ITMI20071713A1 - POINT-POINT RADIO SYSTEM AND METHOD FOR THE FUNCTIONING OF THIS SYSTEM. - Google Patents

Description

“Point-to-point radio system and method for operating such a system” The present invention refers to a point-to-point radio system in accordance with the preamble of claim 1.

In particular, the present invention finds application in configurations with multiple inputs and outputs, i.e. with multiple transmitting antennas and multiple receiving antennas, so-called MIMO (Multiple Input Multiple Output) configurations, which create multiple radio paths between the transmitting side and the receiving side of a connection. . These configurations are used in particular in the technology of cellular radio systems, therefore operating at frequencies much lower than those of microwave radio systems, to create arrays of particularly direct and electrically orientable antennas, so-called intelligent antennas, or to create a spatial multiplexing function. , which allows to increase the transmission capacity with the same bandwidth occupied.

The MIMO configurations can be symmetrical or MIMO (N, N), or asymmetrical, for example MIMO (1, 2), i.e. single input and double output, commonly known as reception diversity systems.

In the context of the present invention, reference will also be made to the adaptability of the physical level. With the term adaptivity we mean the procedure whereby, as the propagation conditions vary over time, the transmission modes on the radio system are dynamically and automatically changed, depending on the same conditions and in a manner adapted to them. More specifically, with the same bandwidth occupied, the physical level is varied between two or more available conditions to better counteract the unfavorable effects of the propagation itself.

In point-to-point radio systems, the technology of the products currently available is based on conventional configurations with single input and single output, ΜΙΜΟ (Ι, Ι), and with fixed or configurable modulation, i.e. not suitable for propagation.

The use of physical layer adaptivity in radio systems has so far been limited in the case of point-to-point cellular systems where a plurality of peripheral terminal stations distributed over a certain geographical area, called a cell, is connected to a central station. Traffic is generally mixed as a type of service, but typically these are services in which data is prevalent.

For example, in the time division multiple access mode (TDMA = Time Di vision Multiple Access), the various terminal stations transmit in distinct and non-overlapping times to maintain mutual orthogonality. The information to be transmitted by each terminal station towards the central station, in the so-called up-link direction, but possibly also in the opposite down-link direction, is consequently segmented into elementary blocks and the latter are appropriately mapped, by means of modulation, on signal packets or bursts transmitted in distinct times, which are suitably adjusted so as not to have overlaps, not even partially, on the central station receiver to avoid any mutual interference.

The traffic mode of the system is of the type with capacity assignment on request, which in practice translates into a variable number of bursts disposed of by each terminal station in the unit of time so as to achieve a transmission capacity (throughput) that varies over time. . In these conditions it is natural to use different physical levels for each burst, in order to optimize the overall throughput on the cell and to better adapt to the conditions, even very different from each other, that the various terminal stations face in their respective connections with the station. central station due to very different distances between central station and terminal stations, different propagation conditions in different areas of the cell, different and time-varying transmission capacity requirements (throughput), from terminal station to terminal station.

Asymmetric MIMO configurations with multiple inputs and a single output have been proposed at the central stations in cellular radio systems. These are generally MIMO configurations (N, l) used for radiant systems of central stations with sectoral coverage and with several elements, in order to increase the overall directivity of the central station and to realize a spatial multiplexing functionality (intelligent antennas). Solutions of this type, combined with the suitability of the physical layer, are for example envisaged for the “WiMax” technology in the standardization referred to in IEEE-802-16 Documents.

MIMO configurations in point-to-point radio systems are described in patent applications EP 1 229 679 and EP 1 448 006. However, these systems use a fixed or unsuitable physical layer.

Document WO 2007/069071 describes point-to-point radio links in adaptive MIMO configuration, applied however to OFDM modulation sources, as it is intended for low frequency ranges (<6GHz), with NLoS (Non Line of Sight) propagation scenarios and incorrelation between multiple paths due to the multiplicity of antennas.

The above solution is however inadequate for single carrier modulation sources, intended for high frequency ranges (> 6GHz) and therefore with LoS (Line of Sight) propagation scenarios.

