TRANSPARENT SATELLITE SWITCHING
- Using Flexible Frequency-band Reallocation

-Funded by CENIIT, Linköping University, Sweden

Project Leader: Per Löwenborg

Introduction

The future society foresees globally interconnected digital communication systems offering multimedia services, information on demand, and delivery of information (data) at high data rates and low cost and with high performance. Terrestrial networks could in principle meet the requirements on communication capacity due to the practically unlimited bandwidth provided by fiber optic cables.
The main terrestrial candidate networks are:

• ADSL - Asymmetric Digital Subscriber Line, which provides asymmetric data transmission on regular twisted-pair telephone lines that can provide up to 8Mbit/s downlink and 256k bit/s uplink.
• HFC - Hybrid Fibre/Coax networks. Communication on the combination of coaxial cables and optical fibres to provide a high data-rate poit-to-point delivery of voice, data, and video. Fibres can provide data rates in the order of Tbit/s.
• LMDS - Local Multipoint Distribution Service providing point-to-point or point-to-multipoint communication using microwave frequency radio transmission over short distances.
• UMTS - Universal Mobile Telecommunications System featuring a 2 Mbit/s radio transmission scheme and is used within the third-generation mobile systems.
• Others - Data transmission can also be provided by cable television networks and power distribution lines.

However, the capacity, in principle achievable using terrestrial networks, is rarely available today; a large investment is required to bridge the distance between the local exchange and the customer. It is therefore internationally recognized that satellite systems will play an important complementary role in providing the global coverage required for both fixed and mobile communications [1] - [3].Typically, each next generation telecommunication satellite is forseen to provide a total capacity in the order of 100 Gbit/s, divided among the users.

Using satellites to provide network access has a number of advantages compared with terrestrial wideband access methods. These are: [4]

• Coverage of sparsely populated areas. A satellite system overcomes the geographical obstacles that prevent or degrades the quality and coverage of other access network methods and can cover isolated areas. Many geographical regions today have population densities that cannot make other access schemes economically justified.
• Flexibility and adaptability to changing markets. The lifetime of a communications satellite is about 8 to 15 years. During that time the market for the communications and multimedia services can change. Due to its flexibility, the satellite can easily be adjusted to such changing conditions.
• Suitability for multi- and broadcast services. The satellite is inherently an excellent broadcasting tool.
• Rapid global deployment. The time to achieve a global coverage is substantially smaller than that of other access networks.
• Maritime network access. Satellite communication is highly suited for providing network access in maritime environments. The excellent coverage of satellites has earlier been demonstrated through the success of the Global positioning system (GPS).

Among other advantages are the possibility provide fast infrastructure for developing countries and independency of local and regional censorship etc.

In order to meet the requirements of the communication systems of tomorrow, it is imperative to develop a new generation of satellite systems, payload architecture, ground technologies, and techniques combining flexibility with cost efficiency. It is envisaged that the improvements required as to the capacity as well as complexity fall in the range of one and two orders of magnitude [1]. The European Space Agency (ESA) outlines three major "standard architecture" for future broadband systems [1]. Two of these are the distributed access network and professional user network which are to provide high-capacity point-to-point and multicast services for ubiquitous Internet access. The satellites are to communicate with user units via multiple spot beams as illustrated in Fig. 1.


Figure 1. A communication satellite providing network access via multi-spot beams. A number of gateway beams are transmitting and receiving data from a large number of user beams.

In order to use the limited available frequency spectrum efficiently, the satellite on-board signal processing must support frequency-band reuse among the beams and also flexibility in bandwidth and transmission power allocated to each user; further, dynamic frequency allocation is desired for covering different service types requiring different data rates and bandwidths. One major issue in this next-generation satellite-based communication system is therefore the on-board reallocation of information. In technical terms, this calls for digital flexible frequency-band reallocation networks which thus are critical components.

Terrestrial Sensor Network Relay Nodes

The Swedish parliament has decided that Swedish armed forces are to be developed according to the concept of Network based defence (NBD) which will greatly enhance the defence efficiency by linking decision-making, information systems and weapon systems, into a single networked organisation. A very important part of the NBD is the exchange of information through wired and wire-less communication networks. The diversity of these networks calls for flexible and efficient network nodes to transmit and receive data with various communication modes, rates, and formats. In this context, satellite and/or terrestrial flexible sensor network relay-nodes (SNRN) are proposed to be used for supporting reliable operation. These devices process information in a very similar way to communication satellite scenario, providing frequency-band reuse and flexibility in terms of bandwidths, transmitted power, and communication modes, through the use of digital flexible frequency-band reallocation networks.

In principle, communication node relaying can either be transparent or regenerative. The frequency-band reallocation principles studied here is a transparent relaying. In the first scenario, the data is redirected without actually retaining the information carried in the received signals before retransmitting it. Hence, this relaying principle is similar to so called circuit switching used in wireline communication. The main advantage of this relaying scheme is that the relay nodes become independent of the communication modes and data transmission principles as long as signal bandwidths and centre frequencies as well as their destinations are knows. The drawback is that the signal-to-noise ratio is successively degraded as the signals travel through a sequence of relaying nodes since during every signal reception, noise is introduced into the information carrying frequency bands. This tends to limit the number of relaying nodes in a sequence.

