L-Band Digital Aeronautical Communication System (LDACS)-Based Non-Cooperative Passive Multistatic Radar for Civil Aviation Surveillance

Kurzfassung

In this thesis, the feasibility of setting up a non-cooperative passive multistatic radar system for detecting aircraft while using communication signals as signals of opportunity is investigated. The motivation for this work is the need to set up a backup non-cooperative surveillance system for civil aviation. To this end, the use of the future L-band digital aeronautical communication system (LDACS) would ensure an appropriate re-use of the already allocated spectrum in the aeronautical radionavigation band. The aim of this work is to assess the suitability and limitations of the LDACS system and signals for radar purposes. Particularly critical in this respect are: i) the low transmit power of the LDACS system and ii) the fact that the LDACS signals are not optimized for radar purposes. In the beginning of the thesis, the envisioned LDACS-based surveillance parameters are introduced and the bistatic and multistatic radar coverage and detection performance are evaluated. The ambiguity function (AF) of the orthogonal frequency division multiplexing (OFDM)-based LDACS communication signals is then studied. Closed-form expressions for the AF are derived and account for the different LDACS signal contributions. This analytical representation enables a direct assessment of the signal ambiguities while also allowing for the multistatic radar AF evaluation. Furthermore, the modified Cramér-Rao bound on the joint estimation of target location and velocity is derived. Closed-form expressions for the signal-dependent elements of the modified Fisher information matrix are obtained and the estimation accuracy performance is analyzed for different antenna configurations, integration times, and signal-to-noise ratio (SNR) conditions. Motivated by the need to improve the SNR of the reflected LDACS signals, the last part of this thesis is dedicated to assessing the limits on the achievable coherent processing gain. The integration time is generally limited by the target movement which leads to changes of both the observed target range and Doppler shift and prevents a coherent SNR increase when traditional cross-correlation processing techniques are employed. Therefore, a more involved approach combining keystone transform and fractional Fourier transform is implemented to compensate for both range and Doppler migration. This approach is however also shown to have limitations, in particular for low SNR and multi-target scenarios. To address these shortcomings, a novel framework is developed to perform target detection and parameter estimation jointly while using a super-resolution sparse Bayesian learning (SBL) approach. This technique uses an observation model which accurately accounts for the underlying range and Doppler migration effects. Moreover, since the search space is not discretized the scheme provides super-resolution estimation capabilities. The simulation experiments demonstrate the effectiveness of the newly proposed approach and the advantages compared to the state-of-the-art compensation approach.