Abdul Samad

Alhamdulilah! our paper titled "Enhanced Human Activity Recognition using Inertial Sensor Data from Smart Wearables: A Neural Network Approach with Residual Connections" got accepted at ICECT2024. a conference of IEEE coming up in Rawalpindi. Credit goes to the team Shaida Muhammad Ataul Aziz Ikram Saad Wazir Kiran Hamza. 👍👍👍 #IEEE #Sensors #AI #PakisaniResearcher #NUST #SEECS #NUCES #kaist #meaurementTechnology #sensorTec #sensorstechnology #mems #DeepLearning #IoT #IoHT

27

2 Comments

Completing AI Python for Beginners: Automating Tasks with Python! Deep Learning

5

2 Comments

Alhamdolillah! I'm gald to share that my 6th Q1 paper of 2024, "Predicting the actual use of Artificial intelligence features of Apple Vision using PLS-SEM", has been accepted for publication in the prestigious IEEE Access journal. Grateful for the opportunity to publish in a top-tier journal with an Impact Factor of 3.4, indexed in the Web of Science (WoS), and ranked Q1 in its category. #IEEEAccess #Research #ArtificialIntelligence #AppleVision #PLSSEM #AcademicSuccess #Gratitude"

10

1 Comment

Alhamdulillah, I am excited to announce the publication of my research paper, titled "MACPE: Mobility Aware Content Provisioning in Edge-Based Content-Centric Internet of Vehicles," in the IEEE Access Journal. This work contributes to the fields of the Internet of Vehicles (IoV) and Information-Centric Networking (ICN). https://lnkd.in/dE9gQVPx

18

6 Comments

Excited to share that I recently took part in IEEE RAS Hackathon! It was an amazing experience working on challenging problems and coming up with new ideas. Learned many new problem Solving techniques. Also Identified my mistakes and will try not to do it again. Now I'm excited for more opportunities like this! #Hackathon #Innovation #Tech #Coding #CompetetiveProgramming #TechNinja #RASit #Comsats

9

2 Comments

"I'm super excited to share that I've started an online course on Transmission Planning using PSSE Software! I've always wanted to get hands-on experience with real-time projects, and this course is making it happen. Huge shoutout to Engr. Hamd- ullah and Power system center - his teaching style is amazing! He breaks down complex concepts into bite-sized pieces, making it easy to understand and apply. I'm so grateful for his guidance and support. I'm passionate about electrical engineering, and I believe that continuous learning is key to success. This course is a big step towards achieving my goals, and I can't wait to see where it takes me. If you're interested in learning more about transmission planning or PSSE Software, let's connect! I'd love to share my experiences and learn from yours. #TransmissionPlanning #PSSESoftware #OnlineLearning Power system center #Electrical Engineering #LearningCommunity #GrowthMindset" #Hamd- ullah

