Algorithm to Automate the Identification of Orbital Positions/Frequency Bands from a Continuously Updated Dataset

Project Description

Yes, it's rocket science.  Yes, it's hard, but do you want to help change the world? To make a disruptive company even more disruptive?  17 years ago little old ManSat from the Isle of Man changed how radio frequency licenes we sought by applying transparent commerical business practices to a logical, yet byzantie process at the ITU.

Today, we need your help to take this one step further by applying machine learning to this process as a next logical step.  If we get it right, this will mean more people on line, more commerce, and more communcations for all on the planet. 

The International Telecommunications Union (ITU) maintains a database of all Geostationary and other orbital spectrum and associated satellite flings. This database is called the Master International Frequency Register, or MIFR for short. PhD level experts use specific ITU software to analyze this database that reference the Radio Regulations. 

Yet, the MIFR is simply a large database that is analyzed using a rigid set of rules, supplemented by a series of equally rigid calculations and equations relating to the power levels of specific satellites and frequency ranges in a logical fashion, to the data. Sound familiar?

We know the data. We know the rules. We know the calculations and equations. What we want is the ability to use machine learning to do all of this, thus freeing the time of our people to actually act upon the results of the data analyzed.  We need your help.

OBJECTIVES

A. Analysis of the International Telecommunications Union (ITU) Master International Frequency Register (MIFR) and ITU databases that record satellite filings filed with the ITU but not yet recorded in MIFR with a view to:

• Identifying available satellite orbital positions and associated frequency bands that would allow for the deployment of new satellite services to a given part of the world (like points on a curve / circle)
• As a part of the above provide a search facility of the ITU SRS database (within which MIFR data is recorded) to search for data on the basis of orbital position, satellite name, satellite filing name, frequency bands, service area, date of submission of filing etc.
• Tasking the algorithms to apply the same IFIC criteria to examine the larger sets of data utilizing the same parameters to seek opportunities for filing and frequency use in the MIFR overlooked by others or not yet anticipated by others: to help us to identify unused or under utilized spectrum and gaps in the orbital arc.

B. Analysis of the bi-weekly Radiocommunications Bureau International Information Frequency Circulars (BR IFICs) and ITU databases (SNS and SNL) with a view to:

• Preparation of responses to IFIC publications, i.e. identification of satellite networks published in the IFIC that may affect a specified satellite filing or satellite networks in operation;
• Prepare a frequency coordination plans for a given filing

• Algorithms could routinely run the numbers on the IFICs when received every two weeks from the ITU giving the same accuracy (or better) than a person doing same.

SUCCESS CRITERIA

A) The algorithm would allow us to identify available orbital positions/frequency bands to offer services to certain parts of the world;

B) Automated preparation of IFIC responses free of any errors.

DATA ASSETS

They data sets we use are referred to by the following acronyms: –SRS, SNS and SNL are available databases maintained by the ITU.

The ITU has also developed a series of software tools (licence free) with which to analyze them.

The software tools can be found here: http://www.itu.int/en/ITUR/software/Pages/spacenetwork-software.aspx

• SRS: contains alphanumeric and graphic information relating to satellite networks and earth stations recorded in the MIFR or in the process of coordination in accordance with Section II of Article 9 of the Radio Regulations or published under the advanced publication of information procedure in accordance with Section I of Article 9.
• SNS: contains Appendix 4 data of geostationary, non-geostationary, and earth station filings.
• SNL: lists basic information concerning planned or existing space stations, earth stations and radio astronomy stations. It includes sections on Advanced Publication Information, coordination requests, notifications, Plans information and their related processing backlog.

Size and Timeliness of the Data 

The IFICs data is published biweekly. These databases range in size from 1 to 10 GB depending on the number of networks published.

Data Collection Mechanism

The IFICs data has to be downloaded from the ITU website every two weeks (requires a subscription). The SRS database is included in these downloads. The SNS and SNL databases are online.

PROPOSALS

Please provide your approach to automate the collection, storing and analysis of the data in the cloud.  We would like a simple system that generates predefined reports on a weekly basis or on-demand. We would also like to understand the number of hours this project would take to get an idea of the budget.