JultikaUniversity of Oulu repository

Counters provide information about the functionality of the base station. That information is highly valuable for mobile operators, who re-configure their networks partly based on that information. Mobile operators also monitor counters to see, what base station is capable of. Each base station has hundreds of different counters, measuring numerous different things continuously. Counter information is provided by the base station software, which takes care of keeping all the counters up-to-date. Nowadays base stations are very efficient, and they are capable to handle thousands of different requests in the blink of the eye. Increasing complexity of the product poses an enormous challenge for counters, since calculating the values for each counter gets more complex. This research was conducted in Finnish large-scale telecommunications company. Although counters are extremely important for customers, they are not verified effectively in case company’s LTE PDCP/RLC/MAC layer’s capacity tests. The goal of this research is to create a mechanism, which makes it possible to easily enable automatic counter verification in any automated capacity test case. Design science research was applied to achieve this goal. In this research, literature review is conducted to gain understanding for LTE, 3GPP throughput counters, and about the capacity testing environment of the case company. Then a new counter verification system for LTE PDCP/RLC/MAC layer’s capacity tests is designed and implemented. After the system is implemented, expected counter values need to be calculated for each test case and counter, which are part of this research. Evaluation of the system is made against the system requirements and to accuracy of limit value calculations. As a conclusion, it can be said that the implementation of the system was a success, but in some of the test cases, counters provided unexpected results. The implemented system was able to catch the faults, but the root causes for problems are not clear. In total, 185 test case – counter combinations were verified, and in almost 13 % of them counter verification failed the test case, because counter provided unexpected value. In future, it would be beneficial to make a root cause analysis for the issues this research pointed out.