One essential component of any remote-sensing mission is the return of the data it obtains. For missions confined to Earth orbit this is not usually a particular challenge, as the link distance is very small and so free space loss is low. For a deep space mission such as this though, the larger link distance means a much higher free space loss and hence a more restricted link budget. The link budget has to meet the data download requirements of the mission; these are discussed in more detail in Section 13 but as a guideline a ten-minute encounter generating one 1024x 1024 x 8-bit pixel image per second will produce 600 MB of data at 8 Mbit/s.

Two principle data return strategies are available for a flyby mission. Data can be transmitted as it is generated, or it can be stored for later transmission. Both approaches have their advantages and disadvantages.

Immediate Transmission removes or renders less important the need for mass storage on board the probe. It allows the primary mission to terminate once the encounter is over, rather than being extended to allow data download. Also, if there is any doubt as to whether the probe will survive the encounter, it ensures that the bulk of the data is still broadcast*.

[*This was the approach adopted with the Giotto mission to Comet Halley, where there was serious doubt as to whether even a well-armoured probe would survive the high-velocity encounter with the particle-rich zone around the nucleus. In the event Giotto did survive, but only after sustaining severe damage.]

The disadvantage of immediate transmission is that it requires that data is transmitted at the rate it is generated. For a probe trying to obtain a large number of images of a target during a short encounter, this rate will be very high. Real-time data compression can be used to reduce the data rate, as discussed below, but to achieve a large reduction lossy compression must be used, which has implications for science data. Also, to maintain the optimum link budget, the probe's high-gain antenna must be kept Earth-pointing during the encounter. Although this may be possible with if the probe features a pointable imaging payload (see Section 8) it adds complexity to both probe design and the planning of the encounter. Finally, if there is no on-board mass storage at all, then any prolonged outage during the downlink (i.e. beyond the capacity of link error correction to handle) will result in permanent data loss.

Downlink from Storage allows data to be transmitted at a much slower rate than it was originally gathered. For instance, data gathered at rate R_d bits/s during a ten-minute flyby can be transmitted over a five-hour downlink session at rate R_d/30 bits/s, resulting in a thirty-fold reduction in downlink bandwidth and a proportional improvement in the received carrier/noise ratio for a given transmission power. The use of a separate downlink session means that the probe can reorient itself from target-tracking to Earth-pointing mode, allowing it to have both a fixed high-gain antenna and imaging payload. Outages during the downlink are also less of a problem as the missing data can be replayed as required.

However, downlink from storage does have problems. The longer downlink session at low data rate means that ground facilities (e.g. the NASA Deep Space Network) must be booked for longer periods. For a rapid-response mission launched at short notice, such prolonged supports may be difficult to arrange. The mass-storage system adds complexity, mass and power demand to the spacecraft design, although the continued advances in solid-stage storage technology are progressively making this less of an issue. Also, as the probe continues to recede from Earth after the target encounter the link budget will become progressively more unfavourable, although this is unlikely to become a significant issue unless the data downlink is seriously delayed.

It is worth noting that compromise data return strategies can be used. A probe could, for instance, archive flyby data in full for later transmission whilst transmitting a compressed subset of data during the encounter. This would provide some insurance against loss of the probe during the encounter and would allow real-time confirmation of the mission's success. If a sufficient reduction in real-time data rate could be obtained, this could even allow the real-time downlink to be via a low-gain omnidirectional antenna, removing the Earth-pointing antenna constraint.

Link Budget Calculation

In order to evaluate the communications design for the mission a link budget is required. Maral and Bousquet [Maral93] give a good introduction to link budgets in the context of communications satellites, but the method is equally applicable to telemetry budgets**.

[**The calculation is carried out for the downlink telemetry budget. There will also be a link budget for the command and ranging uplink from the ground to the probe. However, the data rate for this is certain to be much lower than the data downlink rate, so the uplink budget will be correspondingly more favourable. For this design study therefore, only the downlink budget is considered.]

