The current cellular technology and vehicular networks cannot satisfy the mighty strides of vehicular network demands. Resource management has become a complex and challenging objective to gain expected outcomes in a vehicular environment. The 5G cellular network promises to provide ultra-high-speed, reduced delay, and reliable communications. The development of new technologies such as network function virtualization (NFV) and software defined networking (SDN) are critical enabling technologies leveraging 5G. The SDN provides flexibility in communication administration and resource management, which are of critical importance when considering the ad-hoc nature of vehicular network infrastructures, in terms of safety, privacy, and security, in vehicular network environments. However, the frequent handover activities in the vehicular network remain a challenge to provide the Quality of Service (QoS) to the users. Artificial Intelligence enabled architecture can be used to cater to the sophisticated demands of modern vehicular Internet infrastructures. The inclination towards robust communications in 5G-enabled networks has made it somewhat tricky to manage network slicing efficiently.

wlpang@mmu.edu.my

The demand for the use of renewable energy sources has increased considerably, in recent years, due to the fast depletion of fossil fuels and population growth. Among the renewable energy sources, the solar photovoltaic (PV) system is a popular and the most economical renewable energy source. Electric power generation using solar PV system utilizes solar panels exposed to the sun. As the position of the sun changes throughout the day, solar tracking systems play a vital role in the efficient operation of a solar photovoltaic system. Therefore there is a need for an efficient controller for the tracking system to maximize the power output from the solar photovoltaic system.

jakir.hossen@mmu.edu.my

Analysis of electrocardiogram (ECG) signal has been an active research area in the past two decades. Most of the past works was motivated by applications in clinical diagnoses. Recently, another potential application of ECG analysis has drawn more attention in the research community. Human individuals presents different patterns in their ECG regarding wave shape, amplitude, PT interval etc. due to the difference in the physical conditions of the heart. This characteristic of ECG signal raises the potential of applying ECG for biometric identity recognition. One important issue in the design of a biometric system is the robustness against attacks. However, many of the existing biometrics are easy to be spoofed. EGG based biometric system is expected to be more universal and hard to mimic.

In this project, we will propose a new approach for identification of human subjects from ECG signals.

rosli@mmu.edu.my

The automatic analysis of biomedical signals has become an important process to both identify indicators of diseased states and to provide meaningful information for various applications in physiology, sports medicine and human–computer interface.

Recent advances in Artificial Intelligence (AI) have the potential to manage and analyze large biomedical datasets leading to the clinical decisions and real-time applications. These techniques have been applied successfully to medical imaging, but the analysis of 1D biomedical signals has yet to fully benefit from this novel approach. ECG, EMG, PCG, PPG, BCG, skin temperature, and skin conductance are some of the 1D biomedical signals examples that record the dynamic changes of the human body.

rosli@mmu.edu.my

The amount of information on the web is overwhelming and escalating everyday. Techniques that are able to extract key insight automatically from all this data will help to tackle the issue.

The research aim to extract semantic information from the natural language text. To retrieve the semantic information, the system is expected to be able to (i) detect sentences structure (ii) identify all named entities in text (iii) find relations between entities and (iv) present information in clear and cohesively manner that help human understanding.

hagoh@mmu.edu.my

One of the leading causes of traffic accidents is the drowsiness of the drivers. Around 20% (Guede-Fernández et al. 2019) of car crashes are occurred due to drowsiness. Accurate early identification of drowsy driving is required to minimize such accidents.

Previous studies mostly focused on facial features and applied conventional machine learning algorithms like SVM, KNN, etc., to detect the driver’s drowsiness. These studies also missed the driver’s slight drowsiness stage and failed to detect the early drowsiness. However, only facial features are not sufficient to detect drowsiness. Drowsiness depends on multiple features like driving performance, behavioral features, and physiological response. Although, previous researches did not utilize these features simultaneously.

Therefore, the study aims to propose a noble artificial intelligence system to predict the driver’s early drowsiness. The system will emphasize all the major features related to the drowsiness of the driver. It will also consider the slight drowsiness stage of the driver to make the model more efficient.

All the fractures will be categorized into behavioral, driving, and physiological features. The behavioral feature includes eye blink, yawning, duration of eye-opening. Vehicle velocity, vehicle acceleration, steering wheel acceleration, and distance from lane center will be considered driving performance. And, the features obtained from the EEG and ECG signal will be contemplated as a physiological response. All the extracted features will be arranged according to the drowsiness level after the pre-processing stages. A noble Artificial Neural Network (ANN) will be implemented to classify and predict the early drowsiness of the driver.

