Mohammad Reza Ahmadzadeh
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Mohammad Reza Ahmadzadeh has a total of 26 co-authors. Here are the top co-authors by number of shared publications:
Elham Nazari
2 shared publications- Mehdi Mahdavi
Department of Electrical and Computer Engineering, Isfahan University of Technology, Isfahan, Iran
2 shared publications - Melike Erol-Kantarci
School of Electrical Engineering and Computer Science, University of Ottawa, Ottawa, ON, Canada
2 shared publications - Parisa Parhizgar
Department of Electrical and Computer Engineering, Isfahan University of Technology, Isfahan, Iran
2 shared publications - Saeed Sadri
Isfahan University of Technology
2 shared publications Ali Eslamian
Department of Electrical and Computer Engineering Isfahan University of Technology Isfahan, Iran
1 shared publication
Latest Publications
- 11 September 2026
- Published by Elsevier in Information Sciences
- Vol. 761
FedALPA: Federated learning via adaptive layer-wise partial aggregation
- 11 September 2026
- Published by Elsevier in Information Sciences
- Vol. 761
FedALPA: Federated learning via adaptive layer-wise partial aggregation
- Open Access
- 30 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of Vehicular Technology
- Vol. 6, 2828-2841
Federated Reinforcement Learning for Energy-Efficient D2D-IoT Networks With AoI Awareness
This paper proposes a federated reinforcement learning (FRL) framework for optimizing energy efficiency (EE) and Age of Information (AoI) in device-to-device (D2D) and Internet of Things (IoT) networks. The model leverages simultaneous wireless information and power transfer (SWIPT) with heterogeneous energy harvesting mechanisms-time switching (TS) for D2D users and power splitting (PS) for IoT devices. The objective is to maximize EE while satisfying constraints on data rate, AoI, power transmission, spectrum sharing, and time allocation. The resulting non-convex mixed-integer nonlinear programming problem is addressed using an FRL approach, where a software-defined network controller coordinates distributed agents to optimize resource allocation while preserving data privacy. Simulations demonstrate that the proposed framework achieves up to 25% higher EE and maintains AoI below critical thresholds compared to baseline methods, offering a scalable solution for energy-constrained, time-sensitive communication systems.
- 30 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of Vehicular Technology
- Vol. 6, 2828-2841
Open AccessFederated Reinforcement Learning for Energy-Efficient D2D-IoT Networks With AoI Awareness
This paper proposes a federated reinforcement learning (FRL) framework for optimizing energy efficiency (EE) and Age of Information (AoI) in device-to-device (D2D) and Internet of Things (IoT) networks. The model leverages simultaneous wireless information and power transfer (SWIPT) with heterogeneous energy harvesting mechanisms-time switching (TS) for D2D users and power splitting (PS) for IoT devices. The objective is to maximize EE while satisfying constraints on data rate, AoI, power transmission, spectrum sharing, and time allocation. The resulting non-convex mixed-integer nonlinear programming problem is addressed using an FRL approach, where a software-defined network controller coordinates distributed agents to optimize resource allocation while preserving data privacy. Simulations demonstrate that the proposed framework achieves up to 25% higher EE and maintains AoI below critical thresholds compared to baseline methods, offering a scalable solution for energy-constrained, time-sensitive communication systems.
- Open Access
- 20 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of Vehicular Technology
- Vol. 6, 52-67
Enhancing Information Freshness and Energy Efficiency in D2D Networks Through DRL-Based Scheduling and Resource Management
This paper investigates resource management in device-to-device (D2D) networks coexisting with cellular user equipment (CUEs). We introduce a novel model for joint scheduling and resource management in D2D networks, taking into account environmental constraints. To preserve information freshness, measured by minimizing the average age of information (AoI), and to effectively utilize energy harvesting (EH) technology to satisfy the network's energy needs, we formulate an online optimization problem. This formulation considers factors such as the quality of service (QoS) for both CUEs and D2Ds, available power, information freshness, and environmental sensing requirements. Due to the mixed-integer nonlinear nature and online characteristics of the problem, we propose a deep reinforcement learning (DRL) approach to solve it effectively. Numerical results show that the proposed joint scheduling and resource management strategy, utilizing the soft actor-critic (SAC) algorithm, reduces the average AoI by 20% compared to other baseline methods.
- 20 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Open Journal of Vehicular Technology
- Vol. 6, 52-67
Open AccessEnhancing Information Freshness and Energy Efficiency in D2D Networks Through DRL-Based Scheduling and Resource Management
This paper investigates resource management in device-to-device (D2D) networks coexisting with cellular user equipment (CUEs). We introduce a novel model for joint scheduling and resource management in D2D networks, taking into account environmental constraints. To preserve information freshness, measured by minimizing the average age of information (AoI), and to effectively utilize energy harvesting (EH) technology to satisfy the network's energy needs, we formulate an online optimization problem. This formulation considers factors such as the quality of service (QoS) for both CUEs and D2Ds, available power, information freshness, and environmental sensing requirements. Due to the mixed-integer nonlinear nature and online characteristics of the problem, we propose a deep reinforcement learning (DRL) approach to solve it effectively. Numerical results show that the proposed joint scheduling and resource management strategy, utilizing the soft actor-critic (SAC) algorithm, reduces the average AoI by 20% compared to other baseline methods.
