Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Muhammad Usman has a total of 28 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Corrections to “A Human-in-the-Loop Probabilistic CNN-Fuzzy Logic Framework for Accident Prediction in Vehicular Networks” [Jul 21 15496-15503]

    by Muhammad Usman, Anil Carie Chettupally, Bhaskar Marapelli, Hayat Dino Bedru, Kamanashish Biswas

    In the above article [1] , the surname Chettupally of the author Anil Carie Chettupally was inadvertently omitted from the authors list when the article was originally published.

  • Open Access

    Security Hardened and Privacy Preserved Vehicle-to-Everything (V2X) Communication

    by Azeem Irshad, Muhammad Shafiq, Shehzad Ashraf Chaudhry, Muhammad Usman

    Vehicle-to-everything (V2X) communications have recently gained concentration of researchers for both, academia as well as industry. In the V2X system, the information is communicated from vehicle sensors to other vehicles, infrastructure, pedestrians, and mobile network cloud through high-bandwidth reliable links [1–4]. The technology may greatly improve the driver’s awareness of imminent hazards, thereby reducing the severity of accidents, fatalities, or possible collisions with other vehicles. The V2X technology brings efficiency through creating warning alerts for drivers, imparting the information of alternative routes for avoiding possible traffic congestions and pinpointing available parking spaces. Such critical situations might become problematic if the security and privacy of V2X communication system is compromised [5–8]. Thus, V2X vehicles along with efficiency, reliability, and safety parameters require more secure and robust communication protocols to meet the upcoming security challenges. Moreover, the wireless nature of the system might become challenging in affording secure and ubiquitous connectivity to the V2X network [9–13]. This is crucial to create a fail-safe infrastructure of modern traffic scenario regarding smart cities since security and privacy issues are quite prevalent in our daily lives.


  • Guest Editorial: Computational Intelligence for Human-in-the-Loop Cyber Physical Systems

    by Alireza Jolfaei, Muhammad Usman, Manuel Roveri, Michael Sheng, Marimuthu Palaniswami, Krishna Kant

    The papers in this special section focus on human-in-the-loop cyber-physical systems. Recent advances in computational intelligence, real-time computing and control, have given momentum to Humanin-the-Loop Cyber Physical Systems (HitLCPS) to enable gamechanging communication and collaboration paradigms that operate in connection with humans’ natural behaviour patterns. Despite the ongoing advancement of computational intelligence techniques for analyzing the interactions between the cognitive and cyber domains, there are growing concerns regarding the security, privacy, and safety of humans when they interact with smart cyber physical environments (IoT ecosystems). The large-scale integration of heterogeneous IoT devices to manage and control a wide variety of sensors and settings will hugely increase the attack surface and the scope for misconfigurations. These could lead to unsafe or conflicting behaviour of various devices and subsystems, which in turn, can place the human in unsafe and hazardous situations, both mentally and physically. It is still unclear how to design optimized collaborative systems between people and machines in a scalable manner, how to design triggers for pro-active engagement and disengagement, and how to handle the consequences of implied actions. For example, when the system misbehaves as a result of erroneous data, it is important to have real-time rules that can guarantee a fail-safe state for the HitLCPS. The verification of operations in a large HitLCPS can be very complex due to the evolving nature of human-in-the-loop networks both in terms of physical aspects and the operational environment. Therefore, understanding the semantics of HitLCPS and the context of control behaviour is critical to detect or entirely avoid incorrect configurations and build a proactive resilience and a reactive defence against evolving threats.


  • A Cross-Layer Trust-Based Consensus Protocol for Peer-to-Peer Energy Trading Using Fuzzy Logic

    by Mohammad Jabed M. Chowdhury, Muhammad Usman, Sadek Ferdous, Niaz Chowdhury, Anam Ibna Harun, Umme Sumaya Jannat, Kamanashis Biswas

    Peer-to-Peer (P2P) energy trading platforms are being actively designed, tested, and operated by engineers, power distribution companies, and prosumers. The assurance of the accountability of the conduct of different stakeholders through a robust trust management mechanism is imperative in such platforms. The usage of blockchain, as an underlying technology, can ensure numerous properties, such as immutability, transparency, and traceable execution of transactions, in addition to ensuring trust establishment among different entities of the system. Few blockchain-based decentralized energy trading platforms have been designed in the literature to build trust about the platform and among prosumers. However, none of these proposals have considered human-in-the-loop in the trust establishment process. Moreover, these solutions have considered trust only at a particular layer of the blockchain, such as at the application or consensus layer. To bridge this gap, this article presents a novel cross-layer trust-based consensus protocol that considers human-in-the-loop and employs fuzzy logic to address the issue of vagueness of trust values by offering human interpretable trust level. The experimental results demonstrate the efficiency and effectiveness of our proposed protocol in comparison to established consensus mechanisms. The analysis also shows the protocol is immune against selfish mining, 51% and Sybil attacks.


