An automated framework for testing MUOS voice calls

The Mobile User Objective System (MUOS) is the next generation UHF satellite communications (SATCOM) system for the Department of Defense (DoD). It consists of four geosynchronous satellites and four ground stations, uses a Spectrally Adaptive Wideband Code Division Multiple Access (SA-WCDMA) waveform, and it will provide worldwide voice and data capabilities for the military. Before it can be used by soldiers in the field, the system must be extensively tested to verify that it works and meets specified requirements. This paper discusses an automated test capability developed to analyze the performance of voice calls made over MUOS. There are two types of voice calls that can be made: point-to-point (which consists of 2 MUOS terminals that can talk to each other) and group (which consists of 2 or more MUOS terminals where terminals talk to all other terminals connected to the group). The automation software can be used to measure performance metrics that include voice quality, latency, call completion rate, service connection time, and various other timing metrics of the two types of voice calls. The software also logs status information from the radios and the radio´s measurements of Common Pilot Channel (CPICH) received signal code power (RSCP), CPICH signal-to-interference ratio (SIR), satellite beam carrier (SBC), and bit error rate (BER). The logged information can be used to explain poor performance issues, such as failed calls, long service connection times, etc. In addition to analyzing the performance of MUOS voice calls, the automation software also has several benefits over human testing. Typically, in order to conduct voice call tests, a human tester has to be assigned to and operate each radio. Especially for group voice calls which can consist of as many as 20 or more radios, it can be very costly to conduct hours of testing. However, the automation software only requires a laptop to be connected to each radio and does not require any human operators; thus significantly reducing costs. Furthermore, the results obtained from automated testing are objective and more precise than those obtained from human testing. The laptops used for automated testing are synchronized using the Network Time Protocol which enables the automation software to achieve timing accuracy on the order of milliseconds, whereas human testing only achieves accuracy on the order of seconds at best.