For transmitted power P_t, transmit antenna gain G_t and receive antenna gain G_r the received carrier power C is given by

(12-1)

where L is the total path loss. For the purposes of this calculation atmospheric and rain losses are ignored as they are much less significant than the free space loss L_fs between the probe and receiving station. Free space loss is given by

(12-2)

where R is the free space range and lambda the downlink frequency. We can further specify G_t in terms of and antenna diameter d as

(12-3)

where eta is the antenna efficiency, typically 0.55 for a paraboloid reflector. We thus have an expression for received carrier power in terms of antenna gain and range:

(12-4)

Note that this expression appears to be independent of the operating wavelength, as the increase in transmit antenna gain due to any increase in frequency is offset by an equivalent increase in free space loss. However, G_r will also increase in proportion to lambda^-2 (or f²) so increasing the operating frequency increases received carrier power. Spacecraft telemetry transmission for deep space missions has traditionally been carried out at S-band (2.2 GHz) but for this reason is now increasingly carried out at X-band (8.4 GHz). Furthermore the use of higher frequencies and thus shorter wavelengths allows smaller and thus lighter RF components to be used. The telemetry link for this mission was therefore assumed to be at X-band.

For reliable data detection the received carrier/noise ratio must be above a minimum limit. In fact, as we are dealing with a binary data stream, the energy-per-bit to noise density ratio, E_b/N₀, is used in the link budget. E_b/N₀ is defined from the bit rate b and the carrier powerC as

(12-5)

with T being the receiver noise temperature and k Boltzmann's constant.

We now have an expression for E_b/N₀ in terms of all the link parameters:

(12-6)

It is more convenient to express this in terms of dB, i.e.

(12-7)

where G/T, the ratio in dB/K of receiver gain to noise temperature is the receiver's characteristic figure of merit.

For a range of 0.05 AU, a transmit power of 5W and an antenna diameter of 0.7 metres (as used in the final study configuration) then for a receiver with G/T of 55 dB/K the E_b/N₀ for data rates from 10 kbits/s to 10 Mbits/s is shown at Fig 12.1. This G/T value is for a 34 metre Deep Space Network antenna; for a smaller or higher-noise antenna the data rate would be lower.

The value of E_b/N₀ needed for demodulation depends on the modulation system and type of error-correction coding used. For a typical binary data link, including system margins, a bit error rate of 10^-6 requires an E_b/N₀ of 12 dB [Maral93]. From Fig 12.1, the probe downlink could thus support a data rate of 2 Mbit/s. If error-correction coding is used though then a higher data rate can be supported as the error correction offsets the higher BER. Although the error correction coding increases the data rate by adding check data this can be more than offset by the reduced E_b/N₀ required [Sweeney91]. For example, the use of even a simple half-rate code can reduce the E_b/N₀ required for a BER of 10^-6 by 5dB, which for the system described would allow data transmission at approximately 7 Mbit/s.

In practice such high data rates are rarely used for deep space probe. The DSN can currently support data rates at up to 6.6 Mbps [Larson92] but to allow extra link margin it was decided to specify a data rate of 1 Mbps. As discussed in Section 13 this would allow the download of a typical encounter data set in under ten minutes if lossy data compression was used.

As mentioned earlier, if the probe's antenna is not Earth-pointing then it could still use an omnidirectional antenna to transmit low-rate data. Fig 12.2 shows the probe downlink assuming that this is done. Without coding a data rate of 1 kbit/s is possible, whilst the use of coding as described earlier allows this to be increased to 3 or 4 kbit/s. This is certainly enough to maintain probe telemetry and may allow limited transmission of highly compressed images. Use of image compression and matching of downlink rates with the probe's data return is discussed in Section 13.

Probe Design Aspects. The mass of the communications payload equipment was estimated based on guidelines in Space Mission Analysis and Design and systems used on other small payloads. Given the simple and low-cost nature of the mission it was decided not to provide redundant transmitter and receiver units. Antenna mass was estimated by scaling from composite antennas used on other spacecraft [Wertz96] The communications mass budget is listed at Table 12.1.

Summary. The use of error-corrected X-band telemetry via a high gain antenna and the NASA Deep Space Network allows a data downlink rate of up to a maximum 7 Mbit/s to be supported. This allows for real time data downlink (possibly with some compression) or gives a very large margin for slower replay after the encounter, allowing for a less sensitive ground station to be used. For design study purposes a downlink rate of 1 Mbit/s was assumed. The downlink system could also support kbit/s level data transmission via an omnidirectional antenna if the probe design precluded it from keeping its high-gain antenna Earth-pointing during the target flyby. This would allow spacecraft telemetry and a subset of highly-compressed images to be transmitted.