Measurement and Analysis of Clutter Signal from GSM/DCS-based Passive Radar Antonio De Maio*, Goffredo Foglia t , Nicola Pasquino+ and Michele Vadursi§ *Dept. of Electronics and Telecommunication Engineering - University of Naples Federico II Via Claudio 21,80125 Naples, Italy - Ph. +390817683147, Fax +390817683149, email: ademaio@unina.it tElettronica S.p.A. Via Tiburtina Valeria Ian 13.7,00131 Roma, Italy - Ph. +39064154581, Fax +39064154932, email: goffredo.foglia@elt.it +Dept. of Electrical Engineering - University of Naples Federico II Via Claudio 21,80125 Naples, Italy - Ph. +3908176.83630, Fax +390812396897, email: npasquin@unina.it §Department of Technologies - University of Naples "Parthenope" Centro Direzionale di Napoli Isola C4, 80143 Naples, Italy - Ph. +390815476791, email: vadursi@uniparthenope.it Abstract-Passive radars are not at all new to the radar community though they lately raised interest among researchers for some specific features that make them more appealing than classic active radar systems. In fact, having the TX and RX located at two different positions, and the adoption of signals carrying other kinds of information such as GSM (either classical or the 1800/1900 MHz version) or FM modulated carriers, makes passive radar virtually undetectable and economically advanta- geous. In the paper a measurement system used to characterize the clutter from a GSMlDCS-based passive radar is presented. The main objective is to study the statistical properties of the clutter signal and to find a theoretical model fitting the measured clutter data. I. INTRODUCTION Radars are classically based on a transmitter (TX) and re- ceiver (RX) that are physically close, or even consisting of one single equipment. On the contrary, a passive radar is basically a bistatic system [1], with the RX and TX located at different positions. Moreover the passive radar is characterized by the use of a transmitted signal originally devoted to different purposes, such as audio/video broadcasting [2] or to carry cellular traffic [3]. For this reason, the TX is usually said to be of opportunity. This approach to radar signal processing is not new at all. As a matter of fact, the first radar system tested back in 1935 was a passive radar system [4], though it was abandoned due to its continuous-wave nature that prevented range discrimination of the detected target. In recent years it has reencountered researchers' favor thanks to the wide spread of digital communication and to some peculiar characteristics and advantages, such as: • it doesn't require any specifically allotted bandwidth since it relies upon other transmitting systems; • since it has no proprietary TX, it presents strong econom- ical advantages over its classical counterpart; • being the receiving station physically separated makes radar countermeasures difficult, if not impossible unless some a priori information are available. Current research on passive radar is focused on verifying several characteristic of the used waveforms [5], [6], and its detection behavior [3], [7], though a thorough study on the 1-4244-1539-X108/$25.00 ©2008 IEEE clutter statistical characteristics at the receiver, with reference to the GSM/DCS waveform, has not been carried out yet. This is indeed paramount to correctly design the detector and characterize passive radar's performance and it is the main obj ective of this paper. The paper is organized as follows: in Section II the measurement system and the experimental setup will be outlined; in Section III the results of the experiments and the statistical analysis will be presented, and finally in Section IV the conclusions and proposal for future work will be given. II. MEASUREMENT SETUP The system used to measure clutter signal consists of a radiofrequency (RF) RX connected through a calibrated cable with negligible attenuation to a hom antenna, pointed at different azimuth angles for each dataset, so to have clutter returns from different directions of the measurement site. Analysis over several dataset has been performed showing comparable results, and for this reason in the following only the results from one dataset are presented. A. Measurement site The TX of opportunity is one sector of a GSM/DCS-1800 base station (BS) located on a 30m high tower. Its carrier frequency is 1853.6 MHz. The baseline distance between the receiving antenna and the TX is about 90 m. The most relevant reflecting objects around the TX are: i) a row of pine trees along the SE-NW direction, south-west of the TX; ii) a very busy motorway in the NS direction, east of the TX. B. Antenna The hom antenna is a BBHA 9120-D by Schwarzbeck, working in the nominal frequency range 1 to 18 GHz. The gain is 10.45 dBi and the 3 dB beamwidth is slightly less than 60° in the E-plane and about 45° in the H-plane. The front-to- back ratio is greater than 25 dB for f > 1.3 GHz. The antenna weighs 1.3 kg, the aperture's dimensions are 245 x 142 mm 2 and its total length is 408 mm. Authorized licensed use limited to: Universita degli Studi di Napoli. Downloaded on February 26, 2009 at 05:46 from IEEE Xplore. Restrictions apply.