The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part Bl. Beijing 2008 
1022 
offers two significant advantages over current spacebome 
/airborne radar systems. The first advantage is persistent region 
coverage, because the passive near-space platforms can provide 
stay-and-stare persistence for days, months, and even years. 
The second advantage is the potentials of bistatic radar 
configuration. 
Thus, the detectable maximum range can be expressed as 
R = In a ° Ä2 ° r G ' p 
,nax y ° An AnKT 0 B w SNR miB 
Although the receiver is stationary, an aperture synthesis can 
still achieved by the motion of GNSS satellites (He, et al., 
2005). As SAR (synthetic aperture radar) images are usually 
derived through correlation of the raw data with a two- 
dimensional reference function. Here the range resolution is 
achieved by measuring the correlation power between x d (t) 
and x r (t). 
Similarly, azimuth resolution is obtained by exploiting the 
relative motion between the GNSS transmitter and the target, 
which leads to the target returns having a Doppler bandwidth. 
Additionally, the configuration using a single receiver is also 
feasible. In this case, the received signal would contain the 
signals from both the direct-path channel and the scattered 
channel. Once they are separated, successful matched filtering 
can still be achieved. 
and the maximum sensitivity is determined by 
Snr = *Wo 
(5) 
where y 0 is a constant parameter, and r p is the coherent 
integration time. 
As an example, assuming the following parameters: n o = -130 
dBW/m 2 , cr 0 = 20m 2 , A = 0.19m A = 0.19m, G r =36dB , 
T o =300 , B w = 2.4MHz , G sp =65dB , SNR miD =0dB , then 
R^ is found to be 24.01km. This value validates that the 
requirement of power budget is satiable. Notice that R imx can be 
further increased by increasing the effective area of the receiver 
antenna. 
3. FEASIBILITY ANALYSIS 
Using GNSS transmitter as an illuminator for near-space 
passive remote sensing presents a problem of signal 
detectability because the received signal will be very weak. As 
such, signal detectability is investigated in this section. 
3.1 Detectable Range 
The power of the target reflection available at the near-space 
receiver antenna output is determined by 
^ P,G, ° 0 ¿G r 
AnR 2 AnR 2 An 
(1) 
where P t is the transmitted power, G t is the transmitter antenna 
gain, R, is the transmitter-to-target distance, <j 0 is the radar 
cross section (RCS), R r is the target-to-receiver distance, A is 
the wave-length and G r is the receiver antenna gain, 
respectively. The power flux density near the Earth’s surface 
produced by GNSS can be assumed to be P,G, / AnR 2 = n o 
«3-10“ 14 Wt / m 2 . The noise level at the output of the RF 
front-end can be represented by 
N n = KFB,. 
(2) 
3.2 Signal-to-Noise Ratio 
For radar image formation, the total data samples are processed 
coherently to produce a single image resolution cell. The 
thermal noise samples can be taken as independent from sample 
to sample within each pulse, and from pulse to pulse. After 
coherent range and azimuth compression, the image signal-to- 
noise ratio (SNR) is given as (He, et al., 2005) 
1 
R. 
AnR; KT 0 F n v s -p a 
■V 
(6) 
where A r , F n , v s and 77 are the effective area of the receiver 
antenna, noise figure, satellite velocity and loss factor, 
respectively. As the potential azimuth resolution can be 
rewritten as 
AR, 
v7. 
(7) 
with T s the integration time, we can get 
SNR,. 
n o^o TsO 
AnR 2 F n KT 0 
(8) 
where K , T 0 and B w are the Boltzmann constant, system 
noise temperature and noise bandwidth, respectively. Hence, 
with the consideration of processing gain G sp , the signal-to- 
noise ratio 
depends on 
SNR =-^ = n n 
N n 
A 2 G. 
AnR 2 An KFB,. 
(3) 
As an example, assuming a typical system with the following 
parameters: a 0 = 20m 2 , T s = 1000s , T 0 - 300 , tj = 0.5 q, and 
F n = 2dB , then the calculated SNR is illustrated in Fig. 1. Here 
the SNR is favourable owing to an essentially long integration 
time. Note that this SNR can be further improved by using non 
coherent integration of signals from more than one receiver 
channel.