A comparison of processing approaches for distributed radar sensing.
Doctoral thesis, UCL (University College London).
Radar networks received increasing attention in recent years as they can outperform single monostatic or bistatic systems. Further attention is being dedicated to these systems as an application of the MIMO concept, well know in communications for increasing the capacity of the channel and improving the overall quality of the connection. However, it is here shown that radar network can take advantage not only from the angular diversity in observing the target, but also from a variety of ways of processing the received signals. The number of devices comprising the network has also been taken into the analysis. Detection and false alarm are evaluated in noise only and clutter from a theoretical and simulated point of view. Particular attention is dedicated to the statistics behind the processing. Experiments have been performed to evaluate practical applications of the proposed processing approaches and to validate assumptions made in the theoretical analysis. In particular, the radar network used for gathering real data is made up of two transmitters and three receivers. More than two transmitters are well known to generate mutual interference and therefore require additional e�fforts to mitigate the system self-interference. However, this allowed studying aspects of multistatic clutter, such as correlation, which represent a first and novel insight in this topic. Moreover, two approaches for localizing targets have been developed. Whilst the first is a graphic approach, the second is hybrid numerical (partially decentralized, partially centralized) which is clearly shown to improve dramatically the single radar accuracy. Finally the e�ects of exchanging angular with frequency diversity are shown as well in some particular cases. This led to develop the Frequency MIMO and the Frequency Diverse Array, according to the separation of two consecutive frequencies. The latter is a brand new topic in technical literature, which is attracting the interest of the technical community because of its potential to generate range-dependant patterns. Both the latter systems can be used in radar-designing to improve the agility and the effciency of the radar.
|Title:||A comparison of processing approaches for distributed radar sensing|
|Open access status:||An open access version is available from UCL Discovery|
|UCL classification:||UCL > School of BEAMS > Faculty of Engineering Science > Electronic and Electrical Engineering|
Archive Staff Only