The use of single carrier formats for high frequency ranges is motivated by fundamental technical reasons, such as the worst PAR ratio (Peak to Average Ratio) between peak power and average power that OFDM formats have compared to single carrier, with the same radio frequency emission masks to be respected and with the same band efficiency (ratio between useful tones and total tones for OFDM formats and ratio between symbol frequency and total band for single carrier formats). The worst PAR makes it inconvenient to use OFDM formats for frequency ranges where the availability of high power amplifiers with high saturation powers is critical and very expensive. Furthermore, the high spectral purity and therefore low phase noise imposed on the local oscillators by the OFDM formats are not economically achievable at high frequencies (> 6GHz) and in particular not available for the higher frequency ranges (23-38GHz). Finally, the uselessness of the main feature offered by OFDM formats (compatibility with NLoS propagation), since for high frequency ranges the propagation must still take place in LoS, due to the enormous attenuations from "shadowing" that otherwise would occur. encounter.

Therefore, point-to-point systems of the known type mentioned above are inadequate for single carrier and high frequency solutions.

Consequently, it is very important to have a point-to-point radio system and a method for operating this system that allows to exploit single carrier modulation formats, intended for high frequency ranges and with LoS propagation scenarios.

The object of the present invention is to provide a point-to-point radio system and a method for operating this system having structural and functional characteristics such as to satisfy the aforementioned requirements and at the same time to obviate the drawbacks mentioned with reference to the known art.

This object is achieved by a point-to-point radio system in accordance with claim 1.

This object is also achieved by a method for operating a point-to-point radio system in accordance with claim 13.

The further characteristics and advantages of the point-to-point radio system and of the method for its operation according to the present invention will result from the following description of one of its preferred embodiments, given by way of non-limiting example, with reference to the related figures. , in which:

- figure 1 shows the relationship between the signals in a MIMO configuration (2,2),

- figure 2 shows a block diagram of a MIMO configuration (2,2), - figure 3 shows a block diagram of a combiner / canceller in reception,

- figure 4 shows the relationship between signals in an lxNS (2,2) configuration, - figure 5 shows the relationship between signals in a 2xNS (2,2) configuration, - figure 6 shows the relationship between signals in a lxNS (l, 2) configuration, - figure 7 shows an example of signal segmentation for the two antennas and for formats up to 64QAM, i.e. with 2, 4 or 6 bits per QAM symbol,

- figure 8 shows a diagram relating to the management of switching in MIMO mode,

- figure 9 shows a comparison between point-to-point configurations,

- figures 10 and 11 show the performances of some configurations,

- figure 12 shows by way of example the separation between the antennas with a configuration of figure 10,

- figure 13 shows a block diagram of a point-to-point radio system according to the invention, e

- figure 14 shows an example of a process of adapting the life of the physical layer with a predefined profile.

In the continuation of this description, the term "adaptivity" will be used to indicate the procedure for which, as the propagation conditions vary over time, the modalities of broadcast on the radio system. More specifically, with the same bandwidth occupied, the physical level is varied between two or more available conditions to better counteract the unfavorable effects of the propagation itself.

The term "physical layer" indicates the set of operations carried out on the signal to be transmitted, relating to the modulation format and method and to the associated coding / decoding method for error correction (FEC) and jointly by the method of use of the antennas which determines the signal multiplexing factor m.

Given a radio frequency channeling band to be respected BW, a particular physical layer involves the ability to transmit a net capacity T, also called "payload" or "net throughput"), linked to the band occupied by the simple relationship: T (Mbit / sec) = m * p * BW / (1 + p)

Where

m is the multiplexing factor,

BW indicates the available band in MHz,

η (bit / Hz) denotes the net spectral efficiency, e

p is the roll-off factor of the shaping filter (in practice 0.2 <p <0.5).

With MIMO configurations (2,2) the multiplexing factor m can assume a value of 1 or 2 depending on how the antennas are used.

