Experiment Plan

The characterization of fundamental atmospheric propagation effects (gaseous absorption, rain attenuation, cloud attenuation, scintillation, and depolarization) on the V-band is to be investigated from three Earth stations located in three different rain rate regions. A data collection period of 36 months is suggested, with up to 60 months necessary for statistically valid analysis.

Recognizing the atmospheric parameters and predicting their effects in advance is an integral part of the experiment as it provides a good ballpark estimation of the expected phenomena based on simulation models.

Based on these simulation models, our experiment will validate and further refine the correlation between V-band propagation phenomena and atmospheric parameters using collected data. Through this approach, we will produce a more reliable propagation model to support systems engineering, link assessment, and design of future operational military satellite communication architectures and systems in the V-band.

Table 1 shows the approach adopted to reach each objective.

Table 1. V-Band Experiment approach 

Setup

Each Earth station will be equipped with a two-channel beacon receiver and a two-channel radiometer at frequencies of 71 GHz, 73.5 GHz, and 76 GHz.  Each station will be equipped with a 0.3 m diameter receive antenna (See Antennas and Coverage).  Meteorological stations placed at each site will provide all the important surface parameters such as temperature, humidity and rain intensity. We will also employ the use of synoptic and radiosonde data provided by the U.S. National Weather Service to correlate measured propagation effects to atmospheric phenomena. See the Beacon Design and Receiver Design sections for more information on those components.

Data Acquisition

For each location, our network of data sensors and third party sources will track the following atmospheric and meteorological features:

• Air temperature
• Air pressure
• Air humidity
• Precipitation
• Geometric structure of clouds
• Haze and contents of the air
• Solar and terrestrial radiation

At each receiving station, a vector network analyzer will be used to track the phase distortion by passing the received signal and reference signals through the phase comparator.  In order to characterize depolarization, we use an OMT to separate the V and H polarizations into separate channels so the co-polar strength and cross-polar strength can be calculated and compared against a priori knowledge of the transmitted signal polarity[2]. Scintillation information will be extracted from data collected at the three sites.

Autonomous Operations

Each Earth station must collect data autonomously and measurements of the signal amplitude, XPD, phase, atmospheric and meteorological parameters must be obtained at a sufficient sampling frequency. However, given the long duration of the experiment, continuous collection of data is not feasible given the vast amount of data that needs to be recorded and processed. We have chosen the sampling rate of the signal to be 100 Hz for a duration of 10 seconds per one minute window, which is more than sufficient for a detailed study of tropospheric scintillation[1]. The data acquired has to be time stamped to accurately coordinate the measurements with external weather data. We will achieve that through the use of a GPS receiver at each site to time tag all data locally. To provision for power outages, backup power is installed at each Earth station to provide uninterrupted data collection. All raw data will be stored locally and forwarded to our remote data processing centre on a daily basis. Forwarded data will be checked for validity to ensure that it does not corrupt the data set.

Data Storage and Analysis

At our central data processing location, we require a highly scalable storage solution to store all the data obtained from multi-site measurements and third party sources, which grows rapidly as the experiment progresses. Moreover, the data has to protected from multiple storage failures.

On the other hand, in order to produce good characterization of the V-band channel and examine the effects of complex atmospheric conditions on the channel, it is necessary to develop new atmospheric propagation models.  These models must extract information from multiple parameters that are related to the atmospheric conditions, perform sophisticated classification and mapping in a multidimensional input–output space, rapidly process the large amounts of data that make up atmospheric condition while in an operational mode, and automatically self-update whenever new sources of data are made available.

To meet all the requirements above, we opt for the combined use of EMC Isilon and Greenplum UAP (Unified Analytics Platform) solutions. The Isilon scale-out NAS storage system will enable us to seamlessly expand our storage capacity according to the growth of data while providing automatic replication of data from each site to remote sites without impacting other network resources[3]. The Greenplum UAP features Greenplum Database for structured data and Greenplum Chorus that acts as the productivity layer for our data science team[4]. Greenplum Database features a massively parallel processing (MPP) database which will enable us to manage, store, and analyze petabytes of structured/relational data with its native support for Big Data analytics. Greenplum Chorus will enable our data science team to easily collaborate, run data-mining algorithms, and derive insights from the Big Data.
 All these will enable us to model the complex relationships between atmospheric parameters and propagation phenomena to find patterns in the large data sets that we will collect over the duration of the experiment.

References:

[1] Del Pino, P.G. ;Riera, J.M. ; Benarroch, A., ”Measurements of tropospheric scintillation on millimetre-wave satellite link”, Electronics Letters, Oct. 25 2007.

[2]  R. Acosta, J. Nessel, R. Simons, M. Zemba, J. Morse, and J. Budginer, “W/V-Band RF Propagation Experiment Design,” 18^th Ka and Broadband Communication Conference, Ottawa, Canada, Sept 2012.

[3] http://www.isilon.com/

[4] http://www.greenplum.com/products/greenplum-uap