Last update on 26/10/2017.
In the early design stage of embedded systems design, it is crucial to identify as many as feasible designs to be further refined later at the low-level design. Identifying the schedulable task mapping at the late design stage is less practical since the complexity of the hardware and software is high. This hinder quick alteration of the design since careful debugging process need to be done. At the early design stage, hardware and software complexity is low and hence a NoC mapping can be alter easily if it does not meet with the design objective. For this purpose, metaheuristics are efficient in finding the mapping. However, the complexity and the high number of evaluation required to determine the schedulability of the hard real-time systems may increase the optimisation time, consequently delaying the selection of the right mapping. A constructive mapping algorithm will be proposed as an alternative way to find the mapping for the systems using new metrics to facilitate the finding of schedulable mapping.
Genetic Algorithms (GAs) have been proved as an efficient way in finding schedulable mapping based on single and multi-objective optimisation for hard real-time embedded NoC. Other metaheuristics are available but yet to be studied in detail of their capability in finding the schedulable mapping for the systems. Although the optimal mapping is hard to find by the metaheuristics due to the arbitrary generation of population, the optimisation time can be a determined factor for selection of the best algorithm. A comparison study will be conducted between the GAs and the other metaheuristics to identify which metaheuristic has the fastest convergence.
Open source home automation platform runs on a variety of hardware and it can connect to nearly any home automation hardware on the market today. The number of manufacturer specific apps exists but for more complex control functions, flexible smart homes hub is one of the options. Based on the openHAB platform, your task is to develop the smart homes hub suited to your proposal, by which low cost IoT sensors (MySensors) can be easily integrated with the open source hardware such as Arduino. As part of your project, anlysis on the power efficiency and a simple cost/benefit calculation must be included.
Heavy rainfall has caused major flooding in Kelantan, Terengganu, the state of Sarawak, Johor, Malacca and parts of Negeri Sembilan, causing millions of ringgit worth of damage. But this is nothing compared to lost of human life. For this purpose, flood detection system to alert the community early during heavy rainfall must exist. However, areas that are affected by flood usually very wide, involving hundreds of kilometres. In this project, a flood detection system based on the low-power wide-area IoT network technology known as LoRaWan will be developed. As part of the project, the power efficiency of the system and cost/benefit calculation must be included.