IEEE - Aerospace and Electronic Systems - September 2019 - 51

Angkasa et al.

Figure 7.

Figure 6.
Integer delay estimation between each reference component code
Cn and incoming soft chips fnk g. The reference component code
Cn is correlated against the incoming chips at all its possible
delays of n (see Table 5 of Appendix). The delay that yields the
maximum correlation score is taken to be the correct delay
between subcode Cn and incoming composite code with soft
chips fnk g.

Finally, given that the timing subsystem was able
to estimate the fractional delay of the chips effectively,
it is still necessary to obtain the integer delay of
the chip sequence for regeneration. The traditional
ranging codes, which contain a strong Æ1 chip pattern
in the code, are not amiable to recovery of the integer
delay by simple cross correlation of the composite
code. This is because the range-clock portion of the
code causes the composite code to have a very poor
cross-correlation performance. Instead, we take another
approach of delay estimation by obtaining the integer
delay of the composite code to each of the six reference components codes first, and computing the composite code delay accordingly. This computation relies
on the Chinese remainder theorem approach outlined
in Section III.d [12], and that we review below.
The Iris implementation first generates reference
component codes for components C1-C6 (respective
chip periods which are listed in Table 5 in Appendix)
as well as the target composite sequence. However,
these codes do not have a common time reference with
the incoming ranging code. To align the generated
chips with the incoming chips, the Iris implementation
correlates each component code to the incoming softchip sequence, as illustrated in Figure 6 for a single
component code.
The result of each subcode correlation yields a relative ranking of each possible delay for that subcode
to align well to the incoming soft chips (polarity of
the soft chips is also corrected at this stage to
ensure positive correlation scores). Observe that this
does not immediately yield the composite code delay,
but only the delay of each component code to the
incoming composite code. In order to obtain the reference composite code delay to accurately align to the
incoming composite code, it is possible to convert
the individual subcode delays estimated through the
SEPTEMBER 2019

Regenerative ranging system. The incoming ranging signal (quadrature output of the CTL) is used to estimate the fractional delay
of the incoming chips by tracking the range clock. The fractional
delay is then used to optimally sample the incoming chips.
The incoming soft chips are then used to estimate the integer
delay between the incoming composite code and the regenerated
reference composite code onboard the Iris. A variable delay block
then delays the regenerated code according to this delay.

structure illustrated in Figure 6 to the composite code
delay via
u¼

6
X

an un mod N

(4)

n¼1

where u denotes the integer delay between the reference
composite code and the incoming composite code, un
denotes the delay of component code n to the incoming composite code (obtained through the structure in Figure 6), the
coefficients fan g are listed in Table 3, and N ¼ 1;009;470
denotes the code period of the composite ranging code in
chips (valid for both CCSDS T2B and T4B codes).
Once the integer delay u is estimated, we utilize a variable-delay operation to delay the onboard generated PN
code by the estimated delay u in (4) and output a clean
version of the PN code. In addition, the sampling trigger
produced by the timing-recovery subsystem also serves as
the trigger for the next clean chip generation, thereby
matching any fractional timing drifts between the incoming sequence and the outgoing sequence (up to a fixed
measurable delay through the Iris radio).
Putting the subsystems illustrated in Figures 4-6
together, we obtain the complete regenerative ranging
subsystem illustrated in Figure 7. The full implementation
is modeled in fixed point by implementing the algorithm
in MathWorks Simulink as block diagrams and automatically generating the necessary hardware definition
language code for each individual block. The generated
code is then integrated in the Iris transponder firmware.

MISSION APPLICATION (GALILEO SCENARIO)
When the Galileo spacecraft approached Jupiter orbit
insertion and also during Jupiter orbit (1995-2003), no
ranging was done. Early in the mission, the Galileo highgain antenna failed to unfurl correctly, and radio contact
therefore was only possible using the low-gain antenna
with S-band links. As a result, the uplink and downlink
signal levels were much smaller than planned. One of the

IEEE A&E SYSTEMS MAGAZINE

IEEE - Aerospace and Electronic Systems - September 2019

Table of Contents for the Digital Edition of IEEE - Aerospace and Electronic Systems - September 2019