The spectral efficiency is higher the greater the complexity of the modulation format and the multiplexing factor m and is reduced as the transmission overhead increases, linked to the potential of coding / decoding for error correction. The performance of the physical layer, i.e. the threshold level a of the signal-to-noise ratio S / N between the signal powers and reception noise necessary to obtain a certain probability of error (BER = Bit Error Rate), also changes with complexity of the physical level itself. In other terms, the two parameters η, α are directly linked to each physical level, in the sense that to achieve higher h (more complex modulations), it is also necessary to have higher values of threshold level a.

PHYMIN indicates the minimum physical level, representing the most robust physical level, and with PHYMAX, the maximum physical level, representing the most efficient physical level.

If, as the propagation conditions vary, for example, the additional attenuation on the section increases over time, it switches to a less complex physical level, which therefore requires a lower threshold level a and can therefore better counteract the propagative deterioration in progress on the link . With the same BW band this will therefore result in a reduction in the net spectral efficiency η and a corresponding reduction in the instantaneous throughput T.

In other words, the adaptivity of the physical layer determines the variation over time of the instantaneous throughput T supported by the system. It can therefore be adopted only for the transmission of signals that tolerate the variability of the transmission rate over time, i.e. typically for the transmission of payloads which, at least for a portion, contain data that do not require the transmission in real time, whether of the type "celi oriented" (ATM) or "packet oriented" (IP). The point-to-point radio system according to the invention uses high frequency ranges (typically> 6GHz and more commonly> 13GHz), given the significant bandwidths offered by the relative standardized channeling plans and also given the relatively low lengths to be covered. (typically a few Km in an urban / suburban environment and in any case <15 / 20Km in a rural environment).

The band available for the single link or channel band BW is typically a multiple of 3.5MHz, BW = [3.5 - 7 - 14 - 28] MHz.

These are links in full visibility (LoS = Line of Sight) with very direct antennas and with negligible presence of dispersive effects on the engaged band, due to propagation with multiple paths (multipath). The echoes have small delays compared to the 1 / BW value due to the combined effect of the sections of limited length and the values of the bands themselves which are also limited.

That is, propagation causes non-selective attenuations in the band essentially attributable to two phenomena:

1) Multipath with small echo delays compared to 1 1 / BW value, a situation that can be considered prevalent for low RF ranges, from 6GHz to 13GHz,

2) Rain which can instead be considered prevalent for frequencies greater than 13 GHz.

Under these conditions, consider a configuration with plurality of transmitting antennas (or inputs) and plurality of receiving antennas (or outputs), briefly configuration ΜΙΜΟ (Μ, Ν), in the example schematized in Figure 1 MIMO (2.2). Since the inequality s "<L holds between the separation s between the antennas and the length of the section L, there is a high correlation between the propagation on direct and crossed paths.

In other words, two signals S1 and S2 are transmitted on the two transmission previews, respectively T1 and T2, the signals will be received respectively on the two reception antennas RI and R2:

[1]

In practice, the strong correlation between the various paths implies that the transmission coefficients ν / χχ are all equal to the same value w0e which results in Φ2ι ~ Φ12 ~ Φο.

Therefore, if the phase difference of the crossed paths compared to the direct ones is explicit in the term e <JO> °, this difference is essentially attributable to the different physical length of the paths themselves.

The relationship with

and finally Φο (degrees) ~ 360 * (s <2> / 2LX).

With regard to the characteristic parameters of the physical layer, it is assumed that the variation of the modulation format in the context of adaptivity occurs at the static peak power of the constant constellation, so as to work on the transmission amplifier under the same quasi-linearity conditions.

In this hypothesis the SG system gain will be defined as the difference in dB between the transmitted static peak power PPTX (dBm) and the reception threshold THRX (dBm), for the acceptable BER limit value:

SG (dB) = PPTX - THRX

Where

NF (dB) is the noise figure of the receiver,

BN (MHZ) the equivalent noise band of the same,

a is the ratio (S / N) between peak (static) power of the received signal and thermal noise for the BER threshold, and

BN is tied to the BW channel band.

