Malak Abid Ali Khan

Beijing Institute of Technology

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

Hongbin Ma

Beijing Institute of Technology

Syed Muhammad Aamir

Beijing Institute of Technology

Zia ur Rehman

Quaid-i-Azam University

Imran Khan Mirani

Beijing University of Technology

Ying Jin

Beijing Institute of Technology

Jingxiang Ma

North China Electric Power University

Hongbin Ma

Zhejiang University

Mizanur Rahman

Beijing Institute of Technology

Anil Baris Cekderi

Beijing Institute of Technology

Senlin Luo

Beijing Institute of Technology

Cooperative Institutions

Beijing Institute of Technology

Beijing Technology and Business University

North China Electric Power University

University of Engineering and Technology Peshawar

Jilin University

Shanghai Jiao Tong University

Ministry of Natural Resources

Guangzhou Marine Geological Survey

Shenzhen Institute of Information Technology

Harbin Institute of Technology

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

Real-Time Object Detection in Occluded Environment with Background Cluttering Effects Using Deep Learning

arXiv (Cornell University) (2024)

Syed Muhammad Aamir Hongbin Ma Malak Abid Ali Khan Muhammad Aaqib

Detection of small, undetermined moving objects or objects in an occluded environment with a cluttered background is the main problem of computer vision. This greatly affects the detection accuracy of deep learning models. To overcome these problems, we concentrate on deep learning models for real-time detection of cars and tanks in an occluded environment with a cluttered background employing SSD and YOLO algorithms and improved precision of detection and reduce problems faced by these models. The developed method makes the custom dataset and employs a preprocessing technique to clean the noisy dataset. For training the developed model we apply the data augmentation technique to balance and diversify the data. We fine-tuned, trained, and evaluated these models on the established dataset by applying these techniques and highlighting the results we got more accurately than without applying these techniques. The accuracy and frame per second of the SSD-Mobilenet v2 model are higher than YOLO V3 and YOLO V4. Furthermore, by employing various techniques like data enhancement, noise reduction, parameter optimization, and model fusion we improve the effectiveness of detection and recognition. We further added a counting algorithm, and target attributes experimental comparison, and made a graphical user interface system for the developed model with features of object counting, alerts, status, resolution, and frame per second. Subsequently, to justify the importance of the developed method analysis of YOLO V3, V4, and SSD were incorporated. Which resulted in the overall completion of the proposed method.

10.48550/arxiv.2401.00986

Cite

Citations (0)

Design and Operation Control of an Indoor Storage Crane

Communications in computer and information science (2023)

Mizanur Rahman Yiming Duan Malak Abid Ali Khan Zia ur Rehman Hongbin Ma

Relocation

Data striping

10.1007/978-981-99-7593-8_18

Cite

Citations (0)

Improved Cuckoo Algorithm for Adaptive Adjustment of Discovery Probability

Kun Wang Jia Liang Han Malak Abid Ali Khan Xiaofeng Lian

The Cuckoo Search Algorithm (CS) is a bio-inspired new swarm intelligent optimization algorithm. Aiming at the slow convergence speed and weak search ability of CS algorithm in the late stage of search, an improved cuckoo algorithm (APCS) for detecting the adaptive adjustment of probability parameters is proposed. Firstly, the state discriminant parameter P _s is calculated by Pareto optimal solution and then the balance parameter P _eb is calculated by exploring-exploiting the equilibrium state, finally the adaptive dynamic adjustment of the found probability P _eb of the bird egg is realized. The performance of the two algorithms is analyzed by 10 benchmark functions in 10 and 30 dimensions. The results show that the convergence speed, optimization ability, stability and computation time of APCS algorithm are better than CS algorithm.

Cuckoo search

Benchmark (surveying)

Cuckoo

10.1109/ccdc.2019.8833285

Cite

Citations (0)

When LoRa meets distributed machine learning to optimize the network connectivity for green and intelligent transportation system

Green Energy and Intelligent Transportation (2024)

Malak Abid Ali Khan Hongbin Ma Arshad Farhad Asad Mujeeb Imran Khan Mirani

10.1016/j.geits.2024.100204

Cite

Citations (1)

Optimizing the Performance of Pure ALOHA for LoRa-Based ESL

Sensors (2021)

