based on naming the content to get rid of address-space scarcity, accessing the content via name-based-routing, caching the content at intermediate nodes to provide reliable, efficient data delivery and self-certifying contents to ensure better security. Obvious benefits of ICN in terms of fast and efficient data delivery and improved reliability raises ICN as highly promising networking model for Internet of Things (IoTs) like environments. IoT aims to connect anyone and/or anything at any time by any path on any place. From last decade, IoTs attracts both industry and research communities. IoTs is an emerging research field and still in its infancy. Thus, this paper presents the potential of ICN for IoTs by providing state-of-the-art literature survey. We discuss briefly the feasibility of ICN features and their models (and architectures) in the context of IoT. Subsequently, we present a comprehensive survey on ICN based caching, naming, security and mobility approaches for IoTs with appropriate classification. Furthermore, we present operating systems (OS) and simulation tools for ICNIoT. Finally, we provide important research challenges and issues faced by ICN for IoTs.
- Top Cited Papers
- 1 October 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 6 (2) , 2128-2158
Recent Advances in Information-Centric Networking-Based Internet of Things (ICN-IoT)
Information-Centric Networking (ICN) is being realized as a promising approach to accomplish the shortcomings of current IP-address based networking. ICN models are based on naming the content to get rid of address-space scarcity, accessing the content via name-based-routing, caching the content at intermediate nodes to provide reliable, efficient data delivery and self-certifying contents to ensure better security. Obvious benefits of ICN in terms of fast and efficient data delivery and improved reliability raises ICN as highly promising networking model for Internet of Things (IoTs) like environments. IoT aims to connect anyone and/or anything at any time by any path on any place. From last decade, IoTs attracts both industry and research communities. IoTs is an emerging research field and still in its infancy. Thus, this paper presents the potential of ICN for IoTs by providing state-of-the-art literature survey. We discuss briefly the feasibility of ICN features and their models (and architectures) in the context of IoT. Subsequently, we present a comprehensive survey on ICN based caching, naming, security and mobility approaches for IoTs with appropriate classification. Furthermore, we present operating systems (OS) and simulation tools for ICN-IoT. Finally, we provide important research challenges and issues faced by ICN for IoTs.
- 1 October 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 6 (2) , 2128-2158
Top Cited PapersRecent Advances in Information-Centric Networking-Based Internet of Things (ICN-IoT)
Information-Centric Networking (ICN) is being realized as a promising approach to accomplish the shortcomings of current IP-address based networking. ICN models are based on naming the content to get rid of address-space scarcity, accessing the content via name-based-routing, caching the content at intermediate nodes to provide reliable, efficient data delivery and self-certifying contents to ensure better security. Obvious benefits of ICN in terms of fast and efficient data delivery and improved reliability raises ICN as highly promising networking model for Internet of Things (IoTs) like environments. IoT aims to connect anyone and/or anything at any time by any path on any place. From last decade, IoTs attracts both industry and research communities. IoTs is an emerging research field and still in its infancy. Thus, this paper presents the potential of ICN for IoTs by providing state-of-the-art literature survey. We discuss briefly the feasibility of ICN features and their models (and architectures) in the context of IoT. Subsequently, we present a comprehensive survey on ICN based caching, naming, security and mobility approaches for IoTs with appropriate classification. Furthermore, we present operating systems (OS) and simulation tools for ICN-IoT. Finally, we provide important research challenges and issues faced by ICN for IoTs.
- Open Access
Disaster Management System Aided by Named Data Network of Things: Architecture, Design, and Analysis
Disasters are the uncertain calamities which within no time can change the situation quite drastically. They not only affect the system’s infrastructure but can also put an adverse effect on human life. A large chunk of the IP-based Internet of Things (IoT) schemes tackle disasters such as fire, earthquake, and flood. Moreover, recently proposed Named Data Networking (NDN) architecture exhibited promising results for IoT as compare to IP-based approaches. Therefore to tackle disaster management system (DMS), it is needed to explore it through NDN architecture and this is the main motivation behind this work. In this research, a NDN based IoT-DMS (fire disaster) architecture is proposed, named as NDN-DISCA. In NDN-DISCA, NDN producer pushes emergency content towards nearby consumers. To provide push support, Beacon Alert Message (BAM) is created using fixed sequence number. NDN-DISCA is simulated in ndnSIM considering the disaster scenario of IoT-based smart campus (SC). From results, it is found that NDN-DISCA exhibits minimal delay and improved throughput when compared to the legacy NDN and existing PUSH schemes.
