The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part Bl. Beijing 2008 
Figure 4: Impact of phase synchronization errors 
Figure 5: Block scheme of signal processing algorithm. 
5.2 Bistatic Radar Processing 
f 
\ 
Another challenge is bistatic radar processing. As a starting 
point, most radar imaging algorithms are assumed that the 
transmitted signal is a chirp signal, this is not the case for a 
GNSS signal. Moreover, as the GNSS transmitter follows a 
rectilinear trajectory, while the receiver remains in a fixed 
position on a near-space platform looking down to the 
illuminated scene. This configuration with fixed receiver also 
brings some challenges on image formation algorithms. The 
instantaneous slant range of a given target as a function of its 
geometry is 
* K, ) = H+v 2 •(<„-»* ) 2 + •.4 (13) 
(15) 
where k r , A, f a , v, R(f a ;R l0 ) and c denote the range chirp 
rate, signal wavelength, instantaneous Doppler frequency, 
GNSS transmitter velocity, range in range-Doppler domain and 
light speed, respectively. After range chirp scaling, a range 
processing factor 
where and t dc are the azimuth time and the time while the 
target at the beam center crossing, respectively. R l0 is the 
slant range of closest approach to the transmitter, and RrO that 
to the receiver which moves in a path parallel to the transmitter. 
^(/ r ./ a ;^) =ex P i 
xRjiflf' L nf 2 
CV 1 o). 
(16) 
It can be noticed that the range history of a given point target 
does not depend any more only on the zero-Doppler distance 
and the relative distance from the target to the transmitter, but 
also on the absolute distance to the receiver. But the distance to 
the receiver is only a function of the coordinates R r0 and t dc , 
and independent of the varying variable t m . Then the Doppler 
chirp rate K d can be derived from the instantaneous slant range 
as 
is applied to the two-dimensional frequency-domain signal, 
which completes the range compression and bulk RCM 
correction. This makes the Doppler chirp rate changes with the 
azimuth position in the same range gate. 
To perturb their Doppler chirp rate to be the same, we use one 
perturbation factor (Wong and Yeo, 2001) 
¿/ 2 [/?(t Œ )/à] 
(14) 
tf 3 (0 =ex p 
, v a, 4 a, 6 a, g 
jn —T* +— +—r* 
A6 15 28 
(17) 
Hence the azimuth phase modulation of a target is due to the 
motion of the transmitter alone. As a consequence, there is a 
range ambiguity that does not exist in the monostatic case, i.e., 
two or more targets located at different positions can have the 
same range delay at zero-Doppler but will have different range 
histories. This complex signal model represents a great 
challenge towards the development of precise and efficient 
focusing algorithms. One effective algorithm is non-linear chirp 
scaling (NLCS), as shown in Fig. 5. //, is a chirp scaling factor 
which makes the range cell migration (RCM) of all targets 
along the swath be the same, can be expressed as 
where a x , a 2 and a 3 are constant coefficients. Finally, the 
azimuth Doppler chirp rate corrected azimuth-time domain 
signal can be azimuth compressed with the following reference 
function 
#4 (/<,;*,o) = e x P 
(18) 
In this way, focused radar image can be achieved. However, 
this situation will become more complicated for unflat DEM