7

2 Comments

𝐃𝐞𝐚𝐫 𝐂𝐨𝐧𝐧𝐞𝐜𝐭𝐢𝐨𝐧𝐬 I'm delighted to introduce two accomplished researchers from MERL – Muhammad Shameel Ansari and Ameen Raza Khan . Both have consistently impressed with their technical expertise, dedication, and valuable contributions to our team, particularly in the realm of hardware design and verification. 𝐌𝐮𝐡𝐚𝐦𝐦𝐚𝐝 𝐒𝐡𝐚𝐦𝐞𝐞𝐥 𝐀𝐧𝐬𝐚𝐫𝐢 brings a powerful combination of hardware design and verification skills. His achievements include: Verification Excellence: Played a crucial role in developing a robust verification environment for the AI-FPU, ensuring its accuracy and reliability. Enhanced Usability: Created an intuitive and user-friendly graphical interface for the AI-FPU, significantly improving its accessibility. Peripheral Expertise: Successfully designed and verified the GPIO and PWM peripheral IPs, contributing to the overall functionality and performance of MERL projects. FPGA Proficiency: Utilized his expertise in FPGA emulation to provide a reliable platform for testing and validation of these components. Backend Development: Contributed to the backend development of the Liberty Analyzer, leveraging Python skills to enhance the tool's functionality and efficiency. 𝐀𝐦𝐞𝐞𝐧 𝐑𝐚𝐳𝐚 𝐊𝐡𝐚𝐧 demonstrates exceptional proficiency in PY-UVM methodologies and hardware design. His expertise is evident in: PY-UVM Frameworks: Designed Python-based UVM frameworks for the RISC-V Single Cycle Core and the SweRV EL2 Core, significantly enhancing MERL's verification capabilities and project efficiency. Verification Prowess: Developed a PY-UVM framework for the Timer Peripheral, showcasing his strong grasp of verification methodologies. Peripheral Design and Verification: Successfully designed and verified the I2C Peripheral, ensuring its seamless integration and optimal performance within MERL systems. FPGA Expertise: Played a crucial role in the FPGA emulation of both the Timer and I2C Peripherals, providing reliable platforms for testing and validation. Backend Development: Contributed to the backend development of the Liberty Analyzer, leveraging Python skills to enhance the tool's functionality and efficiency. Beyond Technical Expertise Both Muhammad and Ameen have actively fostered a skilled team by serving as trainers in RISC-V based training sessions. Their professionalism, dedication, and ability to tackle complex challenges have earned them high regard from colleagues and supervisors. Seeking New Opportunities? If your organization is searching for talented individuals with expertise in hardware design, verification, Python, and RISC-V, I encourage you to connect with Muhammad Shameel and Ameen in the comments section below. They are eager to explore potential opportunities that align with their skills and experience. #MERL #HardwareDesign #HardwareVerification #RISC-V #Python #Talent P.S. Sharing this post can help connect these valuable researchers with the right opportunities!

56

2 Comments

The proceedings of The 5th IEEE ICTP 2023 are now available in the IEEExplore. Thanks to my supervisor Nayan Sarker sir and co-author Mahidur Rahman. Here is the link to my paper.

6

The IEEE International Requirements Engineering Conference is the premier requirements engineering conference. That prestigious conference usually comprises several workshops including the AIRE workshop. My student Mithila Sivakumar just got two full papers accepted by the AIRE workshop 🎉. The title of one of these papers is "Design of the Safety Case of the Reinforcement Learning-enabled Component of a Quanser Autonomous Vehicle". I am so proud of her achievements, and of the outstanding quality of her work. I would like to seize this opportunity to warmly thank her, as well as the co-authors of her papers, namely: Jinjun Shan, Oluwafemi Odu, Kimya Khakzad Shahandashti and Mingfeng Yuan for their amazing collaboration. Both papers report focus on some of the research activities Mithila Sivakumar completed during her Msc program. I co-supervised Mithila Sivakumar with Dr. Jinjun Shan. She defended her thesis on April 3, 2024. Her thesis is at the intersection of software engineering and machine learning with applications in the automotive sector. The abstract of one of the accepted papers is the following: Safety assurance is paramount across industries where mission-critical systems operate, mitigating risks of catastrophic failures. Safety cases play a pivotal role, particularly in safety critical systems (e.g., autonomous vehicles), in ensuring system reliability and acceptability, providing a structured argument supported by evidence. However, in the safety case literature, it is challenging to get access to a complete safety case, which is crucial for the research community to contribute in this domain. Hence, in this research, we propose an approach to create a safety case for ML-enabled autonomous vehicle, specifically, the Quanser Qcar. We present a complete safety case for a reinforcement learning algorithm applied on the Quanser Qcar to avoid collisions in an unsignalized 4-way intersection. Finally, we report the lessons learned and provide the safety case for the research community to reuse.

15

2 Comments

I am excited to share that my latest IEEE paper titled "AI-Powered Computer Vision for Early Skin Cancer Detection with IoT Connected Dermascopes" has been published! This paper highlights the success of using artificial intelligence and computer vision to detect skin cancer in its early stages through IoT connected dermascopes. I am proud to contribute to the advancement of medical technology and hope this research will lead to more accessible and accurate skin cancer detection methods. #IEEE #AI #medicaltechnology #skincancerdetection

28

1 Comment

mini project on the alcohol and accident detection system using iot under the guidance of the resource person Harish from Karunadu technologies component used : ●lcd ●ADXL ●arudino uno ●alcohol sensor #miniproject #iotandembeddedworkshop #alcoholdetection #accidentdetection