Malak Abid Ali Khan Hongbin Ma Syed Muhammad Aamir Ying Jin

(1) Background: The scientific development in the field of industrialization demands the automization of electronic shelf labels (ESLs). COVID-19 has limited the manpower responsible for the frequent updating of the ESL system. The current ESL uses QR (quick response) codes, NFC (near-field communication), and RFID (radio-frequency identification). These technologies have a short range or need more manpower. LoRa is one of the prominent contenders in this category as it provides long-range connectivity with less energy harvesting and location tracking. It uses many gateways (GWs) to transmit the same data packet to a node, which causes collision at the receiver side. The restriction of the duty cycle (DC) and dependency of acknowledgment makes it unsuitable for use by the common person. The maximum efficiency of pure ALOHA is 18.4%, while that of slotted ALOHA is 36.8%, which makes LoRa unsuitable for industrial use. It can be used for applications that need a low data rate, i.e., up to approximately 27 Kbps. The ALOHA mechanism can cause inefficiency by not eliminating fast saturation even with the increasing number of gateways. The increasing number of gateways can only improve the global performance for generating packets with Poisson law having a uniform distribution of payload of 1~51 bytes. The maximum expected channel capacity usage is similar to the pure ALOHA throughput. (2) Methods: In this paper, the improved ALOHA mechanism is used, which is based on the orthogonal combination of spreading factor (SF) and bandwidth (BW), to maximize the throughput of LoRa for ESL. The varying distances (D) of the end nodes (ENs) are arranged based on the K-means machine learning algorithm (MLA) using the parameter selection principle of ISM (industrial, scientific and medical) regulation with a 1% DC for transmission to minimize the saturation. (3) Results: The performance of the improved ALOHA degraded with the increasing number of SFs and as well ENs. However, after using K-mapping, the network changes and the different number of gateways had a greater impact on the probability of successful transmission. The saturation decreased from 57% to 1~2% by using MLA. The RSSI (Received Signal Strength Indicator) plays a key role in determining the exact position of the ENs, which helps to improve the possibility of successful transmission and synchronization at higher BW (250 kHz). In addition, a high BW has lower energy consumption than a low BW at the same DC with a double-bit rate and almost half the ToA (time on-air).

Aloha

Duty cycle

10.3390/s21155060

Cite

Citations (23)

Data Analysis of LoRa

Harvard Dataverse (2023)

Malak Abid Ali Khan

NS3 (Network Simulator) is for the simulation to acquis the data, and the LoRa calculator is used for confirming the results. We selected the parameters for designing the network to maintain a smooth data flow. The machine clustering defined the range for the end devices. The spreading factors and the data rate are allocated to those clusters to analyze the results. Different gateways (1, 2, 4, 10, 20) were implemented to measure the SNR, RSSI, and ACK during the concurrent transmission from gateways to end devices. Analyzed the airtime, delay, and duty cycle to mitigate data losses. The bandwidth and spreading factor were modified to minimize the saturation. The duty cycle and bandwidth significantly impact the life span of the end devices.

10.7910/dvn/spbyba

Cite

Citations (0)

LoRa Meets Artificial Intelligence to Optimize Indoor Networks for Static EDs

Transactions on Emerging Telecommunications Technologies (2025)

Malak Abid Ali Khan Senlin Luo Hongbin Ma Amjad Iqbal

ABSTRACT The architectural design of the Indoor Internet of Things ( IIoT ) network targeting static end devices ( EDs ) and gateways ( GWs ) has been innovatively formulated in this paper, integrating LoRa technology to mitigate losses and ensure seamless information reception through meticulous ED allocation. The arrangement of simultaneously transmitted data within the network server ( NS ) employs a deep neural network ( DNN ) with distributed machine learning ( DML ) to adjust transmission parameters, ensuring frequent uninterrupted bidirectional communication. This augmentation is obtained by strategically deploying EDs within distinct clusters determined by K‐means and density‐based spatial clustering with noise ( DBSCAN ), thus optimizing spreading factor ( SF ) and data rate ( DR ) allocation to prevent data congestion and improve signal‐to‐interference noise ratio ( SINR ). The proposed hybrid model ( DR | SF ) for pure and slotted ALOHA amplifies the network's performance metrics for indoor scenarios. A unified framework utilizing a one‐slope model estimates path losses ( PL ) while exploring various bandwidths ( BW ), bidirectional interrogations, and duty cycles ( DC ) to lower the saturation and prolong the active lifespan of the EDs . The results manifest a packet rejection rate ( PRR ) of 0% for the DBSCAN , contrasting a 4.7% estimate for the K‐means. The network saturation is minimized to 9.5% and 10.1%, correspondingly, significantly increasing the efficiency of slotted ALOHA (91%) and pure ALOHA (90.6%), thereby prolonging the longevity of EDs .

