.  Home  .  Publikationen  .  Fopra/SEP/Bachelor  .  edmo17

Edmonds, K. O. (2017):

IoT meets HPC: Securely transferring wireless sensor data to a supercomputer utilizing the LRZ Cloud

Small smart devices communicating together via the Internet may achieve significantly more than single devices would be able to: this is the Internet of Things. There is a plenitude of devices and services available for customers. These devices are often very short lived and are only supported by the developer for a short period of time. The lack of upgrades and difficulty to patch the devices often lead to serious security issues. The protocols used in IoT are designed to be lightweight, as the devices usually have limited processing capacity. The heavy lifting is therefore done by the server the IoT devices send their data to. This Bachelor's thesis aims to design and create an end to end solution for sending data collected by wireless sensors to a supercomputer in a secure way, utilizing a central platform in the cloud. The customer requirements were gathered through interviews leading to user stories and raw requirements. Those were analyzed to determine functional and non functional requirements, which were used to develop a system design. Three implementations of the system design were created, with in the end two working prototypes. The prototypes were evaluated against the customer requirements, their power consumption measured, and the prototype satisfying all customer requirements was tested for reliability over a longer period of time. Finally, the results were discussed with one prototype proving to fulfill all customer requirements.