111 lines
4.0 KiB
Markdown
111 lines
4.0 KiB
Markdown
---
|
|
title: Doppler shift and range rate
|
|
tags: doppler, frequency-offset, clock, ranging, tracking
|
|
summary: How the motion of a satellite changes the frequency it appears to be received at, and how that shift is turned into a position.
|
|
---
|
|
|
|
# Doppler shift and range rate
|
|
|
|
A satellite is observed at a frequency that differs from the one it
|
|
transmits, because it is moving. That difference is the principal observable
|
|
in a tracking system, and turning it into a position is the whole of
|
|
ranging.
|
|
|
|
## Definition
|
|
|
|
**Doppler shift** is the change in observed frequency caused by relative
|
|
motion along the line of sight. **Range rate** is the rate of change of
|
|
range, and it is the same physical quantity: one is a frequency, the other a
|
|
velocity.
|
|
|
|
For a satellite moving at `v_r` radially, at range `R`, the received
|
|
frequency shift is
|
|
|
|
```text
|
|
delta_f / f = -v_r / c
|
|
```
|
|
|
|
approaching raises the frequency. The non-relativistic form is adequate for
|
|
orbital speeds; the relativistic correction is about one part in `10^13` and
|
|
matters only for a clock discipline experiment.
|
|
|
|
## Mathematics
|
|
|
|
### The two-way shift
|
|
|
|
A signal that travels to the satellite and back is shifted twice, so the
|
|
observed fractional shift is
|
|
|
|
```text
|
|
delta_f / f = -2 * v_r / c
|
|
```
|
|
|
|
with `v_r` the range rate. The factor of two is why the same satellite
|
|
appears to shift twice as much when tracked in two-way mode, and a range
|
|
model that ignores it has a factor-of-two error that is easy to miss.
|
|
|
|
### Range from accumulated shift
|
|
|
|
Integrating the two-way shift gives the range,
|
|
|
|
```text
|
|
R(t) = R0 - integral 2 * c * v_r / f0 dt
|
|
```
|
|
|
|
which is why a tracking system measures range as the integral of a
|
|
frequency rather than as a timing. The accumulator drifts with any error in
|
|
the reference frequency, which is why the reference clock is the limiting
|
|
element in the whole chain.
|
|
|
|
### Range rate from the carrier
|
|
|
|
Alternatively the carrier phase is differenced directly, giving range rate
|
|
without an accumulator. This is the common approach and it is noisy, because
|
|
phase noise differentiates into range rate noise amplified by the carrier
|
|
frequency. The trade is a clean velocity against a degraded one.
|
|
|
|
### Clock offset
|
|
|
|
Separately from the motion, the transmitted frequency may not equal the
|
|
nominal one. A **frequency offset** between the spacecraft clock and the
|
|
reference is indistinguishable from a range rate over a short arc, because
|
|
both move the phase. Separating them needs a second observable, typically a
|
|
range measurement from a different station, or an upload that is
|
|
self-referenced.
|
|
|
|
## Unit conventions
|
|
|
|
- Frequencies in hertz; a shift is often expressed in parts per billion.
|
|
- Range rate in metres per second or kilometres per second.
|
|
- Range in kilometres, with the carrier frequency in hertz, so the factor
|
|
`2 c / f` carries its units.
|
|
- The factor of two applies to two-way; one-way is the single case.
|
|
|
|
## Applicability
|
|
|
|
Doppler is insensitive to noise that is common to the uplink and the
|
|
downlink, which is why it is the observable of choice in a jamming
|
|
environment: a jammer must be coherent to spoof it.
|
|
|
|
The shift is large enough to measure and small enough to alias. If the
|
|
estimate of range rate is off by more than half a carrier cycle per sample
|
|
interval, the tracking wraps and the recovered position is wrong by an
|
|
arbitrary amount. This is a real failure mode in practice and it is why the
|
|
sample rate is bounded relative to the dynamics.
|
|
|
|
A one-way system needs a stable reference on the spacecraft; a two-way system
|
|
needs a stable reference on the ground. Either way the clock, not the
|
|
antenna, is usually the limit.
|
|
|
|
## Standards
|
|
|
|
- Range and Doppler conventions: the tracking and data link messages
|
|
specified by the CCSDS Space Link Protocol
|
|
- Reference oscillator discipline: the same documents, in the time and
|
|
frequency parts
|
|
|
|
## See also
|
|
|
|
- [[link-budget]] for whether there is enough signal to measure at all
|
|
- [[interference]] for what a coherent jammer does to this measurement
|
|
- [[tt-and-c]] for the full measurement chain |