Kablosuz Algilayici Aglarda Çoklu Omurga Üzerinden Tüme Gönderim Probleminin Hedef Programlama ile Optimizasyonu
Loading...
Files
Date
2022
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
TOBB ETÜ
Open Access Color
OpenAIRE Downloads
OpenAIRE Views
Abstract
Bu çalışmada, Kablosuz Algılayıcı Ağ (KAA) literatüründe genelde teke gönderim bağlamında ele alınmış olan sıcak nokta probleminin grup haberleşmesindeki karşılığı araştırılmıştır. Çalışmanın ilk aşamasında, tüme gönderim yapan bir KAA uygulamasında çoklu omurgalar kullanılarak maksimum enerji tüketimi enküçüklenmiştir. Çoklu omurga tasarımıyla, ağdaki düğümlere yapılan rol atamasının ve kullanılan bağlantıların değişebilmesine izin verilmiştir. Bu amaçla Akış Tabanlı (İng. Flow Based-FB) ve Düğüm Tabanlı (İng. Node Based-NB) olarak adlandırılan iki farklı Karma Tamsayılı Programlama (İng. Mixed Integer Programming-MIP) modeli kurulmuştur. Performans karşılaştırması yapıldığında NB'nin daha ölçeklenebilir olduğuna karar verilmiş ve çalışmaya bu modelle devam edilmiştir. Ayrıca bir ağdaki kullanılabilecek omurga sayısı için teorik bir üst sınır elde edilmiştir. İkinci aşamada, ağdaki toplam enerji tüketimi ve maksimum enerji tüketimi metrikleri ayrı ayrı eniyilenerek ortak topoloji örneklerinde ağın aynı şekilde davranmadığı kanıtlanmıştır. Son olarak, maksimum enerji tüketimi ve uçtan uca gecikme metrikleri hem ayrı ayrı hem de birlikte eniyilenmiştir. Birbiriyle çelişen bu iki metrik arasındaki ödünleşim çok amaçlı bir eniyileme yöntemi olan Hedef Programlama (İng. Goal Programming-GP) ile analiz edilmiştir. Tüm matematiksel modeller PYTHON kodlama dili ve CPLEX ticari çözücüsü kullanılarak çözdürülmüş, alınan tüm testlerde optimal sonuçlara ulaşılmıştır.
In this study, the counterpart of the hotspot problem in group communications, which is generally considered in the context of unicast in the Wireless Sensor Network (WSN) literature, has been investigated. In the first phase of the study, maximum energy dissipation is minimized by using multiple backbones in a broadcasting WSN application. With the multi-backbone design, the role assignment of nodes in the network and the connections used are allowed to change. For this purpose, two different Mixed Integer Programming (MIP) models called Flow Based (FB) and Node Based (NB) are constructed. When the performance comparison are made, it is decided that the NB is more scalable and the study continues with this model. In addition, a theoretical bound has been obtained for the number of backbones that can be used in a network. In the second phase, the total energy consumption and maximum energy consumption metrics in the network are optimized separately, and it is proved that the network do not behave in the same way in the common topology samples. Finally, maximum energy consumption and end-to-end delay metrics are optimized both separately and together. The trade-off between these two contradictory metrics is analyzed by Goal Programming (GP), which is a multi-objective optimization method. All mathematical models are solved using PYTHON coding language and CPLEX commercial solver, and optimal results are achieved in all tests.
In this study, the counterpart of the hotspot problem in group communications, which is generally considered in the context of unicast in the Wireless Sensor Network (WSN) literature, has been investigated. In the first phase of the study, maximum energy dissipation is minimized by using multiple backbones in a broadcasting WSN application. With the multi-backbone design, the role assignment of nodes in the network and the connections used are allowed to change. For this purpose, two different Mixed Integer Programming (MIP) models called Flow Based (FB) and Node Based (NB) are constructed. When the performance comparison are made, it is decided that the NB is more scalable and the study continues with this model. In addition, a theoretical bound has been obtained for the number of backbones that can be used in a network. In the second phase, the total energy consumption and maximum energy consumption metrics in the network are optimized separately, and it is proved that the network do not behave in the same way in the common topology samples. Finally, maximum energy consumption and end-to-end delay metrics are optimized both separately and together. The trade-off between these two contradictory metrics is analyzed by Goal Programming (GP), which is a multi-objective optimization method. All mathematical models are solved using PYTHON coding language and CPLEX commercial solver, and optimal results are achieved in all tests.
Description
Keywords
Endüstri ve Endüstri Mühendisliği, Industrial and Industrial Engineering
Turkish CoHE Thesis Center URL
Fields of Science
Citation
WoS Q
N/A
Scopus Q
N/A
Source
Volume
Issue
Start Page
1
End Page
60