For example, for QAM modulation formats with optimal distribution of filtering between transmission and reception, we have

where

Fs is the symbol frequency e

p the roll-off factor of the shaping filtering adopted (in practice 0.2 <p <0.5). It should be noted that, for the above relationships, the available system gain difference between two different physical levels used with the same transmitted static peak power PPTX and receiver, i.e. same noise figure NF and same equivalent band BN, is:

ASG (dB) = - Δα

therefore equal to the difference of the relative a, changed in sign.

The threshold condition for a given physical level, or the achievement of the predetermined BER limit, is simply represented by the equality between the system gain available for a given physical level and the total attenuation present on the link. That is, the relationship SG> ASL As must be valid where the following is set:

ASL (dB) = attenuation of free space (including the gain of the transmitting antennas T1, T2 and receiving R1, R2),

As (dB) = additional attenuation (due to propagation).

Given a section length L, that is fixed, that is ASL, and the other parameters PPTX, BN, NF being equal, the switching condition from a physical level with threshold a to the lower physical level, or the connection threshold condition if one is already at the minimum physical level PHYMIN, it is represented by the relation

As <SG - ASL = constant ~~ a - KK - a

which in fact expresses the maximum tolerable As value for a given physical level, or for the connection threshold.

The value of the constant KK depends on all the fixed parameters of the connection, that is, those parameters that remain constant when switching from one physical layer to another, namely L, PPTX, NF, BW, BN, Fs.

The higher the maximum tolerable attenuation As, ie the lower the value of a of the minimum physical level PHYMIN, the lower the probability POUT, defined as the probability over time of having a BER worse than a certain target.

From the above it emerges that, to highlight the basic functionalities of the physical levels for the purposes of their performance on the link, it is advisable to use the pair of characteristic parameters η, RSG, where η (bit / Hz) is the net spectral efficiency and RSG (dB ) is the relative system gain and is equal to - a.

In the context of the adaptivity of the physical layers it is possible to use the MIMO (2,2) configuration in two distinct ways.

"Lx" mode - The physical structure with multiple MIMO antennas is not used with spatial multiplexing functionality but to create only diversity configurations, either of order 2, that is simple reception diversity by turning off a transmitter and actually creating a MIMO scheme (l , 2), or of order 4, i.e. combined transmission and reception using the MIMO base (2, 2) in full. The supported throughput is not doubled, hence the “lx” notation.

“2x” mode - The MIMO (2,2) configuration is used to maximize throughput, ie with spatial multiplexing functionality of order 2, hence the notation “2x”.

The objective of the physical layer adaptivity is essentially to decouple the value of the two "effective" parameters in the system, namely the signal to noise ratio S / N of the connection threshold, or the value a of the minimum physical level PHYMIN, and the average spectral efficiency determined mainly by the values of net spectral efficiency η of the higher physical levels and of the maximum physical level PHYMAX, tolerating the variability of the throughput T over time.

In practice, therefore, it is advisable to size the system with the following distinct objectives: - Minimum physical level PHYMIN: A lower spectral efficiency η is tolerated, however within certain limits, for the minimum traffic T, for which the probability of POUT is to be minimized. This involves both the choice of a robust modulation / coding format, and the use of the MIMO configuration in "lx" mode in order to jointly minimize the value of a, therefore also the POUT probability.

higher physical levels and maximum physical level PHYMAX: more complex modulation formats are used and the MIMO configuration is operated in "2x" mode. in order to maximize the average spectral efficiency η of the link.

For the purposes of performance, the block diagram of a connection direction of the radio system according to the invention can be represented, as shown in Figure 2, the other direction being perfectly reciprocal.

The radio system comprises a physical layer 1 transmission processor, in which, for example, segmentation, service insertion, coding, mapping and double modulation operations are carried out, two radio frequency transmission modules 4,5, each comprising a module translation from an intermediate frequency IF to the RF transmission frequency, a power amplifier and a transmission antenna T1, T2, two radio frequency reception modules 6,7,

Each radio frequency receiving module 6,7 comprises an antenna, a low noise amplifier and a translation module from the RF transmission frequency to the intermediate frequency IF, a combiner / canceller 8, a processor of the physical layer 11, in which for example, double demodulation, demapping, decoding, service extraction and serialization operations are carried out.