In the regenerative relaying scenario, the data is retrieved from the received signal and is recoded and retransmitted again at the relaying node. Provided that the data is sufficiently protected by channel coding, no information is lost during a sequence of relaying nodes. This is similar to so called packet switching used in wireline communication where information is extracted from within the data packages in order to determine the destination of the data. However, this principle can be problematic to use in scenarios where multiple communication modes and data transmission principles are used simultaneously, since the relay nodes then need to act as multistandard transceivers at the same time as a redirecting data. Therefore, transparent relaying appears to be preferable in multi-standard communication scenarios.

On-board processing

The digital part of the satellite on-board processor is a multiple-input multiple-output digital system. The principle architecture of such a processor is shown in Fig. 2.


Figure 2. Principal architecture, describing the fundamental signal processing of a transparent communication satellite.
An antenna matrix receives a combination of signals from multiple sources. These are then frequency demodulated, converted into a digital representation (ADC) and beamformed digitally (DBF). The signals are then switched (FBR) before being transmitted through multi-spot beams. The number of input signals can in general differ from the number of output signals, and the input and output signals can generally use different bandwidths and data rates.

The next-generation satellite-based communication systems discussed above must support several different communication and connectivity scenarios. One main such scenario is based on MF/TDMA (Multiple Frequency Time Division Multiple Access) access schemes. In this case, the bandwidth of each incoming signal is composed of a number of adjacent smaller frequency bands (subbands), each subband being occupied by one (a few) user (users). A main role of the on-board processor is to reallocate all subbands to different pre-specified output signals and positions in the frequency spectrum.

Further, in order to efficiently utilize the available frequency spectrum, the systems must support bandwidth-on-demand which means that the different subbands can have different bandwidths and may vary with time. This is handled by dividing the input beam into a number of granularity bands; any user can occupy one or several such granularity bands and this number may vary with time. Figure 3 illustrates the principle of frequency-band reallocation (in practice, one must also include frequency guard bands between the subbands in order to make the network realizable).


Figure 3. Principle of frequency-band reallocation. Here the subbands are frequency modulated and recombined.
The following main requirements to meet for the next-generation high-performance frequency-band reallocation networks are identified:

• Flexibility - frequency bands of different and variable bandwidths must be handled.
• Low complexity and inherent parallelism - The complexity in terms of arithmetic operations and memory storage must be low. Further, the network (algorithm) itself should not impose restrictions on the feasible throughput; the implementation technology available should be the limiting factor. Meeting these requirements, high-throughput/low-power implementations can be achieved.
• Perfect frequency-band reallocation - Perfect frequency-band reallocation (PFR) means that each subband can be shifted to the new positions without errors. By using a frequency-band reallocation network that is able to approximate PFR as good as desired, the degradation of the overall system performance [often measured in terms of bit-error-rate (BER)] due to these networks can be made as small as desired.
• Spectral efficiency - The available frequency band is limited and must be used efficiently. This calls for spectrally efficient reallocation that, with reference to Fig. 3, means that the decomposition must be performed with a minimum amount of of frequency band overhead.
• Simplicity - Simplicity means that the frequency-band reallocation network should be easily analyzed, designed, and implemented. Although this may not be strictly needed in order to arrive at a high-performance processor, it is indeed recognized in the scientific society that if two different methods are equally good, in some sense, one should preferably use the simplest one.

Trends in satellite communications

The following trends can be identified within the area of satellite communications [5]:

• Development of critical technologies. Critical technologies such as on-board high power generation, on-board processing and switching, advanced antenna technologies using phased-array antennas, and new multimedia services and applications are being developed.
• Broadcast and multicast services. The inherent broadcasting nature of satellites has triggered the development of new applications such as Internet protocol (IP) multicast support, caching technologies and streaming.
• Emerging standardization. The Technical committee for satellite earth stations and systems, (ETSI TC SES) has currently a number of Work groups including "Ka-band earth stations" and "Broadband satellite multimedia". ETSI is also currently cooperating with other major national and international standardization organizations.
• Increasing capacity and functionality. Next generation satellite multimedia and communications systems features very high throughput, flexibility, and multi-beam processing. This pushes the development towards, advanced but computationally efficient digital signal processing algorithms for on-board processing such as beam-forming, advanced circuit switching techniques, and adaptive amplifier linearization.

From these trends it can be identified that flexible satellite switching, as for example through the use of flexible frequency-band reallocation, plays a major role and is a key component in future broadband satellite systems.