  • Anomaly Detection in Automated Vehicles Using Multistage Attention-Based Convolutional Neural Network

    by Abdul Rehman Javed, Muhammad Usman, Saif Ur Rehman, Mohib Ullah Khan, Mohammad Sayad Haghighi

    Connected and Automated Vehicles (CAVs), owing to their characteristics such as seamless and real-time transfer of data, are imperative infrastructural advancements to realize the emerging smart world. The sensor-generated data are, however, vulnerable to anomalies caused due to faults, errors, and/or cyberattacks, which may cause accidents resulting in fatal casualties. To help in avoiding such situations by timely detecting anomalies, this study proposes an anomaly detection method that incorporates a combination of a multi-stage attention mechanism with a Long Short-Term Memory (LSTM)-based Convolutional Neural Network (CNN), namely, MSALSTM-CNN. The data streams, in the proposed method, are converted into vectors and then processed for anomaly detection. We also designed a method, namely, weight-adjusted fine-tuned ensemble: WAVED, which works on the principle of average predicted probability of multiple classifiers to detect anomalies in CAVs and benchmark the performance of the MSALSTM-CNN method. The MSALSTM-CNN method effectively enhances the anomaly detection rate in both low and high magnitude cases of anomalous instances in the dataset with the gain of up to 2.54% in F-score for detecting different single anomaly types. The method achieves the gain of up to 3.24% in F-score in the case of detecting mixed anomaly types. The experiment results show that the MSALSTM-CNN method achieves promising performance gain for both single and mixed multi-source anomaly types as compared to the state-of-the-art and benchmark methods.


  • Fuzzy-in-the-Loop-Driven Low-Cost and Secure Biometric User Access to Server

    by Azeem Irshad, Muhammad Usman, Shehzad Ashraf Chaudhry, Ali Kashif Bashir, Alireza Jolfaei, Gautam Srivastava

    Fuzzy systems can aid in diminishing uncertainty and noise from biometric security applications by providing an intelligent layer to the existing physical systems to make them reliable. In the absence of such fuzzy systems, a little random perturbation in captured human biometrics could disrupt the whole security system, which may even decline the authentication requests of legitimate entities during the protocol execution. In the literature, few fuzzy logic-based biometric authentication schemes have been presented; however, they lack significant security features including perfect forward secrecy (PFS), untraceability, and resistance to known attacks. This article, therefore, proposes a novel two-factor biometric authentication protocol enabling efficient and secure combination of physically unclonable functions, a physical object analogous to human fingerprint, with user biometrics by employing fuzzy extractor-based procedures in the loop. This combination enables the participants in the protocol to achieve PFS. The security of the proposed scheme is tested using the well-known real-or-random model. The performance analysis signifies the fact that the proposed scheme not only offers PFS, untraceability, and anonymity to the participants, but is also resilient to known attacks using light-weight symmetric operations, which makes it an imperative advancement in the category of intelligent and reliable security solutions.


  • A Human-in-the-Loop Probabilistic CNN-Fuzzy Logic Framework for Accident Prediction in Vehicular Networks

    by Muhammad Usman, Anil Carie, Bhaskar Marapelli, Hayat Dino Bedru, Kamanashish Biswas

    The vehicle accident prediction methods are designed to improve the vehicular safety and reduce the rescue response time in the case of an accident. The existing accident prediction methods, however, do not involve Human-in-the-Loop, i.e., do not consider the emotional state of a driver to predict the likelihood of an accident. We propose a Probabilistic Convolutional Neural Network (CNN)-Fuzzy Logic framework that involves Human-in-the-Loop and takes into account the multiple input streams of sensor generated data, i.e., human emotions and traffic data. The features extracted from the CNN model are fed to our designed probabilistic graphbased inference model to determine the accident probability. The probability is then mapped with accident severity through fuzzy membership functions for accident prediction. The experiment results show the promising performance of our proposed framework, i.e., 93.1% accuracy of face expressions, 76.2% accuracy of heartbeat, and 76.9% accuracy of traffic inputs and predicts the accident likelihood with 90% accuracy. The comparison, with related works, shows that the proposed framework can predict accidents with higher probabilities.


  • Closing the Loop in Feedback Driven Learning Environments Using Trust Decision Making and Utility Theory

    by Kanwal Mahmud, Muhammad Usman, Muddassar Azam Sindhu, Alireza Jolfaei, Gautam Srivastava

    Contemporary learning systems are an integration of learning resources with human interactions. To close the loop in feedback driven learning environments, the utility of learning objectives needs to be measured. To this end, a comprehensive trust evaluation model for decision making is required to utilize feedback ratings along with other key parameters such as previous course result percentage, active participation and reputation of learners. This paper proposes a novel utility theory based trust evaluation model, wherein the utility of a learning objective is computed in terms of trust applicable to big data-sets. The utility is computed by allowing users to weigh the course related attributes according to their preferences. The utility value facilitates learners to select trustworthy learning objectives and enables instructors to improve different aspects of learning objectives. In addition, a satisfaction index is proposed for the assessment of the usefulness of the computed utility value. The performance of the model is evaluated on a big data-set, which is collected from learners enrolled in different courses of a postgraduate degree program for the purposes of decision making. The results indicate that the proposed unique intelligent model is effective for dynamic and user-specified trust evaluations of learning objectives for the purposes of decision making.