The values of a for the various physical levels are to be understood at the input of receivers 6 and 7 (section A-A). With respect to the corresponding β values at the input of demodulators 9 and 10 (section B-B) the following naturally applies:

a (dB) - β - GCOMB

where GCOMB indicates the gain in S / N ratio of the combiner / canceller block 8. We will therefore have:

RSG = - a = GCOMB - β

Adopting QAM modulation formats, it is convenient to introduce the compact notation KxNS (N, M) to indicate a generic physical layer where:

K = index for how to use the MIMO structure (K = l for "lx" mode and K = 2 for "2x" mode)

NS = number of states of the QAM constellation

(Ν, Μ) = number of inputs and outputs of the MIMO configuration, respectively used.

For example, with the notation 1x4 (1, 2) we will indicate a physical level that uses the “lx” mode, with a 4QAM constellation and realizes the simple diversity of reception. The notation 2x64 (2,2) will indicate instead a physical layer that uses the “2x” mode, with a constellation 64QAM and realizes a spatial multiplexing of order 2.

The structure of the combiner / canceller 8 can be represented with the basic diagram of Figure 3, regardless of the mode of use of the MIMO configuration. In practice, each received signal R1, R2 is multiplied by the coefficient Ci2, C2i respectively, out of phase of the term and <Jy> and added to the other signal R2, R1 respectively.

Starting from the expressions of the received signals U1 and U2, we obtain, for the various modes of use of the MIMO structure, the following Table 1 of optimal values of the parameters to which the control algorithms of the combiner / canceller 8 must converge.

Table 1

It should be noted that in the physical level lxNS (l, 2) and lxNS (2,2) the output signal U2 is not used. It should also be noted that the physical layer lxNS (2,2) requires the coherence and the remote control of the average phases of the radio frequency signals transmitted.

The point-to-point radio system according to the invention therefore uses MIMO configurations with single-carrier modulation formats and physical level adaptivity, in which the physical level is defined jointly by an operating mode of the antennas and by a modulation format and coding.

A method and relative algorithm to support the functions described above is described below, specifying that alternative methods can be used without prejudice to the functions described above.

As regards the various operating modes of the M1M0 (2,2) configuration, it must be premised that the 2xNS (2,2) mode imposes the phase coherence of the receiving local oscillators to optimize cancellation. The lxNS (2,2) mode, on the other hand, requires coherence on the transmission side, in the following transmission diversity. Since typically the microwave integrated front-ends use a single local oscillator for the frequency translation from IF to RF and the inverse one from RF to IF, it is assumed to have coherent local oscillators both in transmission and reception.

Having said this, first of all it is necessary to provide for an automatic fine adjustment of the level of the two signals received, for example by means of two automatic gain controls AGC1 and AGC2, to make the average powers of the signals RI and R2 equal, thus compensating for any differences due to tolerances and / or imperfect calibrations of the blocks that bring the received signals back to the baseband, such as imperfect pointing and / or slight differences in gain of the antennas, different gains in the RF / 3F translations and the like

Subsequently, the other variables that need to be finely controlled are obtained from an examination of the three MIMO modes provided, reported below.

LxNS mode (2.2ì or reception and transmission diversity (FIGURE 41 For the combined transmission and reception diversity supported by this mode, it is necessary to adjust the transmission phases on the two antennas T1 and T2 so that the received signals arrive with appropriate phases to the For this purpose, a phase shifter is needed on one of the two transmission paths which introduces a phase shift e <Jcx> capable of absorbing any phase shift difference present between the pair of paths outgoing from the antenna T1 and the pair outgoing from the antenna T2. This phase shifter must be adjusted backwards through a telemetry channel with the criterion of obtaining the maximum of the combined signal in reception This maximum occurs for α · ^ 0.

In other words, if a phase difference β is accumulated on the pair of paths exiting the antenna T2 compared to the other pair exiting the antenna TI (difference due for example to inaccuracies in the geometry of the base, rather than to differences in electrical delay in the RF transmission front-ends, calibration imperfections and the like) a must be driven by the algorithm until it assumes the value a = (-β), for the necessary compensation. The situation is shown in Figure 4.