As the proposed system would have the ability to detect drowsiness at the early stage considering all the possible features, it would play a significant role in the driver drowsiness detection system. Moreover, the system would reduce the probability of sudden death of a human being due to drowsiness.

ppem@mmu.edu.my

To develop novel PV system

ervina.noor@mmu.edu.my

Study on the failure analysis of interconnect solder for electronic application

ervina.noor@mmu.edu.my

This project applies Single image estimation techniques to estimate Signal to noise ratio value for images. The images can be material science images, medical images as well as general images.

Note: The research is open for full-time and part-time candidates. For more information, please refer to Prof. Ir. Dr. Sim Kok Swee’s Google scholar page.

Prof. Ir. Dr. Sim Kok Swee

kssim@mmu.edu.my

In healthcare environment, some of medical gloves contain high protein level which increases possibility of healthcare professional to have a risk for latex allergy. The objective of this project is to develop regression methods to estimate protein level through analyses of color difference in glove images. To estimate protein level, latex gloves initially go through chemical test for protein detection. After that, it will be digitally converted into a sample image to perform image processing and calculate color difference values. Those values are used to plot a standard curve for protein level determination.

Note: The research is open for full-time and part-time candidates. For more information, please refer to Prof. Ir. Dr. Sim Kok Swee’s Google scholar page.

kssim@mmu.edu.my

In this work, the effectiveness of ground-based synthetic aperture radar (GBSAR) as an early warning system for surface deformation monitoring will be investigated. In the first part of this study, a 3D model will be derived and simulated to determine optimum working conditions of the GBSAR for different types of natural terrains and man-made structures. Next, field experiments will be carried out to collect measurement data for further analysis and validation. An early warning framework will be proposed to perform real-time monitoring using GBSAR at selected test sites.

vckoo@mmu.edu.my

An accurate estimate of plant disease, its severity and effects is crucial in sustainable crop production. In this work, a new framework of plant disease detection and classification using hyperspectral imaging will be developed. A drone-based hyperspectral imaging system will be used to conduct field measurements at selected plantations / vegetation fields. The dataset will be preprocessed, labeled and trained using state-of-the-art machine learning techniques. The trained model will be applied on a large scale farm to evaluate its effectiveness for early disease detection.

vckoo@mmu.edu.my

Supercapacitor is functioning like conventional capacitor by possessing the potential energy from the opposite charges on the two parallel plates. The current research interest is seeking the maximum effective surface area and the small separation distance of capacitor electrodes in order to increase the capacitance. We intend to prepare the electrode consisting of PANI doped with LiPF6 deposited on the surface of stainless steel plate using chemical deposition technique in this work. A feasible and low cost technique for PANI electrode preparation will be recommended for mass production of supercapacitor in future.

ahyou@mmu.edu.my

This research work presents a novel architecture of an Ultra-Low-Power (ULP) based Hybrid Energy Harvester (HEH) consisting of multiple input sources such as kinetic, thermal and solar energy, harvested from passive human power. Having multiple ambient sources mitigates limitations caused by single sources especially for bodily-worn applications; however, this results in impedance mismatch among the different integrated sources. To overcome this limitation, the proposed ULP based HEH will use one power management unit with Maximum Power Point Tracking (MPPT) algorithm and impedance matching considerations to efficiently manage and combine the power harvested from all three sources to achieve ULP consumptions. Among other crucial sub-modules of the proposed HEH are its Asynchronous Finite State Machine (AFSM) cum resource sharing arbiter to prioritize and share energy sources for overall power reduction, an efficient rectification scheme for the piezoelectric input, an adaptive feedback for low power conditioning, Zero-Current Switching (ZCS) for semi-lossless switching in its on-chip energy storage, a self-start circuit for low ambient startup of 20 mV, a Boost converter to pump-up voltage levels, a Buck regulator to stabilize fluctuating voltage levels, a fuzzy-based micro-battery charger and a multiplexer to switch between harvesting or charging capabilities. All of which is implemented for maximum output extraction and minimal losses. The approach of developing this ULP-HEH begins with a literature review on past HEH architectures. Next, all ULP-HEH sub-modules will be modeled, designed and simulated in PSPICE software with MPPT algorithm; and charging technique in MATLAB. After achieving optimum results, the developed HEH system will then be written in Verilog coding under Mentor Graphics environment and later the designed codes are to be downloaded on the Field Programmable Gate Array (FPGA) board for verification purposes. Once verified, the ULP-HEH will finally be implemented into the layout level using the CMOS 0.13 µm process technology. This ULP based HEH system is expected to operate at 4kHz and deliver up to 3.0-5.0V of regulated voltage output from ambient sources with a minimum input voltage of 35 mV to startup without batteries to aid its proposed (end-to-end) peak efficiency of 90% or above with an output power of about 650 µW when all sources are summed. Also, the proposed HEH is aimed at reducing power consumption to at least 2 times (<70 µW) of the conventional HEH due to multi-sourcing impedance mismatch. The proposed ULP based HEH system can be widely used for low power bodily-worn electrical gadgets, wearable biomedical devices or to power or charge micro-batteries for portable electronic devices.

