A Collaborative Technique for Spatial Interference Reduction in Multi-Node Antenna Arrays with Antenna Diversity
Kenneth L. Collier Jr. and Laurie Joiner, University of Alabama in Huntsville
The unmatched services provided by Global Navigation Satellite Systems (GNSS) such as the United States Global Positioning System (GPS), the European Union Galileo, and Russian Global Navigation Satellite System (GLONASS) have driven the worldwide pervasive use and reliance upon these systems by governments, commercial sectors and private citizens. As dependence upon GNSS for reliable and accurate position, velocity and time measurements increases it becomes more critical that the resiliency and integrity of these systems remain unfettered. GNSS receivers are vulnerable to interference whether from inculpable emitters or intentional sources commonly referred to as jammers. While illegal in the United States and many European nations, the availability and capability of jammers targeting commercial GPS receivers is alarming and has resulted in numerous studies to characterize their behavior and impact. Approaches to mitigate external interference in GPS receivers embedded in small, lightweight systems such as cellphones, autonomous drones, and other smart sensors are limited. Many practical single-antenna antijam algorithms are only applicable to a subset of interference waveforms with narrow frequency bandwidths and slowly changing temporal characteristics. Multi-element antenna arrays provide robust protection against a diverse set of threats; however, in their traditional implementation spacing between elements is fixed near one-half a wavelength. The size and weight of these antenna arrays are not practical for systems whose mobility and portability are distinguishing qualities.
A hybrid concept is devised to deliver robust spatial interference rejection to dispersed embedded GPS receivers working collaboratively in a multi-node antenna (MNA) array through the combination of adaptive null-steering and antenna diversity. The MNA array is formed from a collection of independent devices with single-element antennas thus circumventing SWAP constraints.
It is highly likely that forming a GNSS antenna array from a collection of independent dispersed antenna elements requires accuracy to be sacrificed for jamming-immunity. Unique complications are inherited that must either be accepted or mitigated. Foremost, the distance between individual elements of a practical MNA array will exceed traditional half-wavelength spacing. Thus, applying an adaptive null-steering filter to the MNA array results in aliasing where numerous grating nulls accompany the desired nulls present in the spatial antenna pattern. This severely restricts the useable beamspace of the array. A diversity scheme based on selection combining is postulated to overcome the effects of aliasing. Designating the total number of nodes available to participate in the array as N and the desired number of nodes forming the array as M, the diversity scheme performs adaptive filtering on X array combinations where X is found from the binomial coefficient of choosing M-1 nodes from a collection of N-1 nodes. Each resulting antenna pattern contains both nulls directed towards jammers and grating nulls; however, the grating null locations vary between combinations. Following code correlation in the GPS receiver, selection-combining is leveraged to select the output with the highest carrier to noise ratio per satellite. Simulation results show employing this diversity scheme can increase the useable beamspace from 50% to over 90% with N = 6 and M = 3.
An additional complication arises as the spatial size of the array increases. In traditional array processing the narrowband assumption is often applied such that elementary phase shifts are applied to a common signal to model its propagation across nodes of the array. As the bandwidth of the signal or expanse of the array increases, this approximation loses its validity and when left unmitigated can further restrict the useable beamspace and degrees-of-freedom within the MNA array.
Other difficulties introduced by the MNA array architecture include node synchronization; as well as, network bandwidth and latency requirements. Furthermore, maintaining the ability to accurately estimate the code delay and carrier-phase shift of the desired signal source within a GNSS receiver is more important than simply extracting navigation bits from the modulated waveform. Combining relative time-delayed samples from each antenna element artificially creates a multipath environment for the GPS receiver which can lead to inaccurate range estimates. The MNA array architecture should minimize this performance impact by compensating for the induced multipath environment.
This paper formalizes a concept to combine an adaptive, null-steering filter with selection-combining diversity to reject interference in MNA arrays. The tradeoffs and complications described above associated with large and sparse antenna arrays for GNSS applications are evaluated. Analyses reveal how the severity of these complications vary with MNA array composition and impact the fundamental ability of GNSS receivers to estimate code delay and carrier-phase shift in received satellite waveforms. An end-to-end MATLAB modeling environment is introduced to define MNA antenna arrays, propagate GPS and interference waveforms to array elements, excite adaptive-antenna array filters and assess GPS-receiver baseband processing performance. Finally, simulation results are presented to estimate the jamming immunity, useable beamspace, complexity, and navigation performance of the MNA array concept.