| Summary: | 碩士 === 亞東技術學院 === 資訊與通訊工程研究所 === 102 === In this study, the speed test software Nemo Handy installed on an engineering-mode mobile phone was used to measure the key performance indicators (KPI) of circuit-switched (CS) and packet-switched (PS) signal parameters generated from Node B located around the venue of the Spring Scream festival held in Kenting, Taiwan between April 3, 2013 and April 7, 2013. The measured signal parameters were then processed using the postprocessing software of Nemo Outdoor and Nemo Analyze to examine the quality of the communication signals. The items of the KPI parameters that were analyzed included slope, standard deviation (SD), drop call, UL throughput, DL throughput, and web open time (web time). The resulting data were then compared with the internal data generated at the system end of the switching device to understand the behavior of mobile phone users in various environments. Based on the varying behavioral models, the parameters of the node B were adjusted to optimize the performance of the node B. The mobility test yielded a slope value and SD of 0.1055 and 4.07, respectively; the stationary test produced a slope value and SD of 0.0571 and 4.87, respectively. These results showed that the signal quality of the mobility test was superior to that of the stationary test, as indicated by the slope. In addition, the representativeness of the mobility test result was greater than that of the stationary test, as indicated by the SD. Subsequent analysis revealed that the steady values of small-scale fading in the mobility test were all approximately 0.28, implying that the signals of the mobility tests exhibited Ricean fading. During the mobility test, 1 drop call occurred out of the 55 call attempts, yielding a DCR of 1.818%. According to the Splunk software at the system end, the DCRs during circuit switching and packet switching remained below 2%. The connection success rate of the radio resource control (RRC) and radio access bearer (RAB) was maintained above 98% and 97%, respectively. This high success rate was sustained despite using the mobile phone during peak hours, indicating that the resources available at the system end are sufficient.
|
|---|
