STAP as a solution for imperfections in Multi-Antenna GNSS Receivers

Kurzfassung

GNSS has become the main technology to provide position and timing services in the world. The increasing dependency of the provided services has raised concerns about the vulnerabilities of the systems. Since GNSS signals travel long distances from the satellites and are emitted with relatively low power, the GNSS signals are very weak when they arrive at the receiver. Thus the signals can be easily blocked by interfering radio transmissions. The precision provided by current generation GNSS receivers together with the use of sophisticated processing methods leave multipath reception and radio frequency interference as the dominant remaining error sources affecting GNSS performance. These radio frequency interferences are usually of unintentional nature, but they can also be deliberately produced. Unintentional and deliberate interference signals constitute a challenging problem in Safety of Life GNSS applications. In order to tackle the issue several techniques have been proposed. One of the most powerful technologies arises from the use of antennas formed of several receiving elements. The antenna arrays possess very interesting capabilities, like Null-Steering or Beamforming. Such capabilities, called Spatial Signal Processing algorithms, enable the cancellation or nulling of several interference signals. Loosely speaking the ability of the antenna array to fight interferences depends on the number of elements that form the array. In theory with M elements M-1 possible interferences can be attenuated. Yet there are limitations to the number of antenna elements in the array. Relevant factors are cost and complexity. Antenna design complexity increases with the number of elements and so does the cost. Not only the antenna per se, but also the connectors, cabling, amplifiers, filters, number of analogue to digital converters (ADC), etc., increase. In addition to that, given a limited physical space, placing several antenna elements close to each other could lead to undesired interactions that will degrade the performance. With those ideas in mind, a complementary approach was introduced. By combining Time and Spatial processing it is possible to increase the number of suppressed interference signals without the need of extra elements in the array. This technique based on the use of the temporal and spatial domains is known as Spatial Temporal Adaptive Processing or STAP. In the last years, with the trend towards software defined radio, the analogue frontend part has taken more and more a backseat in recent research for GNSS receivers. However in the presence of radio frequency interference, an adequate design of the analogue frontend stills plays a crucial role and primarily influences the achievable interference robustness of a GNSS receiver. Any error in the design which causes an information loss can never be entirely recovered by digital processing. In the case of multi-antenna GNSS receivers, the analogue frontend properties are even more crucial for the overall system performance and, therefore, the achievable interference robustness. In the modern days most array processing techniques are performed in the digital domain. As the name indicates it requires a digitalization process that is carried out by analogue to digital converters (ADCs). There are several types of ADCs based on the particular way that they are implemented. Some literature has been written about how several types of ADCs affect the performance of the GNSS receivers but not much has been said about the impact on the previously mentioned array digital processing techniques. Referring to STAP techniques, it is commonly assumed that the ADC provides time uncorrelated samples in an optimum interference free scenario. Many array digital processing techniques, not only STAP but also, for example, Beamforming, Direction of Arrival estimation, etc. assume that the so called narrow band signal assumption is fulfilled. Moreover, at least almost identical spectral characteristics of the different channels are often assumed. That means all channels are assumed to behave exactly the same. However, both assumptions are generally not fulfilled. The potential of STAP to cope with the effects caused by the narrow assumption and the imperfections of the hardware is investigated. This is performed with the help of extensive numerical simulations under several interferences scenarios.