10.1002/ett.70060

Cite

Citations (0)

Optimization of LoRa for BIoT based on ML: A case of ESL

Alexandria Engineering Journal (2023)

Malak Abid Ali Khan Zia ur Rehman Jingxiang Ma Hongbin Ma

Several technologies provide cellular connectivity to transmit data to end devices (EDs) in the indoor environment. These technologies have short-range; complex network architecture; short lifespan; and high power consumption. In this paper, LoRa is utilized to design the Building Internet of Things (BIoT) system for electronic shelf labels (ESL) to provide long-range connectivity. To minify the transmission losses, the concurrent data exploits data parallelism for updating transmission parameters at the network server (NS) for non-interruptive communication between the centroid gateways (GWs) and allocated EDs. The EDs are deployed by K-means at different locations in each cluster based on spreading factor (SF) and data rate (DR) to avoid data congestion and intra-SF interference for the SF and hybrid (DR/SF) models. The varying bandwidths (BWs) and duty cycles (DCs) enhance the lifespan of the EDs, while the Bayesian game parameter selection (BGPS) method minimizes the power losses among the EDs. The one-slope estimates medium losses; the adjusted R-square predicts variance; and Pearson finds a correlation among the measured values. The hybrid model improves the network's performance to a spiraling average efficiency of 90.48 % and 89.64 %, with network saturation of 10.185 % and 10.337 % for pure and slotted ALOHA, respectively. The overall results illustrate a packet loss ratio (PLR) of 20.11 % for the SF model and 4.861 % for the hybrid model. The energy dissipation plunged to 0.06032 J per day, prolonging the EDs' life span to 14265.34 days.

Aloha

Duty cycle

10.1016/j.aej.2023.10.064

Cite

Citations (2)

Object Recognition in Different Lighting Conditions at Various Angles by Deep Learning Method

arXiv (Cornell University) (2022)

Imran Khan Mirani Tianhua Chen Malak Abid Ali Khan Syed Muhammad Aamir Waseef Menhaj

Existing computer vision and object detection methods strongly rely on neural networks and deep learning. This active research area is used for applications such as autonomous driving, aerial photography, protection, and monitoring. Futuristic object detection methods rely on rectangular, boundary boxes drawn over an object to accurately locate its location. The modern object recognition algorithms, however, are vulnerable to multiple factors, such as illumination, occlusion, viewing angle, or camera rotation as well as cost. Therefore, deep learning-based object recognition will significantly increase the recognition speed and compatible external interference. In this study, we use convolutional neural networks (CNN) to recognize items, the neural networks have the advantages of end-to-end, sparse relation, and sharing weights. This article aims to classify the name of the various object based on the position of an object's detected box. Instead, under different distances, we can get recognition results with different confidence. Through this study, we find that this model's accuracy through recognition is mainly influenced by the proportion of objects and the number of samples. When we have a small proportion of an object on camera, then we get higher recognition accuracy; if we have a much small number of samples, we can get greater accuracy in recognition. The epidemic has a great impact on the world economy where designing a cheaper object recognition system is the need of time.

3D single-object recognition

Position (finance)

10.48550/arxiv.2210.09618

Cite

Citations (2)

Performance of Slotted ALOHA for LoRa-ESL Based on Adaptive Backoff and Intra Slicing

Malak Abid Ali Khan Hongbin Ma Syed Muhammad Aamir Anil Baris Cekderi Mustak Ahamed

Slotted ALOHA implemented in the internet of things (IoT) uses the LoRaWAN media access control (MAC) protocol to improve its performance. Various backoff algorithms have been proposed in LoRaWAN to evaluate the delay, throughput, and packet loss rate (PLR). However, an adaptive backoff algorithm has been implemented in this paper to examine the performance metrics for electronic shelf labels (ESLs). The use of adaptive backoff optimizes the delay and the bandwidth (BW) for more efficient and meaningful communication with a certain degree of data loss. The results illustrate that the intra-slicing model and adaptive backoff estimate the optimal delay for each slice, starting with the best slicing priority for the end device (ED) which brings mobility into the network.

Aloha

Exponential backoff

10.1109/iccis56375.2022.9998155

Cite

Citations (4)