Satellite-Earth Propagation Effects at Low Elevation Angles; Measurements and Modelling
Abstract
Connectivity and high speed data communication have become one of the basic needs of a modern society. The ability to communicate with high data rates nearly anywhere has allowed development of many applications and services that in turn increased productivity. Communication with government and financial institutions, as well as with friends and family, now takes place mostly over an Internet connection. As a consequence the demand for high data rate services in areas not well covered by existing networks is increasing.
Simultaneously there is a growing interest in high latitude regions (>65°N) for multiple reasons. Climate changes open up new transport and travel routes and allow exploitation of previously inaccessible resources. This in turn leads to increase in the number of search and rescue, and patrol, operations.
Satellite communications are often the only means of enabling high-speed access in these remote areas. The need for high data rates also necessitates the use of frequency bands higher than 10 GHz, with more available bandwidth. Currently, only geostationary (GEO) satellites allow this type of access in high-latitude regions for general use. The main disadvantage in using geostationary satellites is the low elevation angle of the links in these regions and correspondingly long path length through the atmosphere. GEO satellites also do not allow communication above about 81 °N as the satellite is below the radio horizon. Since the highfrequency links are more susceptible to atmospheric impairments, the longer path length leads to significant signal degradation.
Accurate prediction and modelling of the atmospheric impairments is necessary for correct dimensioning of the communication systems, design of user equipment and optimal choice of fade mitigation techniques (FMTs). The current models have limited applicability for low elevation angles and high-latitude regions. The models are, for the most part, empirical, with model coefficient being derived by correlating meteorological data with measurement results. These coefficients are mostly based on measurements carried out at mid-latitude temperate climates and moderate elevation angles. The number of measurements conducted at high latitudes and low elevation angles is extremely limited with some measurements at 4-11 GHz carried out in the 1980s and only one recent campaign at 20 GHz in Alaska. In the course of this work a large-scale long term measurement campaign was conducted in Norway at Ka-band (20 GHz), with 12 experiments located at 8 different locations. The elevation angles were 22°, 21°, 14°, 10°, 6.6°, 3.3° and 3.2°. Satellite beacon signal strength as well as weather data were measured at each of the locations. The used setup allowed high sampling rate of 10 samples / second, as well as a dynamic range of about 40 dB. The sites were chosen to cover the typical elevation angles as well as climates encountered in highlatitude regions.
The results revealed that both attenuation and scintillation have slightly different characteristics at low elevation angles. This necessitated modification of typical data analysis techniques. Analysis of rain, gas and cloud attenuation recorded at the stations revealed moderate attenuation levels, largely consistent with the current prediction models at elevation angles of 10° and more. At 3.2° elevation the recorded attenuation was significantly higher due to increased effects of atmospheric gases, but still not far from models. Numerous small issues with the current models were identified, mostly related to the characteristics of precipitation. The largest discrepancies were found at locations with arctic and oceanic climate.
Both large- and micro-scale diversity was studied as a means of avoiding events with large attenuation. Large-scale diversity benefits were confirmed and in very good agreement with models. The impact of realistic switch-over schemes was also investigated and showed small reductions in performance of the diversity configurations. Micro-scale diversity showed some possible benefits.
Analysis of scintillation revealed large inaccuracies in the current models, especially for very low elevation angles. Scintillation/multipath events with over 42 dB fading depth were measured at elevation angles of 3.2 and 3.3° on two locations. Spectral analysis of scintillation showed scintillation corner frequencies much lower than expected by the commonly used scintillation theory. A possible explanation was found based on a different part of the scintillation theory which might be valid due to the different link geometry. Additional measurements conducted for verifying the theory seem to support it.
The results are important for improving the accuracy of the prediction models and were submitted to ITU-R Study Group 3. From practical point of view, no significant issues that were detrimental for using Ka-band links down to 10° of elevation in high-latitude regions were found. At elevation angles of 3.2-3.3°, some longer periods of service unavailability and/or reduced performance are likely, especially during the summer months. First results from elevation angle of 6.6° did not show a large increase in attenuation and scintillation compared with the 10° link.