| dc.description.abstract | With the integration of the network into a part of modern life, everything could connect with the network and be a member of it. The Internet of Things (IoT) is booming,
gradually becoming the coming future of daily life; many IoT applications include smart home, smart farm, smart medical, smart factory, smart meters, etc.Each device in the network produces data, forms data streams, and consumes the network resource: bandwidth. Because the network bandwidth is exhaustible reousrce, the bandwidth that can be utlized in each unit of time is limited. When the demands of these data streams exceed the upper limit of network bandwidth capacity, the devices compete with each other in the transmission process and result in network congestion and data transmission delay. This problem causes not only poor user experiences, but even irreversible consequences. For example, the anti-theft system notification delays, so that users miss the prime time to catch thieves. Another example is that a fire accident occurs, but the notification system and sprinkler system delay. These situations must race against time. In order to avoid the occurrence of above situations, this thesis uses the basis of edge computing to effectively utilize the bandwidth in smart home. We employ the edge nodes. An edge node can send the data flows into the Software-Defined Network (SDN) architecture, and the SDN Switch nearby could monitor the traffic produced by edge nodes. The SDN controller can dispatch the resource of the other routes of the topology to spread the risk. Thus, the study efforts in this thesis is able to maximum the overall throughput from various devices in smart home space. | en_US |