Akademik Veri Yönetim Sistemi
Tezin Türü: Yüksek Lisans
Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Mühendislik Fakültesi, Elektrik ve Elektronik Mühendisliği Bölümü, Türkiye
Tezin Onay Tarihi: 2017
Öğrenci: BAVER ÖZCEYLAN
Danışman: BUYURMAN BAYKAL
Özet:Internet of Things technology continues to develop as a commercial value and it has become one of the core elements of Industry 4.0 paradigm. Together with that, IEEE 802.15.4e standard provides Time-Slotted Channel Hopping (TSCH) operation mode especially for industrial applications that have strict QoS requirements. In spite of the fact that the standard defines frame structure in MAC layer, there has been no standardization in scheduling for TSCH frame yet. It brings serious challenge for engineering design since industrial applications have stringent demands such as high reliability, low latency and energy efficiency. Industrial applications can generate two types of traffic flows, which are periodic and sporadic. There have been many studies on periodic traffic flow in which it is proven that dedicated cell property of TSCH can fulfill the requirements. Whereas, there is a lack of research on sporadic traffic, which can be handled with shared cell property of TSCH. To this end, a shared cell scheduling method for sporadic traffic flows in Industrial IoT is introduced in this thesis. Proposed method exploits the tree structure of RPL protocol, which is a decisive standard for Industrial IoT, in order to divide the network into independent subnetworks that consist of one parent node and a number of child nodes. The method takes its basis from the idea that the whole network optimization can be achieved by separate subnetwork optimizations, which corresponds to finding an optimum number of shared cells and distribution of child nodes among these cells to handle traffic, which originated from child nodes, with meeting QoS requirements. The method is based on estimations according to a heuristic model of the subnetworks. The results obtained by simulations prove that proposed method minimizes energy consumption without violating QoS requirements. It is also observed that optimum results can be accomplished with this method whether traffic generation rate of child nodes in a subnetwork are the same or different. Thus, granting parent nodes the ability to manage trade-offs between energy consumption, latency and reliability is the outcome of this thesis.