In this article, we present implementation details of a generative satellite maritime propagation channel model at L-band. In an earlier paper, we presented the parameterization of the channel model for the narrowband case based on recordings of GPS L1 $C/N_{0}$ time-series in different sea state conditions. Additionally, we performed wider band measurements using GPS E5a (≈10 Mcps—chips per second— > 20 MHz bandwidth) and obtained estimations of power delay profiles. Due to lack of dynamic range and delay resolution, in these measurements, no conclusive information was gathered on the time spreading caused by the channel. What was clear is that the time spreading was not significant outside the first signal chip, i.e., $\approx 0.1~\mu \text{s}$ . Still wideband modeling is required within this spreading range. Now, we want to extend the model to the wideband case. To achieve this goal, we have used a rough surface scattering model applicable to sea surfaces and employed it to characterize the time dispersion to be expected. Additionally, we also used this scattering model to help verify the validity of the existing Doppler spread model. Furthermore, ship movements in the waves play a role. We try and analyze this issue. Finally, we present a detailed discussion on various implementation issues we came across when coding the simulator.

Satellite tropospheric propagation studies strongly rely on beacon receiver measurements. We were interested in performing a measurement campaign to characterize rain attenuation statistics. In this article, we outline some of the characteristics and drawbacks one faces when trying to perform a radio wave satellite beacon propagation experiment at the Ka-band with low-cost measurement equipment. We used an affordable beacon receiver consisting of a commercial low-noise block down-converter, an outdoor dual-reflector antenna, and a software-defined radio unit. To measure rain attenuation events, we needed to work out where the reference signal level was at all times. However, as we did not have a radiometer to remove the impact of gases and clouds, since it is a very expensive device, we used a procedure that involved the subtraction of a stable and reliable reference level (template) from the raw received beacon level. This template was extracted from observations during non-rainy periods. The procedure implemented for extracting the template was based on the same data processing methodology used by other authors in this field. Here, we describe through specific examples the main characteristics of the templates extracted on non-rainy days, as well as the impact of some meteorological parameters and unavoidable, but small antenna pointing errors.

In this article, we present a statistical narrowband mobile satellite maritime propagation channel model at the L-band based on empirical observations using GPS C/N ₀ measurements from a receiver onboard a vessel in different sea conditions. The model is fully parameterized for all possible elevation angles. In the measurements, we have used a low-gain, 6.2 dBi, hemispherical antenna. The parameters provided, thus, are valid for this and similar antenna gains.