Frequency management and related aspects such as the impact of Radio Frequency Interference (RFI) are growing concerns not only for the design of future Earth Observation (EO) missions, but also operation of EO missions. This presentation will address an overview not only of the raising threats (e.g. a larger demand of spectrum in terrestrial services), but also of the main tasks that ESA needs to carry out to address these issues, such as:
• to interface multiple Member States in CEPT and other Agencies in Space Frequency Coordination Group (SFCG) to ensure a good outcome of World Radiocommunication Conference (WRCs) for the Satellite (vs Terrestrial) Services
• to promote the different interests and Services in many ESA Directorates: e.g. D/TIA in Fixed Satellite Services (FSS), Broadcast (BSS), Mobile (MSS), D/NAV in Radio Navigation (RNSS), D/EOP in Earth Exploration (EESS), D/SCI or D/HRE in Space Research (SRS) services.
• to prioritize: some of these frequencies and services seek new business opportunities and job creation, whereas some others need to be protected for societal benefits (e.g. Climate Change or Numerical Weather Forecast - NWP).
• to perform technical studies:
- to justify requests to ITU regarding the use of specific frequencies for missions (potentially) affected by RF Interference (RFI) from the same or from other services in the same or in adjacent frequencies,
- to propose technical solutions (e.g. exclusion zones, guard bands or acceptable power levels) when compromises for sharing the spectrum are needed
• to (file or) request the use of frequencies for all ESA missions, and give consultancy for other (e.g. commercial in InCubed) projects to make those projects possible.

RF Interference impacts missions in several ways, resulting in additional work (e.g. development of additional filters for sensors or reporting to national authorities in the regulatory status). The risk of RFI goes from simple measurement biases not always detected, through loss of data when detected, and in the worst cases up to potential damage to sensors.

This presentation will provide an overview of all these issues, and will also expand a bit on specific practical cases in Earth Observation missions.

The European space astronomy centre (ESAC) Radio frequency interference (RFI) Monitoring and Information Tool (ERMIT) is the tool developed and used at ESAC with the purpose of handling and managing the information of RFI cases affecting SMOS operations and science data retrieval.

Basically, the RFI ERMIT Tool is made up of 3 parties: a set of implemented software to detect and monitor RFI from the SMOS products, an RFI database where all the processed information is stored, and the third level is the server set, comprising two servers, an application server, and a web server. The RFI monitoring tools implemented at ESAC generate some information about interferences detected by SMOS which is stored in the RFI DB, such as coordinates, brightness temperatures, SMOS passes visibility, etc., as well as maps, reports, statistics…

To achieve the current status of this tool, the road traveled has been very long. SMOS was launched in 2009 and, as it was the first passive L-band radiometer in orbit, the effects of L-band interference were not known globally. Since the first observations of the SMOS mission, its radiometer in the 1400-1427 MHz passive band detected RFI in various areas of the world. Any emission in this band is prohibited by the ITU Radio-Regulations (RR No.5.340).

At the beginning of the mission, when the interference problem became evident, the RFI detection process was manual, analysing the SMOS products one by one and storing the information in Excel files. This worked with the first very strong permanent sources detected, but a large number of sources made that manual process unmanageable. The reporting method consisted of Excel tables and screenshots sent by email together with their cover letter to the respective National Regulatory Authorities (NRA).

As the number of sources for reporting grew, the need for automation became essential. For this, several automatic interference detection algorithms were developed and an FTP was configured to save the information from all the sources detected in Excel tables. A standard document format was also created to inform NRA and quarterly summaries of the global interference environment experienced. More tools were also developed to generate brightness temperature maps by source and probability maps by region.

Finally, the increase in the amount of information has made it necessary to create a database and an interface to access it. This is what the ERMIT tool is all about. But this whole process has been carried out as needs have arisen and it has not been as efficient as if these tools had to be developed now from scratch.

This presentation details the lessons learned throughout this process and how all of these utilities would be developed now that we know what we need for effective and efficient RFI detection monitoring and reporting tools. All this knowledge and tools are and will be of great use to existing and future Earth observation missions, as for example the Copernicus Imaging Microwave Radiometer (CIMR) mission.

