Previous Abstract Return to Session B12 Next Abstract

Session B12: GNSS Resilience to Interference, Jamming, and Spoofing

Model and Observation of the Impact of JTIDS/MIDS on GNSS C/N0 Degradation
Axel Garcia-Pena, Christophe Macabiau, ENAC, France; Mikael Mabilleau, Pierre Durel, GSA, France
Alternate Number 1

Global Navigation Satellite System (GNSS) received signals processing can be affected by received additive signals such as noise, multipath and interference. Radio Frequency Interference (RFI) sources are of various sorts and their nature and impact depends on the user application. In the context of civil aviation, it is important to identify and characterize the radio frequency interference relevant to the airborne GNSS receivers processing signals in the L1/E1 and L5/E5a bands, to determine the vulnerability of airborne GNSS receivers in L1/E1 and L5/E5a equipped with their relevant antenna, to issue minimum requirements on these L1/E1 and L5/E5a antennas, as well as to issue minimum requirements to be imposed to airborne GNSS receivers operating at L1/E1 and L5/E5a bands. A long thread of activities led to the elaboration of various ICAO, RTCA and EUROCAE standards considering RFI. Currently, [RTCA DO292, 2004] reflecting the relevant interference to L5/E5a is being updated to incorporate the evolutions of the RFI environment defined by DME/TACAN, JTIDS/MIDS, LDACS, SSR equipment and other GNSS systems operating at these bands, as well as the usage of this L5/E5a band for GALILEO E5a and SBAS L5/E5a datalink airborne signal processing. In addition, ICAO RFI mask of GNSS L5/E5a is now under definition. These elements will then complement the current ICAO SARPs, draft EUROCAE and RTCA MOPS for GNSS L5/E5a airborne receivers.
In the course of the elaboration of the update of [RTCA DO292, 2004], it has been proposed to revisit several elements of the worst-case scenario link budget analysis in order to consolidate the overall link budget margin. This was deemed necessary since the link budget margin is expected to be very small. Among the axes of revision are:
- the analytical model representing the effect of the AGC/ADC and temporal blanker
- the DME/TACAN environment and its impact on minimum operational/system performance requirements for a GNSS L5/E5a receiver
- the JTIDS/MIDS environment and its impact on minimum operational/system performance requirements for a GNSS L5/E5a receiver
- The consideration of SSR, systems operating within the Aeronautical Mobile (en-route) Service [(AM(R)S] to include LDACS and RPAS (Remotely Piloted Aircraft System) Command and Control (C2) Data link, and commercial systems that are authorized in the band such as commercial PMSE (Programme Making and Special Events) equipment.
This article specifically looks at the consolidation of the model of the effect of pulsed interference on an airborne GNSS receiver.
The RFI impact on a GNSS receiver in civil aviation is usually modelled as the overall carrier to receiver thermal noise density (C/N0) degradation observed at the receiver’s correlator output, or equivalently, as an increase of the effective receiver noise density N0 with added interference denoted as N_(0,eff). Therefore, a decrease of the minimum available C/N0, derived from the link budget and from the N_(0,eff) calculation, implies a reduction of the C/N0 margin between the minimum available C/N0, eff and the different L5/E5a GNSS and SBAS signal processing, acquisition, tracking, demodulation, C/N0 threshold values. Concerning the revisit of several elements determining the C/N0 margin, first, the model for the GNSS airborne receiver RF processing chain, namely the model for AGC/ADC and blanker is reviewed. In particular, the model for blanking function has gone under new scrutiny, with the prospect of the definition of a minimum blanker model. Second, the DME/TACAN environment is being reviewed. Models of impact of DME/TACAN on C/N0 degradation are also revised. Next, the JTIDS/MIDS environment will be re-assessed, and the relevant models updated. Systems operating within the [(AM(R)S] to include LDACS and other datalinks and commercial systems that gain authorization in the band will also need inspection.