Note: Full-time only. Candidate needs own financial support.

shabiul.islam@mmu.edu.my

Microchannel flow has applications in electronics cooling and others. The thermal effect of an oscillatory surface on a fluid-filled microchannel is lacking in research and is the subject of this project. The investigation uses analytical and/or numerical methods using principles of fluid mechanics and heat transfer.

Note: Candidate must have a degree in mechanical engineering or in applied mathematics.

cptso@mmu.edu.my

The focal point calibration of antenna is a prime parameter to reduce the background noise of reflected scattering objects for tumor detection in microwave imaging technique. Very small size tumors or any unwanted cell can be detected with high accuracy bycalibrating the focal point of the antenna with minimizing the background noise. However, due to low spectral power density and weak scattering behavior of the body tissue, reflected scattering objects must rely on the very low power reflections, which are often drowned out by the background noise and reflections from the nearby scattering objects. Thus, it is essential to overcome the drawbacks in terms of susceptibility to background noise and scattering objects, false detections, low-resolution scanning, and uncomfortable processing. It is necessary to determine the antenna phase center for accurate ranging, signal processing, and image reconstruction since it makes the antenna simple as a point source. The main aim of this project is to introduce and explore a new approach of antenna focal point calibration by utilizing a new optimization technique. The new parallel-start variant of iterative optimization (PSVI) approach will be assessed to heuristically optimize the focal point of the antenna with respect to the imaging domain. The difference between the two data sets will be calculated and measured due to the generated noise from reflected scattering objects. After that, the antennas will be used to reduce the background noise of reflected scattering objects that will detect the tumor or any unwanted cell inside the body phantom. The noise and artifact reduction of the images using the optimized antenna will be evaluated by comparing with the antenna without the optimal focal point. The expected outcome of this project has the potential to generate noiseless images that significantly paves the detection of tumor or any unwanted cell.

Note: FRGS related Topic – Application submitted.

shabiul.islam@mmu.edu.my

FML are fibers in resin sandwiched between metal layers. In this work, the characterization and strength of FML are studied. 1. To characterize GLARE laminate. 2. To study the relationship between metal thickness and layers of fiber glass of the developed FML. Characterization of fiber metal laminates as follows. The fiber metal laminate (FML) characterized with 2 layers of aluminium alloy sheets of different thickness sandwiched with woven fibers of different thickness joined using epoxy, hardener mixture. The top and bottom layer are aluminium alloy of 0.3/0.5/0.8 mm thickness sheets added with 5 layers of woven E-Glass fiber (200/400/600 gsm) is compared with 7 layers of glass fiber laminates. Firstly, aluminium sheet to be decreased, cleaned. Bonding strength of aluminium is increased by scratching with emery paper. After cleaning the aluminium sheet with acetone. Epoxy preparation: Epoxy resin and the hardener are mixed in a bowl, stirred. Experimental facilities used for this experiment are, tensile testing, Impact test machine used for impact strength analysis as per ASTM D 5628 FD. The deformation, energy absorbed and force upon impact are considered to evaluate the characteristics of the laminates.
The research is expected to contribute in the new knowledge in the field of developing a GLARE (FML) laminate. The developed material will be a stronger, less weight, cost effective alternative to existing FMLs. The new knowledge will be the doorway to open new opportunities of further research and development utilizing the material in automobile, aerospace and Industrial sector. The new alternative material for automotive and other potential industrial application. And the new knowledge will open up for further research and development. (GRA Vacancy)

palanisamy.chockalingam@mmu.edu.my

The project aims to develop a reliable real-time grading system from H&E images of breast carcinoma and evaluate its performance in terms of accuracy and speed.

faizal1@mmu.edu.my

In this project, the aim is to first explore and investigate the fundamental effects of different mobility models on the performance of routing protocols. It provides information on how the natural behaviour of the node affects the transmission with different protocols using different mobility models. A new model will be developed based on the behaviour of the nodes. Hence, a novel dynamic routing protocol will be designed to search for a better efficiency route between the source and destination nodes by utilizing the statistical data obtained from the model. The proposed protocol will be tested using different mobility models in a multi-hop environment. System performance will be analysed in terms of energy consumption, routing overheads and quality-of-service metrics, including throughput and delay.

Note: only full time.

gcchung@mmu.edu.my

Fundamental knowledge and design innovation are imperative to improve energy efficiencies in engineering applications like heat exchangers and electronics cooling. This study proposes a passive control method to enhance heat transfer through the
incorporation of fins into a cylinder in cross-flow. The project aims primarily to correlate the hydrodynamic and thermal boundary layer structure on a heated cylinder in cross flow to the drag coefficient and Nusselt number under low and moderate Reynolds numbers numerically and to be verified experimentally.
Note: Only full-time candidates are accepted.