To obtain the control signal C, the difference is made between the modules of the signal RI and the signal R2.

Mathematically, the following applies:

| U1 | = | (4wo) Sl * cos (Oo / 2) cos (ot / 2) | maximum for a - ^ 0, a even Φο

Telemetry control to converge arH)

C - | R11 - | R2 | = - (4wo) Sl * sin (<J> o / 2) sin (a / 2)

2xNSf2.21 or spatial multiplexing mode (FIGURE 5)

In this mode, the signals received after the combination, assuming that the loop on the transmission has already led to "> 0, are valid:

The situation is shown in Figure 5 and the convergence condition is γ12, γ21τ> (π - Φο).

It is possible to control the value of the coefficients γ in feedback, for example by means of a process of the recursive MMSE type ("Stochastic gradient algorithm"), known and therefore not further described, minimizing the errors in the decision Si = <Ul> - Ul and 82 = <U2> - U2

where the notation <> indicates "estimated value" downstream of the decision.

In other words, the recursive algorithms can be used:

where μ is the step-size value of the process.

It should be noted that a relatively low probability of error (roughly better than 10%) is required to allow the convergence of the MMSE algorithms. This involves the need to enable Γ algorithm starting from values of γ already close to those of convergence, values to be identified and stored for example by means of a self-training routine upon initial system power-up.

Mathematically it holds that in the cancellation condition involving γ12, γ21 (π - Φο), we have:

LxNSHJf mode) or diversity in reception (FIGURE 61

To achieve a combination in phase (maximum power with simple diversity in reception), it is necessary to adjust the combination phase of the signal RI received by the first antenna with that of the signal R2 received by the second antenna. Being S2-0 for this mode, it is a question of phase-shifting the signal of the second block antenna and <jY> and making γ * ^ (-Φο) converge, if a sum of the two signals is then carried out. If, on the other hand, a difference is made between the two signals, it is necessary to realize γ ~> (π-Φο), this solution more convenient as it coincides with the convergence of the case 2xNS (2,2) (figure 6).

Mathematically, the following applies:

| U1 | = 1 wo * [SI- Sle <i (Cw >>] |

maximum for γ 4 (π- Φο) so

| U1 | = 2wo * YES.

In general, the main functions to be guaranteed for a good functioning of the system during the switching of the physical levels are the following: "Hitless" switching: the switching of the physical level must occur instantaneously and without transients involving possible error packets (that is, they must be of hitless type). From this point of view, the modulation and / or coding format changes (with the same MIMO Mode) are irrelevant, while the MIMO Mode changes must be appropriately managed.

Continuity of controls: the various loops that control the variables involved must not be subject to discontinuities in correspondence with the MIMO mode switches. That is, it is necessary to continuously maintain control of all the variables involved, regardless of the MIMO mode in use and without disturbing the relative control mechanisms. The control loops to manage are the following:

- AGC1, AGC2

- Control of a

- Controls of γ12 and γ21.

For this purpose, the methodology proposed here is to prepare service fields in the segmentation of the data, reserved for appropriate unique words (UW = Unique Word) always present in the frame, on which to calculate the various error signals necessary for the operation of the loops. control above.

Complete independence (orthogonality) is achieved between the various loops and, consequently, their continuity of operation, regardless of the MIMO Mode in use and without any disturbance in correspondence with the physical level switches.

Therefore, two UWs are foreseen for each street with the following purposes:

U identical on the two channels TI and T2, for the control of a and other "housekeeping" functions (frame start, Link identification)

To two different words (Al and A2) on the two channels T1 and T2, for the controls of AGC1, AGC2, γ! 2 and y21.

Considering QAM modulation formats with number of states NS <~> 2 <2L> (therefore 4QAM, 16QAM, 64QAM, ...) it is convenient for example to adopt a coding structure with independent and equal Codecs in parallel, each operating on a pair of bits (the 4QAM format activates only 1 Codec, the 16QAM activates 2, the 64QAM activates 3,.)