- Open Access
Det-SLAM: A semantic visual SLAM for highly dynamic scenes using Detectron2
According to experts, Simultaneous Localization and Mapping (SLAM) is an intrinsic part of autonomous robotic systems. Several SLAM systems with impressive performance have been invented and used during the last several decades. However, there are still unresolved issues, such as how to deal with moving objects in dynamic situations. Classic SLAM systems depend on the assumption of a static environment, which becomes unworkable in highly dynamic situations. Several methods have been presented to tackle this issue in recent years, but each has its limitations. This research combines the visual SLAM systems ORB-SLAM3 and Detectron2 to present the Det-SLAM system, which employs depth information and semantic segmentation to identify and eradicate dynamic spots to accomplish semantic SLAM for dynamic situations. Evaluation of public TUM datasets indicates that Det-SLAM is more resilient than previous dynamic SLAM systems and can lower the estimated error of camera posture in dynamic indoor scenarios.
Open AccessDet-SLAM: A semantic visual SLAM for highly dynamic scenes using Detectron2
According to experts, Simultaneous Localization and Mapping (SLAM) is an intrinsic part of autonomous robotic systems. Several SLAM systems with impressive performance have been invented and used during the last several decades. However, there are still unresolved issues, such as how to deal with moving objects in dynamic situations. Classic SLAM systems depend on the assumption of a static environment, which becomes unworkable in highly dynamic situations. Several methods have been presented to tackle this issue in recent years, but each has its limitations. This research combines the visual SLAM systems ORB-SLAM3 and Detectron2 to present the Det-SLAM system, which employs depth information and semantic segmentation to identify and eradicate dynamic spots to accomplish semantic SLAM for dynamic situations. Evaluation of public TUM datasets indicates that Det-SLAM is more resilient than previous dynamic SLAM systems and can lower the estimated error of camera posture in dynamic indoor scenarios.
- Open Access
- 26 December 2020
- Published by Institution of Engineering and Technology (IET) in IET Image Processing
- Vol. 15 (7) , 1439-1458
A level‐set method for inhomogeneous image segmentation with application to breast thermography images
Various level‐set methods have been suggested for segmenting images with intensity inhomogeneity as local region‐based models. The challenge in these methods is segmenting the inhomogeneous images with smooth edges. These methods cannot properly segment regions with smooth edges in inhomogeneous images. This paper presents a new local region‐based active contour model called local self‐weighted active contour model. In the proposed method, a novel different weighting technique is applied. In this model, the weight of each neighbour pixel in the energy function is set by a function of its intensity and not its geometrical distance regarding the central pixel as previous methods. Considering this, the presented approach can segment regions with smooth edges in the presence of inhomogeneity as breast thermography images. The experimental results of applying the model on heterogeneous images containing synthetic images and medical images, especially breast thermography images, are compared with well‐known local level‐set methods which show the perfect capability of the model. The segmentation results were evaluated using the F‐score, accuracy, precision and recall criteria. The results show values of 0.8, 0.62, 0.73 and 0.82 for the average accuracy, F‐score, precision and recall criteria on the segmentation of breast thermography images, respectively.
- 26 December 2020
- Published by Institution of Engineering and Technology (IET) in IET Image Processing
- Vol. 15 (7) , 1439-1458
Open AccessA level‐set method for inhomogeneous image segmentation with application to breast thermography images
Various level‐set methods have been suggested for segmenting images with intensity inhomogeneity as local region‐based models. The challenge in these methods is segmenting the inhomogeneous images with smooth edges. These methods cannot properly segment regions with smooth edges in inhomogeneous images. This paper presents a new local region‐based active contour model called local self‐weighted active contour model. In the proposed method, a novel different weighting technique is applied. In this model, the weight of each neighbour pixel in the energy function is set by a function of its intensity and not its geometrical distance regarding the central pixel as previous methods. Considering this, the presented approach can segment regions with smooth edges in the presence of inhomogeneity as breast thermography images. The experimental results of applying the model on heterogeneous images containing synthetic images and medical images, especially breast thermography images, are compared with well‐known local level‐set methods which show the perfect capability of the model. The segmentation results were evaluated using the F‐score, accuracy, precision and recall criteria. The results show values of 0.8, 0.62, 0.73 and 0.82 for the average accuracy, F‐score, precision and recall criteria on the segmentation of breast thermography images, respectively.
- 1 November 2020
- Published by Springer Nature in Signal, Image and Video Processing
- Vol. 14 (8) , 1583-1590
Quantitative analysis of SNR in bilinear time frequency domain
- 1 November 2020
- Published by Springer Nature in Signal, Image and Video Processing
- Vol. 14 (8) , 1583-1590
Quantitative analysis of SNR in bilinear time frequency domain
- 17 December 2019
- Published by Springer Nature in Intelligent Service Robotics
- Vol. 13 (2) , 235-250
Qualitative vision-based navigation based on sloped funnel lane concept
- 17 December 2019
- Published by Springer Nature in Intelligent Service Robotics
- Vol. 13 (2) , 235-250
Qualitative vision-based navigation based on sloped funnel lane concept
- 1 October 2018
- Published by Elsevier in Signal Processing
- Vol. 151, 160-171
WIRIF: Wave interference-based rotation invariant feature for texture description
- 1 October 2018
- Published by Elsevier in Signal Processing
- Vol. 151, 160-171
WIRIF: Wave interference-based rotation invariant feature for texture description
- 2 September 2016
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 76 (14) , 15251-15278
A new efficient feature-combination-based method for dynamic texture modeling and classification using semi-random starting parameter dynamic Bayesian networks
- 2 September 2016
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 76 (14) , 15251-15278
A new efficient feature-combination-based method for dynamic texture modeling and classification using semi-random starting parameter dynamic Bayesian networks