Objectives, key innovations and significance of the paper
Traditionally, the countermeasure adopted against pulse interference which is analyzed in civil aviation is the temporal domain pulse blanking method as described in [RTCA DO292, 2004]. Temporal domain blanking method is easy to implement and computationally efficient. It can thus be considered as representative of what could be implemented in a minimum airborne receiver.
The expression of effective N_0 is given below:
N_(0,eff)=N_0/(1-bdc)*(1+I_(0,WB)/N_0 +R_I ) (1)
where I_(0,WB) are all the wideband (non-pulsed) continuous RFI contributions (usually the other GNSS signals falling in the L5/E5a band). Setting the blanking threshold correctly can be challenging due to the trade-off between the Blanking Duty Cycle, bdc, (percentage of samples set to zero by the blanker) and the R_i (the below-blanker interfering-signal-to-thermal-noise ratio) parameters since both of them directly impact the effective noise, N_0, of the received signal after blanking. On one hand, a low threshold removes the majority of the signal samples containing interference (reduction of R_I) but a higher percentage of time the noise alone is enough to trigger the zero-setting process causing a “false alarm” (increase of bdc). On the other hand, a high threshold value decreases the “false alarm” events (decrease of bdc) but also does not appropriately suppresses the interference term (increase of R_I). Proper threshold selection is thus a crucial factor of performance in such blanking methods.
In this paper, the model and observation of the impact of JTIDS/MIDS on GNSS C/N0 degradation as a RF pulsed interference is specifically tackled. JTIDS/MIDS stands for Joint Tactical Information Distribution System / Multifunctional Information Distribution System and is a military aeronautical digital tactical communication, navigation and identification system which is operated on land, sea and airborne platforms in many countries world wide including the US, and countries within Europe and Asia. JTIDS/MIDS, also known as Link 16, is a hybrid direct sequence and frequency hopping spread-spectrum system that operates on 51 different carrier frequencies in the frequency bands of 969 – 1008 MHz, 1053 – 1065 MHz and 1113 – 1206 MHz. In particular, it operates in the region surrounding and including the L5/E5a band. Recently, a remap capability has been implemented where it would have the capability to operate on as few as 37 carrier frequencies which could result in added pulse density within the L5/E5a band when compared to the 51 carrier case. Moreover, JTIDS/MIDS employs TDMA to accommodate multiple users in a network with 128 timeslots per second. Each time slot lasts 7.8125ms with transmission message intervals of 929µs, 3.354ms or 5.77ms within assigned time slots. The information transmission period is constituted of 72, 258 or 444 pulses for the transmission periods respectively. The 6.4µs pulses are formed by 32 chips of 200ns. From this previous description, the pulsed nature of the JTIDS/MIDS signal is clearly observed.
The review of the JTIDS/MIDS signal impact analysis on the C/N0 degradation with respect to the analysis presented in [RTCA DO292] is of interest for several reasons. First of all, in [1], the proposal of a new formula for the modeling of the pulsed interference R_I term was made. This formula should thus be compared to the one presented in [RTCA DO292] for a JTIDS/MIDS signal for different values of GNSS receiver bandwidths (e.g. 24, 20, 18, 15 and 12 MHz) and blanker thresholds (e.g. -120 and -121 dBW). In this regard, the conducted analysis has shown a very good match of both formulas for the nominal case in [RTCA DO292] (Rx bandwidth equal to 20MHz and threshold equal to -120dBW) but some differences are observed for other values (e.g. 12 MHz and -121 dBW). Additionally, the application of [RTCA DO292] formula is not completely straightforward from the definition given in the MOPS and thus some clarifications are given in this paper: a weighting of the given formula by cardinal sinusoidal (width determined by the L5 signal chip duration) is required.
Second, the validation of the two previous R_I formulas must be conducted. This disclosed validation is difficult to obtain since RF samples of the signal cannot be obtained by unauthorized users and the exact mathematical definition of the signal is not publicly released. Therefore, an equivalent mathematical model (waveform, time structure, etc.) from the RFI impact on C/N0 degradation point of view is presented in this article in order to allow the generation of an equivalent JTIDS/MIDS signal to conduct C/N0 degradation simulations. The signal mathematical model is compliant with the public release information such as the type of modulation, the power spectrum density, etc., and should be acceptable for standardization purposes. From the derived equivalent mathematical model, the two previous formulas, [1] and [RTCA DO292], have been tested and while [1] shows a very good match for all receiver values, the [RTCA DO292] formula shows some difference for non-nominal values.