Furthermore, considering that the remaining overhead (for error correction (FEC), MAC telemetry and EOW End-to-End service channel) is concentrated in two further fields (respectively FEC and SRV-services) the suggested frame on the 2 transmission paths it is as shown in Figure 7. Different (but equivalent) segmentation choices are always possible in compliance with the discussed functionalities.

It should be noted that the SRV services field can only be provided for the two most significant bits (4QAM), while it is available for the payload for the least significant bits. The modulation format of the UWs (U and A) is here assumed 4QAM, but it is possible to choose an even more robust one (2PSK), to better guarantee the correct reception of the UWs themselves.

Clearly the single word U is the same on both ways because it is addressed to the control of a, which is necessary in the lxNS (2,2) mode. The words Al and A2 (different) are addressed in particular to the management of γ, which are basic for the 2xNS (2,2) mode. Apart from the Overhead mentioned above, the payload transmissions on the two channels TI and T2 are organized in detail as follows, for the various MIMO modes provided:

MIMO mode Antenna 1 (Tl) Antenna 2 (T2) 2xNS (2,2) Payload 1 Payload 2

lxNS (2,2) Payload 1 Payload 1

lxNS (l, 2) Payload 1 Tx2 = OFF

Normal techniques of mixing with known pseudo-random sequences, or of rotation in time, allow to avoid that the systematic UWs cause unwanted concentrated components in the transmitted spectra.

Figure 8 presents a detailed diagram of the transmission and reception functionalities addressed in particular to the management of MIMO mode switches. For the sake of simplicity, other functions are not shown, not specifically related to the management of the aforementioned switches, but evidently essential in the context of the radio system. The omitted functions, however well known to those skilled in the art, concern circuits such as IF / RF and RF / IF translators and related local oscillators, high power amplifiers and low noise amplifiers, IF synthesizers, QAM modulators and demodulators, AGC 1F (coarse), cable interfaces and the like.

The processing functions of the physical layer in transmission (segmentation, FEC coding, insertion of UWs and services and modulation mapping) are carried out in block 101 which prepares the data on the two transmission paths for subsequent D / A conversions and IF modulations .

Block 101 is set up under the external control of the physical layer selection command provided by the physical layer controller, not shown in the figure.

A transmission line, in the example the T2 line, is rotated with respect to the other by an angle a in block 102. This block 102 is loop-controlled from the reception side through a telemetry channel.

Leaving aside the RF functions not highlighted, in the receiving part the signals RI and R2 are finely tuned for the same output power by the amplifiers 103 and 104, which represent the amplifiers with automatic gain control AGC1 and AGC2.

The relative control signals are obtained by measuring with the detectors 106 and 107 the outgoing powers downstream of the A / D converters only in correspondence with the unique words A1 and A2 respectively. This is made possible by means of a special enabling waveform SA provided by the synchronization block 115, suitable for acquiring frame synchronization.

The control signal of the transmission phase shifter a is obtained from the comparator 105 by directly calculating the difference of the modules of the signals RI and R2, only in correspondence with the single word U, by means of a special enabling waveform Su- The sum of the signals RI and R2 provides the combined output for the lxNS (2,2) mode instead.

Subsequently, the signals are processed in an erasing structure represented by the sum of the same signals R1 and R2 with their respective out-of-phase and crossed versions.

The phase shifts γ12 and γ21, produced by the phase shifters blocks 108 and 109, are controlled by the MMSE processors of block 112 which process the stochastic correction signal only in correspondence with the unique words Al and A2, thanks to the appropriate SA enabling waveform.

For the lxNS (l, 2) mode, the difference between the signals RI and R2 out of phase by γ21 is also carried out.

For correct operation, the quantization and quality of the arithmetic used must be appropriately dimensioned, jointly responsible for the accuracy of the controls, and a correct synchronization of the synchronization block 115 is required. search for frame synchronism, well known from the field of numerical multiplexers / demultiplexers or decoders for block codes.

