Multiprocessor System-on-Chips based Wireless Sensor Network Energy Optimization
Wireless Sensor Network (WSN) is an integrated part of the Internet-of-Things (IoT) used to monitor the physical or environmental conditions without human intervention. In WSN one of the major challenges is energy consumption reduction both at the sensor nodes and network levels. High energy consumption not only causes an increased carbon footprint but also limits the lifetime (LT) of the network. Network-on-Chip (NoC) based Multiprocessor System-on-Chips (MPSoCs) are becoming the de-facto computing platform for computationally extensive real-time applications in IoT due to their high performance and exceptional quality-of-service. In this thesis a task scheduling problem is investigated using MPSoCs architecture for tasks with precedence and deadline constraints in order to minimize the processing energy consumption while guaranteeing the timing constraints. Moreover, energy-aware nodes clustering is also performed to reduce the transmission energy consumption of the sensor nodes. Three distinct problems for energy optimization are investigated given as follows:
First, a contention-aware energy-efficient static scheduling using NoC based heterogeneous MPSoC is performed for real-time tasks with an individual deadline and precedence constraints. An offline meta-heuristic based contention-aware energy-efficient task scheduling is developed that performs task ordering, mapping, and voltage assignment in an integrated manner. Compared to state-of-the-art scheduling our proposed algorithm significantly improves the energy-efficiency.
Second, an energy-aware scheduling is investigated for a set of tasks with precedence constraints deploying Voltage Frequency Island (VFI) based heterogeneous NoC-MPSoCs. A novel population based algorithm called ARSH-FATI is developed that can dynamically switch between explorative and exploitative search modes at run-time. ARSH-FATI performance is superior to the existing task schedulers developed for homogeneous VFI-NoC-MPSoCs.
Third, the transmission energy consumption of the sensor nodes in WSN is reduced by developing ARSH-FATI based Cluster Head Selection (ARSH-FATI-CHS) algorithm integrated with a heuristic called Novel Ranked Based Clustering (NRC). In cluster formation parameters such as residual energy, distance parameters, and workload on CHs are considered to improve LT of the network. The results prove that ARSH-FATI-CHS outperforms other state-of-the-art clustering algorithms in terms of LT.
|Keywords||MPSoCs, Mapping, Scheduling, IoT, Wireless Sensor Networks, Energy-efficiency|
|Publisher||University of Derby|
|Department of Electronics, Computing and Mathematics|
|Web address (URL)||http://hdl.handle.net/10545/625262|
File Access Level
File Access Level
|Publication process dates|
|Deposited||15 Oct 2020, 09:02|
|Publication dates||08 Oct 2020|
Attribution 4.0 International
|Contributors||Panneerselvam, John (Advisor) and Xue, Yong (Advisor)|
0views this month
0downloads this month