National Project:

Project name: Developing software for building the geological, land composition and organism database for Vietnam’s natural museums (DATP.03/15-17).

Principle investigator: Dr. Vu Van Thieu

Duration: 6/2016 – 12/2019

Funding: Vietnam Academy of Science and Technology

Description: The project aims to develop the software used in building the geological, land composition and organism database for Vietnam’s natural museums. The software has the following requirements: fast data access and writes (maximum query time is 0.4s, according to MySQL’s QRTi standard), accuracy (data is not lost during queries, sufficient data is displayed when performing keyword searches), security, and other properties equivalent to software being utilized at major museums internationally.

National Foundation for Science and Technology Development (NAFOSTED)

1. Project name: Towards energy unlimited IoT networks by nature-inspired algorithms.

Principle investigator: Assoc. Prof. Huynh Thi Thanh Binh

Duration: 4/2020-4/2022


Description: Along with the development of microprocessor technology, and intelligent data collection and processing techniques, Internet of Things (IoT) have been used more widely and plays an important role in the fourth industrial revolution. Most IoT networks are self-operating networks, whose components are usually limited in energy. Therefore, one of the most critical problems in IoT networks is to optimize the energy consumption of the nodes, thereby maximizing the network’s lifetime. This problem has attracted a lot of attention from the academic community. Energy consumption of network nodes depends on many factors, from the installation to the operation. As a result, optimizing energy consumption is a difficult problem, especially in the context of new generation of IoT networks, which has complex configurations and uses advanced technologies.

Although there have been many studies tackling energy optimization for IoT networks, the studies so far only focus on certain types of networks and solve the optimization problems under ideal assumptions. In this research, we aim at a comprehensive energy optimization solution for IoT networks. Specifically, we propose network models that are closer to reality and provide energy optimization solutions using the most advanced technologies. Energy optimization problems are NP-hard problems and we often cannot find the exact solution for large datasets. Our approach is to exploit nature-inspired algorithms in order to obtain the most efficient solutions.

2. Project name: Algorithmic and combinatorial methods on some discrete structures

Principle investigator: Assoc. Prof. Do Phan Thuan

Duration: từ 4/2017 – 4/2019


Description: applying combinatorial and graphs algorithms to some combinatorial optimization problems in communication networks, network intrusion detection, optimization problems in artificial intelligence.

3. Project name: Developing metaheuristic techniques for solving optimization problems in distributed and software systems (DFG 102.01-2016.03).

Principle investigator: Assoc. Prof. Huynh Thi Thanh Binh

Duration: 7/2017-7/2019


Description: Nowadays, integrating different types of resources in distributed systems and building software systems that run across heterogeneous network infrastructures are emerging as new technology trends, especially in IoT. With the current technology, IoT applications are required to process huge amounts of data in a distributed and real time environment with limited capacity of processors, memory, batteries and power source of the terminal devices. This research includes  two main related parts to address the aforementined problems:

  • IoT infrastructure: Solving optimization problems in the deployment of WSNs when applied to IoT systems; Reducing the latency in network communications of IoT applications.
  • Software development for IoT applications: Solving the problem of IoT software optimization; Solving the problem of automated data generation for software testing in IoT environments; Assessing and testing services in IoT environments.

It has turned out that all of these above problems could be formulated as combinatorial optimization problems. They are all NP-Hard problems. Currently, IoT applications are solely utilizing old technology or less-developed metaheuristic techniques to solve optimization problems in IoT infrastructure and IoT software development. Therefore, developing metaheuristic techniques to solve NP-hard problems in distributed and software systems has become a promising research direction, and further studies into this area are highly necessary.

4. Project name: Prediction-based optimization for dynamic transport scheduling. (FWO. 102.2013.04)

Principle investigator: Dr. Pham Quang Dung

Duration: 4/2014-4/2016


Description: The objective of the project is to develop algorithms for solving optimization transportation problems in both static and dynamic scenarios. In the first phase, we focus on designing and constructing a framework for solving large scale combinatorial optimization problems, especially, optimization vehicle routing problems. Then, we develop models and learning algorithms for predicting transportation requests. Based on the prediction information, we develop algorithms for solving dynamic vehicle routing problems in which transportation requests are revealed online during the execution of the planning.

Ministry of Education and Training projects

1. Project name: Developing optimal data transmission algorithms and code-generation tool for GPUs in several high-performance computing problems (B2015-01-90).

Principle investigator: Dr. Vu Van Thieu

Duration: 2015-2017

Funding: Ministry of Education and Training

Description: The project aims to solve 2 problems affecting the performance of CUDA programs: synchronization and CPU-GPU communication. Additionally, manual implementation of CUDA programs are often time-consuming and sub-optimal. The project thus proposes methods to deal with the aforementioned bottlenecks, as well as develop a code-generation tool for generating efficient CUDA programs.

International collaboration projects

1. Project name: Multitasking Evolutionary Algorithms for Optimizing Artificial Neural Network and Graph-based Models

Principle investigator: Assoc. Prof. Huynh Thi Thanh Binh

Duration: 2019-2022

Funding: US Army Research Lab, US Army International Research Center – Asia Pacific.

Description: Many applications in economy, industry and science need to solve combinatorial optimization problems such as routing, scheduling, network design, and neural network training. Most of these problems are also NP-hard. Approximation algorithms do not guarantee optimal solutions to these problems. However, they provide acceptable execution time on large datasets. As no known technique have been found to adequately solve large-scale NP-hard problems exactly, evolutionary algorithms have become a viable method for approximating them in recent years. This project aims to propose and develop multitasking evolutionary models for solving graph optimization, neural network training problems applied in various problem domains.