Third, the degradation of the equivalent C/N0 at the GNSS receiver correlator output will depend on the number of JTIDS/MIDS transmitters, their peak power, their Time Slot Duty Factor (TSDF), their geographical situation with respect to the victim GNSS receiver and the network for each modelled user. All these elements define the interfering scenario. In [RTCA DO292], different scenarios are defined following the Geographical Area (GA) methodology used for managing permissible JTIDS/MIDS TDSF levels with the most constraining scenario being the case 8 scenario. However, the interfering scenarios depend on the agreements for the JTIDS/MIDS TSDF limits between the department of transportation and the department of defense of each country. For example, within the US the GA is equivalent to the case 8, while for many European countries another methodology called an Any Point In Space (APIS) GA method is used. Similar to the case 8 scenario used in [RTCA DO292] the APIS GA methodology is modelled conservatively for the C/No degradation analyses. In this article, the worst case scenario which can be derived from the APIS methodology will be presented; this scenario will be acceptable for standardization purposes. For both the case 8 and APIS scenarios, the total C/N0 degradation have been calculated for several receiver configurations, for both theoretical formulas and for simulated signals. The results showed that case 8 scenario introduced a smaller degradation than the APIS scenario; in the case of nominal receiver values, case 8 scenario has a degradation 1.8dB while 50NM APIS has a degradation of 2.9dB. Moreover, it was observed that the degradation can be further reduced to 1dB and to 1.8dB respectively by decreasing the receiver bandwidth and by decreasing the blanker threshold. An analysis to clearly justify why the threshold value decrease improves the final C/N0 degradation will be shown. The final observation which was made in this part is that while the theoretical formulas provide a very good match for the C/N0 degradation results, the intermediate values, Bdc and RI, deviate from the simulated ones for the -121dBW case (easily observed in the 50NM APIS scenario). This last observation makes the case of study of the last and final point.
Fourth and last, the C/N0 degradation formula presented previously in this abstract from [RTCA DO292] makes some assumptions about the pulse collisions between the different interference sources. One of these assumptions is that the triggering of the threshold by an interfering source x is going to always affect the other interfering sources; and thus, that the power of the other interfering sources should be decreased by the Bdc associated to the x interfering source, Bdcx. However, this assumption as well as other implicit assumptions may and may not be fulfilled by the real scenarios. For example, the previous results presented for JTIDS/MIDS case 8 assume that the pulses from all modelled users arrive at the receiver at different times. This assumption implies that the collision between JTIDS/MIDS users may never happen, which is not realistic. Therefore, in this paper, the collision of the JTIDS/MIDS signals for the different scenarios will be analyzed and the general C/N0 degradation formula will be updated.
To summarize, the specific objectives of this paper can be listed as follows:
1- To introduce an equivalent mathematical model for the JTIDS/MIDS signal generation from a C/N0 degradation point of view for standardization purposes.
2- To compare and to clarify the RI theoretical formulas of [1] and [RTCA DO292] for different receiver parameters.
3- To validate the RI theoretical formulas of [1] and [RTCA DO292] with respect to the JTIDS/MIDS simulated signal results.
4- To define the case 8 and 50NM APIS JTIDS/MIDS signal interfering scenarios for standardization purposes
5- To determine the equivalent C/N0 degradation for the case 8 and 50NM APIS geographic area modelled scenarios and for the different receiver parameters.
6- To determine and to model the influence of the JTIDS/MIDS pulse collisions on the C/N0 degradation formula

[1] A.Garcia-Pena C. Macabiau, M. Mabilleau et P; Durel, "GNSS C/N0 Degradation Model in Presence of Continuous Wave and Pulsed Interference", ION ITM 2020
[RTCA DO292] Assessment of RF interference relevant to the GNSS L5-E5a band, RTCA, DO 292, July 29, 2004

Previous Abstract Return to Session B12 Next Abstract