Please use this identifier to cite or link to this item:
Title: Pemilihan Pemimpin Mesin Pengendalian Cerdas Nirkabel Dinamis Pada Pemrosesan Terdistribusi
Other Titles: Dynamic Wireless Intelligent Control Machine Leader In Distributed Processing
Authors: Wibowo, Ferry Wahyu
Keywords: Chimpanzee Leader Election Optimization;Classification;Clustering;Energy Efficiency;Group Leader;Wireless Sensor Networks
Issue Date: 28-Feb-2023
Abstract: Optimization is one strategy that is quite capable of getting the results of the approach of a computational problem. The problem of big data, of course, will not be able to be solved using ordinary mathematics quickly or even very impossible to get the results. Developing a meta-heuristic model approach is one solution in placing this field for multi-high dimensional searches in big data. The ability of meta-heuristic algorithms to quickly get optimal values from a problem is an advantage in big data computing. This dissertation aims to create a population-based social meta-heuristic algorithm that only uses random values. It eliminates setting the constant parameters, so the process is more straightforward without having to be reset, just specifying the population value and iterations. The implementation of this meta-heuristic algorithm is applied to the creation of meta-protocols in intelligent wireless sensor networks (IWSN). This meta-protocol helps specify other protocols on the IWSN, such as routing protocols, communication protocols, network clustering protocols, etc. Energy efficiency strategies in wireless sensor networks (WSNs) are critical. The more sensor nodes that report data to the sink, the more energy the WSN consumes, so it doesn't take long for the WSN to shut down or cease to function due to running out of power. Scheduling whether the sensor node is active or not is a matter of urgency; this should be proportional to the energy expended in the WSN with the ability of the WSN to charge for each sensor node. One of the effective strategies for energy efficiency in WSN is to group sensor nodes and choose a leader from each group. This strategy is enough to prolong the life of WSN. The energy optimization approach in WSN is more efficient than the conventional approach. Slow computing will result in delays in getting data from WSN. In addition, the only practical computation for low dimensions will result in the solution-finding process not being optimal because the sensor node consists of hundreds or even thousands of units. This study proposes an algorithm to fulfill the implementation of WSN performance. This study also applies two clustering models: clustering sensor nodes near and far from the sink. Groups near or around the sink will be considered as a classification, while groups far from the sink will be clustered using the CLEO algorithm. After grouping, the sink will compute each group leader according to predetermined criteria. The performance of a proposed meta-protocol has also been compared with other protocol models, such as low-energy adaptive clustering hierarchy (LEACH), stable Election Protocol (SEP), and distributed energy-efficient clustering (DEEC). As a result, WSN-CLEO has the advantage that the sink in the center of the WSN where the mean and standard deviation at the first node dead (FND) is 3103.10±307.42, at half node dead (HND) is 4022.13±34.19, while the last node dead (LND) of sink position at the top center is 4648.87±36.34.
Appears in Collections:D - Doctor of Computer Science

Files in This Item:
File Description SizeFormat 
D_982019010_Judul.pdfJudul3.18 MBAdobe PDFView/Open
D_982019010_Bab I.pdfBab I777.71 kBAdobe PDFView/Open
D_982019010_Bab II.pdfBab II1.11 MBAdobe PDFView/Open
D_982019010_Bab III.pdfBab III1.76 MBAdobe PDFView/Open
D_982019010_Bab IV.pdfBab IV2.29 MBAdobe PDFView/Open
D_982019010_Bab V.pdfBab V745.47 kBAdobe PDFView/Open
D_982019010_Daftar Pustaka.pdfDaftar Pustaka791.41 kBAdobe PDFView/Open
D_982019010_Formulir Pernyataan Persetujuan Penyerahan Lisensi Nonekslusif Tugas Akhir dan Pilihan Embargo.pdfLisensi902.87 kBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.