PhD position: Direction of Arrival (DOA) estimation techniques for 3-Dimensional phased arrays

Updated: almost 2 years ago
Location: Shrivenham, ENGLAND
Job Type: FullTime
Deadline: 08 Apr 2022

Context and aims

Existing missile Radio Frequency (RF) seekers use a mechanical rotating antenna to steer the radiating beam in the direction of a target. Latest collaborative research between Cranfield University, the French LAB-STICC and Thales Optronique has been investigating the replacement of mechanical antenna components of the RF-seeker with a novel 3-Dimenstional (3D) antenna array conformal to the missile shape that can steer the beam electronically.

3D conformal antennas may offer significant advantages with respect to 2D mechanical antennas, such as a much faster beamsteering and a better coverage but, at the same time, introduce new technical challenges resulting from a much more complex interference pattern between the array radiating elements. Firstly, a 3D phased array consists of antenna elements whose phase centres are arranged on a 3D coordinate system and, as result, most of the signal processing solutions developed for 1D or 2D antennas cannot be directly applied. Secondly, unlike existing 2D mechanical or electronic solutions, the antenna elements forming the 3D array are directive elements arranged with different orientations and, therefore, each element contributes to the total received radar signal with a different response to a given polarisation.

One of the key tasks of a missile RF seeker is the ability to measure the Direction Of Arrival (DOA) of target echoes. Our previous collaborative research has investigated and characterised the DOA accuracy of 3D phased arrays consisting of directive elements through a study that has resulted in the analytical expressions of the lower bound for the variance of unbiased DOA estimators for the one-source scenario [1]. However, despite providing an understanding of the array geometrical characteristics that affect the DOA estimation performance and an insight on what antenna design parameters can improve it, the lower bound does not directly indicate how to estimate the DOA of a target (neither it indicates how far the performance of an unbiased estimator are from the optimal).

A recent study at Cranfield University (ref: P10139) has investigated the performance of the Maximum-Likelihood (ML) estimator with Matlab simulations alone. This background work has clearly shown that finding the estimates of the DOA, using 3D arrays, requires to minimise a complex multi-dimensional likelihood function and therefore requires a significant computational load. This computational cost can become prohibitive for the multiple sources scenario and for real-time applications. To solve this problem, a study of the analytical expressions of the estimator and the development of suitable low-computational approximations for the estimator is an imperative. For example, these approximations could include the extension of the monopulse technique, subspace approaches (MUSIC/ESPRIT), sparse reconstruction algorithms to the 3D case (Lasso) or deep learning-based approaches. The proposed research is novel with very little, if any, published works in the existing literature.

The main objectives of this research are to 1) develop low-computational DOA estimation techniques for 3D conformal arrays of directive elements, 2) validate the signal model with experimental signals, and 2) assess their performance with a set of experimental trials using a unique bespoke 3D active phased array prototype developed at LAB-STICC.

The technical objectives of this research proposal have been defined as a result of our successful and long-standing research collaboration with Thales Optronique and the French LAB-STIC. To date this collaboration has delivered two successful MCM-ITP projects and 3 PhD student programmes (one completed and two in near completion) [2][3][4].

Cranfield University has also recently been contracted by our French partners (ref: P13268) to support the development of 3D phased array at LAB-STICC. The subcontract will run in parallel with this proposed PhD programme so that the PhD student activities can inform and complement ongoing cutting-edge research defence activities and, at the same time, benefit from additional Travel & Subsistence (T&S) funding for regular meetings in France.

The student will benefit of unique opportunities for dissemination of the results with participation to regular Thales workshops, meetings at LAB-STICC and, whenever possible, international IEEE radar conferences. Based on the novelty of the proposed research there is a strong expectation for this research to deliver peer-reviewed publications in IFQ1 journals.

References

    1. Fourtinon L, Balleri A, Quere Y, Person C, Lesueur G and T Merlet, “Direction of Arrival Estimation with 3D Conformal Polarised Arrays”, in submission to IEEE Transactions on Antennas and Propagation.

    2. Kocjancic L, Balleri A and Merlet T, “Multibeam radar based on linear frequency modulated waveform diversity”, IET Radar, Sonar and Navigation, vol. 12, no. 11, pp. 1320-1329, 2018.

    3. Fourtinon L et al., “Directivity and ellipticity study for planar and 3D conformal RF-seeker antennas”, 2017 IEEE Conference on Antenna Measurements & Applications (CAMA), Tsukuba, Japan, 2017.

    4. Abedrrabba S, Ahiska K, Allanic R, Balleri A, Chatterjee M, Merlet T and Quendo C, “DRAGON: Adaptive RF Seekers based on 3D Conformal Antennas”, in International Radar Conference 2019, Toulon, France, 2019.