11

Level Up in Parallel Processing!😍 I've just completed the NVIDIA course titled "Getting Started with Accelerated Computing in CUDA C/C++" under the expert guidance of Dr. Muhammad Saeed at UBIT, Karachi University. Huge thanks to Professor Saeed and my school for their sponsorship and support! #UBIT #KarachiUniversity This comprehensive course equipped me with the fundamentals of harnessing the power of GPUs to supercharge my programs through parallel processing. Here's a glimpse into the key concepts I mastered: *CUDA Programming Model: I can now seamlessly manage memory and launch kernels on the GPU, understanding the clear distinction between the host (CPU) and device (GPU) environments. *CUDA Unified Memory: Say goodbye to complex memory management! CUDA Unified Memory simplifies data sharing between CPU and GPU with a unified address space, streamlining development and boosting program efficiency. *Nsight Systems: This powerful profiling and debugging tool empowers me to analyze CUDA application performance, identify bottlenecks, and optimize my code for peak performance. *Kernel Execution: I'm now adept at defining and executing CUDA kernels – parallel functions running on the GPU – using the <<<>>> syntax to configure grid and block dimensions. *Memory Management: From dynamic memory allocation/deallocation with cudaMallocManaged and cudaFree to cudaMallocHost and more, I've gained a comprehensive understanding of memory management on both host and device. *Data Transfer: Efficient data transfer between host and device is key! I've explored methods like cudaMemcpy and asynchronous cudaMemcpyAsync for non-blocking transfers, optimizing performance by overlapping computation and data movement. *Concurrency and Streams: CUDA's advanced features for concurrent execution are now in my arsenal. Stream usage for scheduling and ordering memory operations is essential for maximizing GPU utilization and performance. *Intermediate Concepts: This course delved deeper into page-locked memory, pinned memory, and concurrency requirements – foundational knowledge for advanced CUDA programming. Throughout the course, I gained valuable insights into the GPU architecture, including the role of Streaming Multiprocessors (SMs). This knowledge empowers me to configure kernels and memory operations for optimal performance. By the end, I was confident in writing efficient CUDA programs that effectively leverage the parallel processing power of NVIDIA GPUs. This is just the beginning of my journey into high-performance computing! I'm excited to explore the possibilities of CUDA programming and apply these newfound skills to real-world projects. #NVIDIA #CUDA #GPUProgramming #ParallelComputing #HighPerformanceComputing #UBIT #KarachiUniversity

24

4 Comments

🌟 Day 4 at the DNN BootCamp! 🌟 Another productive day at the DNN BootCamp at Ghulam Ishaq Khan Institute of Engineering Sciences and Technology in collaboration with Skylabs AI and SkyElectric Pvt. Ltd Theory by Dr. Ali Imran Sandhu • Training Neural Networks • Choosing Activation Functions • Multiclass and Multilabel Classifications • Adam Algorithm • Cross Validation • Debugging Lab by Sir Amin: • TensorFlow Implementation of Neural Networks • Plotting Neural Networks Feeling more confident with each passing day as we dive deeper into neural networks! #DNNBootCamp #MachineLearning #DeepLearning #AI #GIKI #Skylabs #SkyElectric #DataScience #NeuralNetworks #TensorFlow #AdamAlgorithm #CrossValidation #MulticlassClassification

20

🎉 Proud to share the culmination of weeks of hard work on our latest project with my classmate, Fares Hegazi We’ve successfully designed and implemented an Arithmetic Logic Unit (ALU) with a robust debouncing circuit using Verilog HDL. This project allowed us to explore the intricacies of digital circuit design and put our skills to the test. 🔧 Project Highlights: Debouncing Circuit: Designed using the delay detection technique, our debouncing circuit ensures reliable input signal processing by mitigating the effects of signal bouncing. The circuit architecture includes a Finite State Machine (FSM) for state transitions based on the input signal and a timer circuit to provide necessary delay information, enabling accurate decision-making. Arithmetic Logic Unit (ALU): The core of our design, the ALU handles a variety of operations, from basic arithmetic to more complex logical functions, all controlled via a simple interface. Additional Blocks: The project also includes a clock divider, 7-segment display decoder, and a binary to BCD converter, all integrated to provide a complete and functional system. 🛠️ Tools Used: Verilog HDL: For designing and implementing the digital logic. FPGA: To test and validate the functionality of the entire system. 🚀 Looking Ahead: This project has not only deepened our understanding of digital design but also fueled our passion for embedded systems and hardware design.