Radio frequency interference (RFI) is a threat that is affecting globally the Earth Observation (EO) missions, in particular the microwave passive ones. These sensors estimate the natural electromagnetic emissions from ground, so even signals below the noise floor might affect the measurements. When these contaminated data are used for Numerical Weather Prediction (NWP), it can lead to weather forecast errors.
Ground RFI Detection System (GRDS) is a new concept developed by Zenithal Blue Technologies for RFI detection and mitigation in EO missions. The software is capable of ingesting data products from SMOS and AMSR2 passive microwave satellites, but other sensors will be added in the close future. Then, a number of RFI detection techniques scan for abnormal behavior produced by RFI in any of the intensity, polarimetric, temporal, spatial or statistical domains. All these techniques can be applied with different threshold levels, that can be defined as function of the polarization, latitude, incidence angle, or ground pixel classification. The software also takes into account previous detections over the same area through different internal databases and external information such as airport radars databases or other mission’s RFI sources databases when appropriate to adjust the thresholds in those regions, prone to have RFI contamination. The flags from the different algorithms are combined into a single one per measurement and the EO data product is modified accordingly. GRDS is meant to be placed in the value chain of operative remote sensing missions, between the EO ground processors and the NWP systems.
The system has been successfully tested with SMOS data. In collaboration with the European Center for Mid-Range Weather Forecast (ECMWF), a month of data was processed and then used to feed the ECMWF integrated forecasting system. There, weather information such as temperature, atmospheric pressure, etc. are used to estimate the expected brightness temperature around the globe using the Community Microwave Emissivity Modelling platform (CMEM). The SMOS measurements are collocated over a grid and the differences with the estimated ones are computed, obtaining thus the so-called First Guess Departures (FGD). The statistics of the FGD, in particular the standard deviation, were used as a figure of merit to evaluate the performance of the GRDS system. The brightness temperature of RFI contaminated pixels differs from the expected from their geophysical properties, therefore the standard deviation of the FGD increases. After removing the data screened by the GRDS system, a drastic reduction of the FGD in the areas with stronger RFI contamination is observed (see Figure 1).
The original data (left) shows strong RFI contamination in East Europe, Middle East, or India when no screening is applied. SMOS own RFI flagging (center) leaves some contaminated data, which is drastically reduced with GRDS system (right).
GRDS can also extract statistical information from the RFI scanning process. The generated RFI probability database for SMOS L-band (see Figure 2) shows a high concentration of RFIs at East Europe (Croatia, Greece, Hungary), Middle East and Arabia Saudi, India and Pakistan, China and Mongolia, the Korean peninsula and Japan, and some of them in Africa and America. Observe also four trails of RFI contamination over Australia, one in the center-north, one in the center-south, and two at the middle-east, cause by a new strong RFI source located there.
A data collection from AMSR2 has also been analyzed in order to assess the techniques in the higher frequency channels and to get a preliminary evaluation of the RFI scenario encountered at AMSR2 frequency bands. The results show, among others, RFI presence observed in the reflection of 10.7 GHz satellite signals on the Mediterranean Sea and the Atlantic Ocean near Europe; the reflection of 18.7 GHz satellite signals in the east and west coasts of the USA; 7.3 GHz RFI widespread in Indonesia, Turkey, and Ukraine; and 6.9 GHz widespread RFI in India.
Currently, new capabilities are considered for implementation, such as the addition of new sensors; sea ice detection capability for SMOS, to avoid misclassification as water creating thus false positives over the sea ice; new RFI detection techniques; the use of machine learning ; and the capability to analyze a desired area of the World map to study how the RFI probability evolves along time.
The system architecture and all implemented novelties, the results from the SMOS experimental campaign, and the results from the ASMR2 data processing will be presented at the conference.

