Detecting Changes in UHF Relay Communications Data Volume Performance for Mars Surface Rovers

The Mars Science Laboratory (MSL) Telecom OperationsTeam and the Mars2020 (M20) Telecom Operations Teamat the Jet Propulsion laboratory (JPL) are developing statisticalmethods for detecting anomalous data volume trends in UHFrelay passes. The MSL and M20 projects rely on UHF relayfor nearly all their scientific data and engineering telemetry.Although variations between predicted and actual UHF relayvolumes are inevitable, it is essential to detect changes in UHFrelay volume performance trends. Changes in volume performancemay be indicative of Telecom hardware health issues,which was the case when the Mars Reconnaissance Orbiter(MRO) developed problems with one of its two Ultra-StableOscillators (USO), leading to intermittent loss of entire UHFrelay passes and ultimately requiring the orbiter to switch tothe second USO. Such changes can also be induced by terrain,and significant losses in average volume performance have adisruptive effect on mission planning and execution since theuplink planning process requires good estimates of upcomingdata volume returns to plan science and engineering activities onMars. The ability to detect changes in long-term volume trendsnot only enables UHF hardware health monitoring but alsoprovides valuable information to the rover planning team duringperiods when terrain and multipath issues may impact volumereturn. We use multiple statistical tests, including Welch’s ttest[1], the Kolmogorov-Smirnov test [2] [3], the chi-square test[4], and a test based on the Irwin-Hall [5] [6] distribution todetect changes in the cumulative distribution function of relaydata volumes. Statistically significant changes, detected using anappropriate statistical significance threshold, are used to alertoperations personnel to unusual trends in volume.