37

16 Comments

🚀 Day 18 of 21 Days of Embedded Programming Challenge ! 🚀 MULTIPLICATION AND DIVISION WITH DISPLAY------> 🔌 Key Task: Perform Any Digit Multiplication and Division Today’s challenge was focused on extending my previous project to perform multiplication and division of any number of digits using a keypad interfaced with an Atmega 2560 (Arduino MEGA). 🔌I also ensured that the results of these operations are displayed properly on a 7-segment display or an LCD, with error handling when results exceed a specific range. 🔌New Tasks Covered: i)Multiplication: Programmed the system to multiply any number of digits. If the product exceeds 9999, it will display "EEEE". Example: Input: 25 × 40 → Output: 1000 Input: 1111 × 8 → Output: 8888 Input: 999 × 12 → Output: "EEEE" (product exceeds limit) Division: The system was programmed to handle two types of division: ii)Quotient Division: It displays only the quotient of any digit division. Example: Input: 1561 ÷ 6 → Output: 260 Input: 765 ÷ 9 → Output: 85 iii)Remainder Division: It displays only the remainder of the division. Example: Input: 1561 ÷ 6 → Output: 1 Input: 765 ÷ 9 → Output: 0 🧑‍💻 Learnings and Progress: 🧑‍💻C Programming Mastery: Writing the embedded code for handling large digit operations was simplified by my constant practice of C programming. Handling edge cases such as overflow in multiplication and ensuring correct display output was a challenge I enjoyed. 🧑‍💻Debugging and User Interaction: The program was designed to wait for a user’s input through the keypad and based on that input, either multiplication or division is performed. Debugging and working with real-time input using interrupts and polling mechanisms helped me refine the project. 🧑‍💻Limit Handling: One key point I tackled is limiting the results to 9999. This was an exciting challenge, ensuring the program recognizes when a result exceeds the 4-digit display limit and returns "EEEE". 🕹️ Future Implementations: I’m excited to extend this project further, such as: 🕹️Displaying floating-point results for division 🕹️Using a touchscreen interface to replace the keypad for a smoother user experience. Stay tuned for more updates! Any feedback or suggestions are welcome! hashtag #EmbeddedProgramming #EmbeddedSystemsProgramming #ATmega2560 #Keypad #keypad #CProgramming #Firmware #ECENAcademy #21dayschallenge #RealTimeProcessing #ArithmeticOperations #18thday #18

39

IEEE 1901.2a: Narrowband Power Line Communication (NB-PLC) Standard IEEE 1901.2a is a standard that specifies communication protocols for Narrowband Power Line Communication (NB-PLC) technology. NB-PLC utilizes existing electrical wiring to transmit data between devices. Here's a closer look at what 1901.2a offers: Applications: Smart Grid Applications:Enables two-way communication between smart meters and utilities for automated meter reading, demand response programs, and grid management. Home Automation:Allows devices like thermostats, lighting systems, and appliances to communicate over existing power lines, facilitating home automation and remote control. Industrial Automation:Provides a communication channel for industrial sensors and actuators integrated into power grids within factories and manufacturing facilities. Technical Characteristics: Frequency Band: Operates in the low-frequency range (typically below 500 kHz) to leverage the good signal propagation characteristics of power lines. Data Rates: Supports moderate data rates, sufficient for transmitting sensor data and control messages in NB-PLC applications. Modulation Technique: Employs Orthogonal Frequency-Division Multiplexing (OFDM) for efficient data transmission over noisy power line channels. OFDM divides the signal into multiple subcarriers, improving robustness against interference. Security: The standard incorporates security features to protect data communication from eavesdropping and tampering. Advantages of 1901.2a NB-PLC: Leverages Existing Infrastructure: No need for additional wiring, making it cost-effective to deploy in buildings with existing electrical wiring. Long Range: Signals can propagate well over long distances within power lines, reducing the need for repeaters or additional infrastructure. Reliable Communication: Operates in a lower frequency band, which is less susceptible to interference compared to higher frequency wireless technologies. Security Features: Provides a level of security to safeguard data communication over the power lines. Limitations of 1901.2a NB-PLC: Data Rates: Lower data rates compared to other communication technologies like Wi-Fi or cellular networks. This might not be suitable for applications requiring high-bandwidth data transfer. Electromagnetic Interference: NB-PLC signals can be susceptible to electromagnetic interference from household appliances or industrial equipment. Standardization: Multiple NB-PLC implementations exist, and interoperability between devices from different vendors might be a challenge. In conclusion, IEEE 1901.2a is a valuable standard for Narrowband Power Line Communication, enabling data transmission over existing electrical wiring. While it offers advantages like cost-effectiveness, long range, and reliable communication, its limitations in data rate and potential for interference need to be considered when choosing a communication technology for a specific application.

