Center for Mapping
Receiver
Performance Considerations
In general, the receiver performance criteria are function of the number of satellites visible, occupation time, observation conditions, obstructions, baseline length and environmental effects, as well as atmospheric conditions, and consequently, the factual performance will vary as these conditions change. It is very important to fully understand what can be expected under what conditions, and how the receiver copes with the unfavorable environment. A few important aspects are discussed below.
Thermal
noise: This is the most
basic kind of noise, produced by the movement of the electrons in any material
that has temperature above 0 Kelvin. The commonly used measure of the received
signal strength is the signal-to-noise-ratio, SNR. In case of RF and IF, the
most commonly used measure of the signal’s strength is the
carrier-to-noise-power-density ratio, C/No, defined as a ratio of
the power level of the signal carrier to the noise power in a 1 Hz bandwidth.
C/No is considered a primary parameter describing the GPS receiver
performance (thus, how well the tracking loops can track the signal), as its
value determines the precision of the pseudorange and carrier phase
observations.
Bandwidth of
a tracking loop: An
engineering design of a receiver must always involve some trade-offs. For
example, if the bandwidth of a tracking loop is made very narrow, the receiver
will provide very low noise level, but might fail tracking under some dynamics
(or due to rapidly changing ionosphere), and vice versa – if somehow larger
noise level that comes with a wider bandwidth can be accepted, a larger
measurement noise should be expected. Thus, in case of a narrow bandwidth, very
rapidly changing ionospheric conditions might cause losses of lock, especially
on the L2 frequency under Anti-Spoofing (AS). The estimated maximum
rate-of-change of ionospheric propagation delay, under conditions where
tracking is still possible, is about 19 cm per second, which corresponds to
about 1 cycle on L1.
Interference and Multipath Effects: Radio interference represents another phenomenon contributing to the overall receiver noise level. Several kinds of interference can potentially disrupt a GPS receiver’s operation: in-band emission, nearby-band emission, harmonics, and jamming, most common sources being radio communication, mobile phones, power lines, radar systems, equipment operated by police and emergency cars, etc. Under less than ideal conditions, radio interference can, at minimum, reduce the GPS signal’s apparent strength (that is reduce SNR by adding more noise) and consequently – the accuracy, or, at worse, even block the signal entirely. The most common interference protection used in GPS receivers are null-steering antennas, narrow front-end filters and narrowed, aided tracking loops (on-receiver techniques, generally limited to the kinds of interference they can reject), and most recently developed interference suppression unit (ISU). ISU consists of a simple patch antenna and electronic unit that plugs directly into the GPS receiver antenna port.
Another important aspect of
the receiver performance is its resistance to multipath. Multipath distorts
both the signal modulation and phase of the carrier signal, thus degrades the
accuracy of conventional and differential GPS. As opposed to interference,
which disrupts the signal and can virtually provide no or useless data,
multipath would allow for data collection, but the results would be wrong.
Reflected signals can introduce centimeters of error to the carrier phase
observable and several meters to the code phase. In case of a non-moving
receiver, multipath has a strong effect on accuracy, as the antenna remains
stationary to any multipath-generating objects nearby. In kinematic situation,
however, multipath signals may be less troublesome, as GPS antenna is moving
relative to multipath-generating objects, which can result in better accuracy,
as compared to the static case.
In-receiver Data Smoothing and Correlation: The RMS of the residuals from the position estimation process does not necessarily indicate the factual achievable accuracy. In fact, it provides a meaningful accuracy estimate only if the data points in the sequence used to evaluate RMS are statistically independent. However, most of the GPS receivers provide the data averaged (or smoothed) over some period of time, and thus – statistically dependent. This practically means that the error noise in one measurement is almost identical to that of another measurements, and each new data point does not provide much additional information. Thus, the RMS derived from these data will depend on how much smoothing was applied. On the top of data smoothing (or averaging), dual frequency GPS receivers represents even more complications scenario due to the encryption of the P-code, as a consequence of the Anti-Spoofing (AS) policy. To battle AS, codeless and quasi-codeless techniques that enable access to L2 code and carrier phase observations are used. Most receivers reconstruct the L1 carrier by code correlation using C/A-code, and then one of the codeless techniques follows to reconstruct the L2 carrier and pseudorange. Virtually all geodetic-grade receivers guide their encrypted pseudoranges by L1 carrier phase estimate, ultimately introducing correlation among the measurement types.