Goal of the project

The goal of this project is to develop the theory and design methods for a new proposed filter bank (FB) based frequency-band reallocation network. preliminary results show that the proposed technique can substantially outperform the existing ones when all aspects on flexibility, low complexity and inherent parallelism, perfect frequency-band reallocation, and simplicity are simultaneously considered. In particular, compared with previous techniques that can approximate PFR as good as desired, the new technique may have one or two orders of magnitude lower complexity. Thus, with reference to the introductory text, the proposed technique has the potentials of becoming a standard solution for the next-generation satellite-based communications systems.

As can be seen from Fig. 2 other key system components are the analog/digital interfaces and efficient beamforming networks. We intend to include these components into the focus of this project and study them in the context of frequency-band reallocation systems. The analog/digital interfaces need to simultaneously achieve highsample rate and resolution, something that is very difficult to achieve in practical implementations due to fundamental limitations in circuit technology. Our approach to overcome this is to increase the sample-rate/resolution product by parallelization in conjunction with digital estimation and correction techniques for removing harmful distortion caused by the parallelization. Efficient beamforming architecture are also of interest and we intend to make use of optimization-based temporal filter design.

Project status

A new technique is being developed that can substantially outperform the existing ones when all desired aspects such as flexibility, low complexity and inherent parallelism, perfect frequency-band reallocation (PFR), spectral efficiency, and simplicity are simultaneously considered. In particular, compared with previous techniques that can approximate PFR as good as desired, the new technique may have one or two orders of magnitude lower complexity. Thus, with reference to the text above, the new technique has the potential of becoming a standard solution for the next-generation satellite-based communications systems. This technical breakthrough was enabled through the invention of a new class of multirate filter bank structures, incorporating concepts that with respect to other conventional filter bank applications has been of no practical use previously, but with respect to transparent switching has appeared as an excellent solution.

In technical terms, we are developing a frequency-band reallocation technique based on a new type of complex- modulated N-channel filter bank which uses decimation and interpolation by M and handles Q input and output subbands, with Q being variable and M and N fixed. By properly selecting N, M, and bandpass filters (channel filters), given a maximum predefined value of Q, this new class of FBs can: 1) handle all possible frequencyshifts, 2) handle all possible bandwidths, 3) achieve as low complexity as in regular complex-modulated DFT FBs, 4) achieve as much parallelism as in any of the previously existing methods, 5) approximate perfect frequency- band reallocation as close as desired via a proper design, 6) achieve a high degree of frequency selectivity, 7) easily be analyzed, designed, and implemented. Compared to existing methods of interest [6]-[8], the proposed technique appears to be superior.

The core of the proposed technique is a novel on-line variable oversampled complex-modulated filter bank. A paper describing this filter bank and its use for flexible frequency band reallocation has been published at IEEE International Conference on Acoustics, Speech, and Signal Processing, 2005. A journal paper manuscript has been prepared and is about to be sent for publication. As is obvious from Fig. 2, in order to yield high-capacity relaying of signals, efficient digital static and adaptive beam forming, beam steering, and high-performance A/D and D/A converters are also required together with the digital core of flexible frequency band-reallocation. Today, these are all important key research topics in electronics for communications as well as medical imaging, and radar applications. We have also developed very promising techniques for digtal linearization of very high-speed A/D converters which are useful in this context.

These and other related results have been published and are given in the list of publications.

References

[1] B. Arbesser-Rastburg, R. Bellini, F. Coromina, R. De Gaudenzi, O. del Rio, M. Hollreiser, R. Rinaldo, P. Rinous, and A Roederer, "R&D directions for next generation broadband multimedia systems: An ESA perspective," in Proc. Int. Comm. Satellite Syst. Conf., May 2002.
[2] E. Del Re and L. Pierucci, "Next-generation mobile satellite networks," IEEE Comm. Mag., vol. 40, pp. 150-159, Sept. 2002.
[3] M. Witting, "Satellite onboard processing for multimedia applications," IEEE Comm. Mag., vol. 38, pp. 134-140, June 2000.
[4] H. Skinnemoen and R. Rusch, "Multimedia satellite communications," in Proc. Global Telecom. Conf., 1999, pp.1167-1171.
[5] H. Skinnemoen and H. Tork, "Standardization activities within broadband satellite multimmedia," in Proc. IEEE Int. Conf. Communications, May 2002, pp. 3010-3014.
[6] M. L. Boucheret, I. Mortensen, and H. Favaro, Fast convolution filter banks for satellite payloads with on-board processing, IEEE J. Selected Areas Comm., vol. 17, no. 2, pp. 238 248, Feb. 1999.
[7] G. Chiassarini, and G. Gallinaro, Frequency domain switching: Algorithms, Performances, Implementation aspects, in Proc. 7th Tyrrhenian Int. Workshop Digital Comm., Viareggio, Italy, Sept. 1995.
[8] H. G. Göckler and B. Felbecker, Digital on-board FDM-demultiplexing without restrictions on channel allocation and bandwidth, in Proc. 7th Int. Workshop Digital Sign. Processing Techn. for Space Appl. DSP 99 , Noordwijk, Netherlands.

---

For information and questions regarding this web site, please contact Per Löwenborg, E-mail: perl@isy.liu.se

---