2

Advancing PCB Quality Control with AI Exciting research alert! "Automating PCB Fault Detection with YOLOv8 Deep Learning" Authors: Rehman Ullah Khan, Fazal Shah, Ahmad Ali Khan, Hamza Tahir Abstract: Printed Circuit Boards (PCBs) are crucial in modern electronics, connecting various components. As components integrate, quality control challenges arise. This research explores YOLOv8 deep learning for real-time PCB fault detection, evaluating YOLOv8 Small and Nano architectures. Results: - YOLOv8 Small model achieved: - Precision: 98.7% - Recall: 99% - Accuracy: 98.6% - F1 score: 0.98 This research demonstrates the potential of YOLOv8 Small to significantly improve PCB manufacturing quality control, providing a reliable and efficient fault detection solution. Read the full paper here: https://lnkd.in/dVqV_nJu #PCB #QualityControl #AI #DeepLearning #YOLOv8 #FaultDetection #Manufacturing #Innovation

4

1 Comment

We are excited to share that our paper "A DRL Approach for RIS-Assisted Full-Duplex UL and DL Transmission: Beamforming, Phase Shift and Power Optimization" has been accepted for publication in IEEE Transactions on Wireless Communications. In this article, we present a two-stage deep reinforcement learning (DRL) approach for a full-duplex transmission scenario that predicts the phase-shifts of the reconfigurable intelligent surface (RIS), active beamformers at the base station, and the transmit powers of the base station and uplink users to maximize the weighted sum rate of uplink and downlink users. Existing methods require channel state information (CSI) and residual self-interference (SI) knowledge to calculate exact active beamformers or the DRL rewards, which typically fail without CSI or residual SI. Especially for time-varying channels, estimating and signaling CSI to the DRL agent is required at each time step and is costly. We propose a two-stage DRL framework with minimal signaling overhead to address this. The first stage uses the least squares method to initiate learning by partially canceling the residual SI. The second stage uses DRL to achieve performance comparable to existing CSI-based methods without requiring the CSI or the exact residual SI.  The proposed method works well in scenarios with mobile users, too. Further, we propose a DRL framework that can handle pseudo-discrete action space to predict i) quantized RIS phases and ii) reduced phase information with one phase information per group. The quantized phase DRL method is attractive because it results in lesser signaling with better convergence than the continuous phase DRL method. I thank my co-authors, Prof. Sheetal Kalyani and Nancy Nayak at the Indian Institute of Technology, Madras for their invaluable contributions and excellent collaboration! #comsoc #reinforcementlearning #6g

127

6 Comments

Excited to share the completion of my Final Year Project, "Security Analysis of IIoT Devices." Under the guidance of Engr. Umair Shahid. Project aims to tackled the critical challenge of securing IoT devices in industries, which often have vulnerabilities due to manufacturers prioritizing smart features over security.🎓 Project Milestones: --)Developed an advanced IIoT system --)Conducted in-depth studies on Cyber-Attacks targeting IIoT networks --)Created a chatbot for cyber-attack prevention inquiries I extend profound gratitude to my esteemed supervisor, Umair Shahid, for his invaluable mentorship and unwavering support.I am enthusiastic about my involvement in the realm of cybersecurity and eagerly anticipate the forthcoming challenges and prospects. #IIoT #CyberSecurity #FinalYearProject #IoT #IoTSecurity #Engineering #TechInnovation #BahriaUniversity

20

4 Comments