July 30, 2017
July 30, 2017
This part of The Beam-Hopping blog post focuses on core technical aspects of a Beam-Hopping system elements. We address the technology challenges which arise from the theory and need to be solved practically in the design. Part one of the series was published last week.
Traditionally satellite communication payloads are designed as “bent pipe”, where the uplink signal is received, filtered converted to the down link frequency, amplified and filtered again for transmission back to earth. The operation of the satellite is practically transparent to the users. The introduction of High Throughput Satellites has naturally complicated the payload design, adding the need for de-multiplexers, switches, multiplexers etc. Modern design is using channelizers based on digital technology for that purpose.
Beam-Hopping is a time-domain technique, for which time alignment and time slot allocation are very important. If the payload is to remain transparent, accurate time alignment has to be kept between the uplink receive antenna all the way to the downlink transmit antennas. In addition, some switching time is required, depending on the antenna, transmission channel and switching device. Practical values are in the order of magnitude of few μs.
A regenerative payload is proposed for a microsatellite scenario. Regenerative payloads require on board full demodulation and decoding of the uplink and then routing, encoding modulation of the information in the downlink. Those payloads provide higher flexibility in routing and in link adaptation. It allows mesh connectivity, protocol conversion, and transcoding between different earth stations. Thus a suitable waveform can be selected according to each link in accordance with its SNR. In the context of beam-hopping, with regenerative payloads, the uplink hopping plan is decoupled from the downlink one, and the gateway transmissions need not be strictly synchronous with the beam hopping transmission plan. The payload, in this case, should be able to make the routing, queuing and buffering for all the beams.
However, using regenerative payloads would limit the users to only those waveforms supported by the payload, which may limit the type of applications supported by the satellite. This limitation may be even more restrictive during the lifetime of the satellites as needs and technologies develop. This limitation gets more significant as the size, cost and lifetime expectancy of the satellite increases. But, with further development of digital techniques such as software defined radios, the confidence of launching satellites equipped with regenerative payload would certainly grow, as over-the-air adaptation and modifications would make this solution as future proof as a transparent payload.
As it is well-known, the receiver in a wireless system has to synchronize to the transmitter in frequency, time, and in coherent receivers in phase as well. It also has to adapt to the channel impairment. A satellite terminal typically performs this synchronization constantly, using the forward link which is transmitted continuously. In a beam hopping environment, this is not the case, as the satellite beam illuminates the terminal only during its dwell time. The rest of the time the terminal is in the dark without receiving any signal.
The terminal has to acquire the satellite signal in two, rather different circumstances:
It can be quite safely assumed that the burst receiving conditions are more of the signal loss type rather than cold start, but, depending on the off-time interval, oscillator’s drift and instability and dynamic changes may require that the receiver performs re-acquisition.
The acquisition engine of SatixFy’s SX-3000 ASIC has the capability to lock onto frequency deviations of up to 30% of the symbol rate.
The key component in the acquisition is the programmable correlator, which correlates the signal with the expected header. The same component is also used to detect the PLS, SOF, and SOSF for non-VL-SNR frames and for superframes. The correlator is fed with the symbol stream. At the correlator, each symbol is multiplied by the complex conjugate of the expected value and all the results are summed up to produce the correlation value. If there is no frequency offset, the summation produces a peak when the correct symbol sequence has entered the correlator. However, in a case of a frequency offset, the phase shift due to this offset should be compensated for in order to detect the header. This compensation is performed by a FFT operation to which the products are fed.
Beam Hopping inherently requires synchronization, especially so when multiple teleports and many terminals are involved, and we assume that no time adjustments or buffering are to be done on the satellite (without the regenerative payload described above).
The critical component of the synchronization accuracy is that beam transmission, at the satellite output. The required accuracy for synchronization is highly dependent on the switching time between beams and on the accuracy of the means used for synchronization. One can consider two options for synchronization: synchronization of the teleports and gateways by the standard synchronization means in an open loop or using a satellite loopback.
Standard methods of synchronization have the following accuracy:
As for the terminals, the jitter and accuracy of chips like the SX-3000 ASIC are in the order 10-100 ns. Thus, synchronization accuracy of the order of 1μs seems to be feasible.
Although the open loop concept described above could be made to meet the specifications, it might be beneficial to use a Satellite auxiliary transmitter that would add robustness to possible variations in the system.
The Satellite auxiliary transmitter can be considered as a device that provides to the teleports information about actual hop times and transmission times. It can be a pulse train transmitted synchronously with the hop times or timestamps with the transmission times or hop time information. A hybrid synchronization system will then consist of a GPS or equivalent master clock at each teleport and a synchronization module that synchronizes the clock to the actual transmission times. This additional hardware makes it possible to compensate for variations in propagation delay, the stability of hardware and clock inaccuracies.
Terminals do not have to be accurately synchronized to the network. In that case, a terminal would be required to search for the satellite signal and re-acquire it every dwell time. However, network synchronization would enable the terminal management to be aware of transmission times and to be power down or revert to sleep mode when possible, thus saving battery power.
The transmission waveform is an important part of a beam-hopping system. The waveform should provide ample time for beam hopping and means for the receiver to reacquire the signal without loss of information. On the forward channel, the considerations for selecting a waveform are dependent whether the standard or non-standard waveforms are selected. For a transparent payload, non-standard waveforms can be used, as long as the transmissions are made within the frequency and time slots allocated for the user. However, the DVB-S2/S2X waveform is often selected in both broadcast and broadband transmissions, as it provides high-efficiency high throughput waveform for a wide SNR range very well matched to the satellite channel, with all the advantages that come by using a standard.
The DVB-S2/S2X waveform, on the other hand, was not designed for beam-hopping. It is indeed a TDM, but of variable frame length dependent on the chosen bandwidth, modulation and coding. For a 10 MHz bandwidth frame length would range between 0.35 to 3.5 ms, at higher bandwidths for which one mainly aims for, e.g. at 500 MHz bandwidth, the frame length would range between 7 μs to 70 μs. This order of magnitude suggests that network synchronization and beam switching time of the order of a few μs would present a large overhead for frame-by-frame beam-hopping. This high overhead at high frequency and the frame variability makes the usage of this waveform problematic for beam-hopping.
The DVB-S2X Annex E, a superframe waveform is (see figure below) is a container of a fixed large number of symbols, which, for a given symbol rate, forms a fixed time frame. The number of symbols in a superframe is 612,540 and its duration is about 1.3 ms for a 500 MHz bandwidth. The superframe also provides for a longer header, fixed time pilot signals for improved synchronization and set of dummy symbols at the end of the frame to allow for smooth switching during hopping. The large container enables embedding a number of S2/S2X frames within the superframe, each transmitted to a different user with a different modulation and coding as adequate. However, the superframe duration dictates the granularity of the hopping scheme and adds to the total latency of the channel.
On the return channel, the situation is quite different. One should distinguish between two different access technologies that are used for the return channel, random access, and controlled access. While the first access method the terminal transmits whenever it has information, such as the case for the various Aloha and spread-Aloha techniques, whereas for the controlled access the terminal is allocated a time slot and frequency slot for its transmission, such as the case for the DVB-RCS2. While the latter lends itself to beam hopping without any modifications, random access return channel would have to be modified to allow transmissions only on predetermined time slots, leading to a Slotted-Aloha structure. In any case, beam hopping on the return would be applicable if equipment and resources savings could be made at the payload by using it. On the return channel, the effect on power saving is not as substantial as that of the forward channel since there is no transmission power involved.
(end of part 2)
In the next part, we shall explore the whole eco-system and discuss the satellite operator benefits.