EMC and Interference Aspects

An insight to electromagnetic compatibility (EMC) in avionics systems has been done in deliverable D12. Since all avionic devices are subject to electromagnetic interference (EMI), any new electronic device to be installed in an aircraft must be assessed for certification. For this reason it was found necessary to describe the proposed way of interaction tests to determine worst-case interference with passenger PEDs. The main objective of the introduced test method is to identify possible interactions due to intentional emissions of wireless services and must be carried out by all existing standards and their associated frequencies; spurious emissions from PEDs cannot be covered by this test. The figure below (left) depicts a test setup for aircraft interaction tests, where aircraft systems behavior shall be investigated while stressing the aircraft systems at an artificially increased power level. Several different positions for the testing antenna in the aircraft have to be investigated from electronic bay to cockpit and different passenger positions, as it is shown in the right figure.


(a) Possible Test Setup for Aircraft Interaction Test

(b) Indication of test positions



A second part of the interference study was dedicated to the interference that the wireless on-board network may produce to licensed terrestrial networks.

Extensive out-cabin measure-ments in two aircraft bodies at GSM and UMTS frequencies have been documented, and results have been deeply analysed. On the left, a picture during out-cabin measurement in the A330-200 aircraft with a crane to reach positions above ground level.


On the left, a coloured power map of the received power ground level is shown. Ground positions are marked as described in D12. In this scenario, the Tx. antenna working at 1950 MHz was placed beside a window, close to the red spot in 4G.

After analysing the power distribution outside the cabin depending on the transmitter position, aircraft conditions (doors open), shadowing (due to the wings) and signal path, a realistic worst case fuselage attenuation estimation of 10 dB has been obtained.

Without the application of countermeasures interference may occur and straight communication between terrestrial and on-board systems could happen. For example, for an aircraft flying at 10 Km height with a base station transmitting at maximum power the signal would arrive to ground UMTS base stations over their sensitivity level in an area up to 11 Km of radius from the vertical projection of the aircraft, see figure on the right.

Hence, countermeasures to avoid interaction between the on-board network and terrestrial networks were proposed by lappropriate power control. This was demonstrated by link budget calculations for GSM systems. Additionally, these calculations also provided the dimensioning of this transmission control unit feeding a leaky line antenna.

It has been concluded that airborne GSM 1800 pico-cells will not interfere with terrestrial cellular networks if mobile use the onboard network, and if power control is applied by the airborne pico-cell to keep emissions at lowest level. Moreover, it has been demonstrated that in a GSM 1800 system signals do not interfere with terrestrial mobile systems when aircraft is at cruising, and that the on-board wireless network must be coordinated at gate position, ascending and descending maneuvers for interference reasons. Anyway, since no parameter is fixed (aircraft height, ground base stations antenna parameters,…) a statistical analysis estimating the probability of interferring has also been carried out.

This statistical analysis has been done to characterise the co-channel interference to terrestrial GSM 1800 networks based in worst case parameters, see signals schema in the figure below (left). The scenario taken into account considers real aircraft positions in an area of more than 140 Km2 around Frankfurt airport in a busy-traffic hour, considering that all aircraft have an on-board GSM system working. This is considered to be a realistic worst case scenario in terms of high aircraft population. Other parameters like BTS antenna sensitivity, tilt angle, transmit power, main beam width… are always taken in a worst-case basis. See figure below (right) for the relative position between aircraft and ground BTS.

(a) Uplink and downlink carrier and co-channel interference signals

(b) Relative position between aircraft and ground BTS

In a first step, the impact of all the parameters taken into account has been analysed individually, and afterwards, in a second step the cumulative density function of co-channel interference (Ic) and co-channel to interference ratio (C/Ic) of a user on the coverage edge of its cell is calculated for 16 scenarios with different parameters combinations results have been deeply analysed. The main output of this study is the following:

Assuming as a real case that ground BTSs use sectorised antennas and that each sector operates 4 channels, the worst result is obtained when 26 GSM channels[1] are available, and it yields to a probability of 0.154. That is 15.4% of the ground BTS in the area under study use the same channel as the BTS on-board. For bigger or smaller areas, if proportional to the one here studied and if all channel available are operated in the area, the results do not change. Under these conditions, since the ground BTS are placed uniformly, the size of the area under study does not play a role.

The conclusion is that in the worst case, an on-board wireless telephone system operating in Germany does not interfere the 85% of the ground cells. The interference of the remaining 15% of the cells is characterised by the cdf’s provided in step 2 of the study.


D (Km)

t (°)

v (°)


scn. 15





The figure below shows the cdf of Ic (left) and C/Ic (right) for scenario 15 of D12, see table, showed for two situations: (i) by considering all aircraft (blue lines), and (ii) by considering the aircraft flying above a certain threshold height (green lines).

To have no influence in GSM 1800 terrestrial networks, it would be desired that:

­          in Ic cdf plots: to keep the uplink interference always below the ground BTS sensitivity level and to keep the downlink interference always below the MS sensitivity level.

­          in C/Ic cdf plots: to keep uplink and downlink C/Ic levels above the reference C/Ic level.

In more than 99% of the time Ic is kept below the sensitivity levels, and downlink C/Ic is also kept above the reference level in more than 99% of the time. Only uplink C/Ic falls below the reference level during 45% of the time for users that are on the coverage edge of the cell. Therefore, the only consequence would be that the ground BTS can not keep with enough quality the connection with a MS attached to it, and therefore a handover would be performed.




[1] Minimum number of GSM channels operated by a mobile telephone operator in Germany.