Designed, developed and operated to measure radio noise naturally emitted by the Earth and its constituents, space-borne passive remote sensing applications in the Earth exploration-satellite service (EESS(passive)) largely rely on the precondition that no man-made emissions are disturbing the natural radio environment to be measured. This precondition is established thanks to the Radio Regulations (RR) provision No. 5.340, protecting a number of frequency-bands by prohibiting there all emissions.

Operating under the principle that they may coexist with primary spectrum users by employing an underlay spectrum sharing model, UWB (ultra-wideband) technologies enable important solutions across different industries with significant economic value. Whilst this spectrum sharing model is, under certain conditions, generally feasible among radio telecommunication applications, very different is the case where EESS(passive) is to be considered in such spectrum sharing model.

A number of working items involving UWB technologies and EESS(passive) have been running within the CEPT (European Conference of Postal and Telecommunications Administrations) across many frequency-bands. One of them aimed at allowing the operation of next generation of UWB radiodetermination applications for measuring different physical parameters for which it has been considered possibilities for designating radio frequency spectrum resources in the frequency range from 116 GHz to 260 GHz, which includes bands covered by RR No. 5.340. The concept, technical properties and deployment characteristics of these applications were communicated to the CEPT with a request for authorisation of use of spectrum in ETSI system reference document (SRDoc) TR 103 498.

Even though it was agreed that the studies on this very specific type of application (in particular very low number of devices) should not be understood as precedence for general allowance for studies in bands covered by ITU RR 5.340, a new similar request has been made recently for considering possibilities for designating radio frequency spectrum resources in the frequency range from 3.6 GHz to 12.4 GHz in order to allow the operation of Low Frequency MicroWave Security Scanners (MWSSc) (see ETSI SRDoc TR 103 730).

In this contribution, the risks posed by these ETSI requests to the scientific retrieval of space missions operating in the EESS(passive) bands protected by provision No 5.340 are analysed in the framework of the considered spectrum sharing model, aiming to trigger discussions about how to better process them considering the feasibility of possible solutions for mitigating their associated risks.

Passive microwave satellite Earth Observation instruments and even active ones are experiencing increasingly more instances of Radio Frequency Interference (RFI) coming from nearby services or from illegal emitting sources. The problem occurs even in protected frequency bands where man-made emissions are not allowed. At the L-band protected portion of the spectrum for passive Earth Exploration Satellite Services (EESS), instruments such as the European Space Agency’s (ESA) satellite SMOS or the National Aeronautics and Space Administration’s (NASA) satellite SMAP are largely affected by RFI. Large areas in Europe, the Middle-East or Asia are particularly degraded [1]. Some areas are completely lost for SMOS data users.
At higher frequencies, the C, X, and K bands, the Japanese satellite AMSR2, NASA’s GMI or the US Navy WindSat are also reporting the presence of RFI that are affecting their observations at the different channels, such as the 6.9 and 10.65 GHz for AMSR2 [2]. Globally, RFI distribution at these bands varies widely. In general, the different distributions correspond to different frequency allocations for the three International Telecommunications Union (ITU) regions or the presence of certain radio-frequency deployed system in a particular country.
But across the different instruments, similarities appear in the distributions when observing through the same channels. RFI contamination at L band is similarly observed in SMOS and SMAP, or at C band by AMSR2 and Windsat, although differences appear as a result of the instrument specific characteristics. At present however, the interference information from Earth Observation satellite missions is scarce, sparsely disseminated and following different methodologies. In order to obtain a valid assessment of RFI present in a frequency band over time a defined methodology must be consistently followed. If the documentation methodology is not consistently followed over time then a false or misleading trend may be reported.

The ITU differentiates between permissible interference, accepted interference and harmful interference [3], depending on the degree of disruption to the communications service that the interference causes. This definition is not directly applicable to Earth Observation, particularly the passive case, where the signal of interest is the thermal noise emitted by the Earth surface or atmosphere. For that case, Recommendation ITU-R RS.2017 provides a definition of what level of interference is “acceptable” for space based remote sensing operations in all the frequency bands allocated for usage by remote sensing.
However, there is not an exact correspondence between the exceedance of the “acceptable” level of interference and the RFI detected by a mission.
To this purpose, the Frequency Allocation in Remote Sensing technical committee (FARS-TC) from the IEEE Geoscience and Remote Sensing Society (IEEE-GRSS) has proposed the development of a Standard for Remote Sensing Frequency Band RFI Impact Assessment. The purpose of the standard is “to define the quantitative assessment of man-made RFI in a given frequency band. Specifically, this standard is intended to be used in RFI impact evaluations and monitoring of frequency bands allocated to space-based remote sensing. The standard will provide a definition of RFI as it relates to space-based remote sensing operations” [4].

