Vi utbildar för framtiden och skapar samhällsnytta genom vår forskning som levandegörs i nära samarbete med näringslivet. Vi bedriver forskning inom computer science, datateknik, software engineering och interaktionsdesign - från grundforskning till direkta tillämpningar. Institutionen har en stark internationell prägel och är delad mellan Chalmers och Göteborgs universitet.
We are engaged in research and education across the full spectrum of computer science, computer engineering, software engineering, and interaction design, from foundations to applications. We educate for the future, conduct research with high international visibility, and create societal benefits through close cooperation with businesses and industry. The department is joint between Chalmers and the University of Gothenburg.
(2022) HRUSTIC, ADI; Chalmers tekniska högskola / Institutionen för data och informationsteknik; Chalmers University of Technology / Department of Computer Science and Engineering; Papatriantafilou, Marina; Duvignau, Romaric
As we shift towards the introduction of 5G in the world, it is indisputable that tools for continuously testing such an environment are needed in order to ensure its success. Ericsson wants to become one of the leaders of 5G solutions with their Evolved Packet Gateway (EPG), which acts as a gateway between a radio network and the Internet. A core component of the EPG is the User Plane (UP), who is responsible for inspecting and transferring payload data of network packets, one of the processheaviest parts of the entire communication. To test the efficacy of the UP, Ericsson has several simulators at their disposal depending on the intended goal. Ericload is one such simulator that can send bidirectional data to the UP as a means to load test it. Currently, Ericload is able to generate model-based traffic, i.e., traffic calculated using mathematical models. While this approach is sensible, it does not entirely reflect the common properties found in real world traffic, such as self-similarity and long-range dependency. Trace-based modelling is a second approach of traffic generation, whose goal is to introduce the real world traffic properties into a simulation by replaying already captured network traces back to networks. The purpose of this thesis is to extend Ericload and implement scalable trace-based traffic modelling and load testing, as well analyze how this adage compares to model-based load testing. The analysis was done using metrics such as packet and data throughput, latency, as well as several verification approaches to ensure a correct trace-based simulation. The results conclude that the implementation of a basic trace-based traffic modelling
was successful, and performs load tests well exclusively by itself or together with model-based traffic, without causing any severe performance issues. The results also shows, theoretically, how scaling captured traces affects its self-similar properties, and how to take these effects into consideration when making any further future improvements.