Open AccessDisaster Management System Aided by Named Data Network of Things: Architecture, Design, and Analysis
Disasters are the uncertain calamities which within no time can change the situation quite drastically. They not only affect the system’s infrastructure but can also put an adverse effect on human life. A large chunk of the IP-based Internet of Things (IoT) schemes tackle disasters such as fire, earthquake, and flood. Moreover, recently proposed Named Data Networking (NDN) architecture exhibited promising results for IoT as compare to IP-based approaches. Therefore to tackle disaster management system (DMS), it is needed to explore it through NDN architecture and this is the main motivation behind this work. In this research, a NDN based IoT-DMS (fire disaster) architecture is proposed, named as NDN-DISCA. In NDN-DISCA, NDN producer pushes emergency content towards nearby consumers. To provide push support, Beacon Alert Message (BAM) is created using fixed sequence number. NDN-DISCA is simulated in ndnSIM considering the disaster scenario of IoT-based smart campus (SC). From results, it is found that NDN-DISCA exhibits minimal delay and improved throughput when compared to the legacy NDN and existing PUSH schemes.
- 11 January 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 5 (2) , 1070-1080
Hierarchical and Flat-Based Hybrid Naming Scheme in Content-Centric Networks of Things
Information-Centric Networking (ICN) approaches have been considered as an alternative approach to TCP/IP. Contrary to the traditional IP, the ICN treats content as a first-class citizen of the entire network, where names are given through different naming schemes to contents and are used during the retrieval. Among ICN approaches, Content Centric Networking (CCN) is one of the key protocols being explored for Internet of Things (IoT), names the contents using hierarchical naming. Moreover, CCN follows pull-based strategy and exhibits the communication loop problem because of its broadcasting mode. However, IoT requires both pull and push modes of communication with scalable and secured content names in terms of integrity. In this paper, we propose a hybrid naming scheme that names contents using hierarchical and flat components to support both push and pull communication and to provide both scalability and security, respectively. We consider an IoT based Smart Campus (SC) scenario and introduce two transmission modes namely (1) unicast mode and (2) broadcast mode to address loop problem associated with CCN. Simulation results demonstrate that proposed scheme significantly improves the rate of interest transmissions, number of covered hops, name aggregation, and reliability along with addressing the loop problem.
- 11 January 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 5 (2) , 1070-1080
Hierarchical and Flat-Based Hybrid Naming Scheme in Content-Centric Networks of Things
Information-Centric Networking (ICN) approaches have been considered as an alternative approach to TCP/IP. Contrary to the traditional IP, the ICN treats content as a first-class citizen of the entire network, where names are given through different naming schemes to contents and are used during the retrieval. Among ICN approaches, Content Centric Networking (CCN) is one of the key protocols being explored for Internet of Things (IoT), names the contents using hierarchical naming. Moreover, CCN follows pull-based strategy and exhibits the communication loop problem because of its broadcasting mode. However, IoT requires both pull and push modes of communication with scalable and secured content names in terms of integrity. In this paper, we propose a hybrid naming scheme that names contents using hierarchical and flat components to support both push and pull communication and to provide both scalability and security, respectively. We consider an IoT based Smart Campus (SC) scenario and introduce two transmission modes namely (1) unicast mode and (2) broadcast mode to address loop problem associated with CCN. Simulation results demonstrate that proposed scheme significantly improves the rate of interest transmissions, number of covered hops, name aggregation, and reliability along with addressing the loop problem.