The information derived from the use of this standard is to be used to inform policy decision makers and the public regarding the status, over time, of man-made RFI in any given remote sensing frequency band and its impact on remote sensing operations and products. The information is needed for frequency managers to allocate the efforts to enforce radio-regulations, for space agencies to determine the remote sensing instruments that will provide more benefit to society, and/or to allocate efforts in RFI mitigation techniques, and for researchers to understand the error quantities associated with their retrievals.

The standard development process started in June 2021 with the creation of a first Working Group, and has been having regular meetings ever since. Currently, the team defined the outline of the standard document and is moving towards establishing the definitions of the main key concepts and the main structure for performing the RFI quantification assessment.

The process is still open to anyone interested in the topic trough the IEEE Standard Association usual process.

REFERENCES
[1] A. Llorente et al., "Lessons Learnt from SMOS RFI Activities After 10 Years in Orbit: RFI Detection and Reporting to Claim Protection and Increase Awareness of the Interference Problem in the 1400–1427 MHZ Passive Band," 2019 RFI Workshop - Coexisting with Radio Frequency Interference (RFI), 2019, pp. 1-6, doi: 10.23919/RFI48793.2019.9111797.
[2] D. W. Draper and P. de Matthaeis, "Radio Frequency Interference Trends for The AMSR-E and AMSR2 Radiometers," IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium, 2018, pp. 301-304, doi: 10.1109/IGARSS.2018.8518061
[3] Vol I (Articles) of the Radio Regulations, International Telecommunications Union, Edition of 2016, RR S1-166:168
[4] P4006 – Standard for remote sensing frequency band RFI impact assessment. IEEE Standards Association
[https://standards.ieee.org/project/4006.html]

The Frequency Allocations in Remote Sensing (FARS) Technical Committee of the IEEE Geoscience and Remote Sensing Society (GRSS) was established in the year 2000 to provide an interface between the remote sensing community and the regulatory world of frequency allocations.
Its mission is to educate scientists and engineers on current spectrum management issues and processes relevant to remote sensing, coordinate GRSS technical recommendations to regulatory organizations, track current and future spectrum user requirements, and investigate potential interference issues and promote development of interference detection and filtering techniques.
The technical committee is involved in educational initiatives, such as conference participation, planning of technical sessions, workshops, seminars and tutorials organization. It also participates in international spectrum management meetings and follows the decision-making process of the US Federal Communication Commission (FCC). Online tools publicly accessible on the GRSS website are also currently under development. They include a database of Radio Frequency Interference (RFI) observations and a tool to search and graphically display frequency allocations.
Many ongoing and new activities of the FARS Technical Committee focus on the challenges that the Earth Observation community is facing in regard to Radio Frequency Interference (RFI). For example, the high risk of interference from mega constellations of telecommunication satellites in non-geosynchronous orbit that could result from frequency allocations near the operational bands of the Copernicus Imaging Microwave Radiometer (CIMR) is being discussed in some of the contributions of the technical committee to the ITU-R Study Groups. The potential for RFI from new wireless 5G systems is also being considered very seriously, with a number of activities planned to monitor the evolving situation in this respect. In cooperation with the IEEE Standards Association, the committee is also working to establish recommended procedures to evaluate how much remote sensing frequency bands are affected by interference. Finally, the FARS Technical Committee is developing a document to express the point of view of the remote sensing scientists and engineers on agenda items of the World Radiocommunication Conference 2023 relevant to the Earth Observation community. In the proposed talk, all these activities will be presented and discussed.