The selectors 110 and 111 finally select the signals towards the decision blocks 113 and 114 under the external control of the physical layer selection command, with the following logic:

MIMO mode lxNS (2,2) = position 1

MIMO mode lxNS (l, 2) = position 2

MIMO mode 2xNS (2,2) = position 3

The decided signals are finally processed by the processing block 116, which is also conditioned by the external physical layer selection command. Block 16 comprises the functions of demapping, FEC decoding, services extraction and serialization of the received data.

By way of example, Table 2 details the characteristics of a possible PHY Level profile to be used for the MIMO configuration (2.2).

The modulation formats include 4QAM, 16QAM and 64QAM, the latter assumed as a format of reasonable complexity. The mode of use of the PHYMIN physical layer is of the "lx" type with combined transmission and reception diversity. It also takes into account reasonable practical losses (implementation loss) and a margin (3dB) to absorb the hysteresis to be introduced in the transitions between adjacent physical levels. It is assumed to use RS (Reed Solomon) type block encoding with short block (86byte) and ability to correct 3 errors per block with hard decoding. It is also assumed a not very high frame efficiency (0.86) which takes into account the overheads for services (the two unique words UW, the word U which can also serve as a link identifier and the word A, the EOW = service channel End-toEnd, and telemetry for the MAC) and those for error correction (FEC).

The uncoded BER level = 1E-3 (which is lowered by the FEC to Coded BER »1E-7) is assumed as the BER threshold for the connection. Note that these choices are purely illustrative, and other segmentation / coding solutions may be perfectly compatible with the MIMO solution proposed here, MIMO profile (2,2) 'Φο {degrees) = 45.00 Efficiency. Brothers =; 0.86

PHY Level Muxing (S / N) avg age QAM Impl-loss Peak / avg margin bela GCOMB age RSG <dS) (dB) (dB) (dB) (dB) (dB) (net) (dB) 1x4 (2, 2) 1.00 0.54 2.00 1.50 0.00 0.00 11.04 8.34 1.72; -2.7p 1x16 (2.2) 1.00 16.53 4.00 2.00 2.55 3.00 24.08 8.34 3.44 -15.74.

2x16 (2.2) 2.00 16.53 4.00 2.00 2.55 3.00 24.08 0.00 6.88 -24.08 ^ 2x64 (2.2) 2.00 22.76 6.00 2.50 3.68 3.00 31.94 0.00 10.33: -31.94Ì! <'>! <■> Table 2 - Example of a 4-state profile for the proposed MIMO configuration (2.2) For comparison yes considers a traditional single-fronted configuration both in transmission and in reception STND (1,1) which, using the adaptivity of the physical layer with the same formats, would have the characteristics shown in Table 3.

STND profile (1, 1) 1.1; Frame efficiency = (0, & 6

PHY Level Muxing (S / N) avg age QAM ImpMoss Peak / avg margin beta GCOMB age RSG (dB) (dB) (dB) (dB) (dB) (dB) (net) (dB) 1x4 (1, 1) 1, 00 9.54 2.00 1.50 0.00 0.00 11.04 0.00 1.72 <'> - 11.04: 1x16 (1.1) 1.00 16.53 4.00 2.00 2 .55 3.00 24.08 0.00 3.44: ■ -24.08.

1x64 (1.1) 1.00 22.76 6.00 2.50 3.68 3.00 31.94 0.00 ^ 5.16; -31.94:

Table 3 - Corresponding profile for the conventional system STND (1,1) Representing the two profiles in the plane (RSG, η), the great advantages offered by the MIMO configuration (2,2) object of the present invention are immediately highlighted, both in terms spectral efficiency and available system gain at the threshold (Figure 9). It should be noted that the value of Φο influences the RSG values of the higher physical levels, while it does not appreciably change the relative gain of the threshold RSG system.

The quality of the link in systems with PHY Level Adaptivity

The quality for a connection with adaptivity of the physical level can no longer be defined only with the probability of the POUT threshold, which is recorded at the minimum physical level PHYMIN. parameter that expresses the ability of the system to provide throughput higher than the competent one at the minimum phase level PHYMIN- It is convenient for this purpose to use a parameter that measures the average spectral efficiency and provides for example a measure of how far it deviates from the efficiency spectral level competent at the maximum physical level PHYMAX.

It is therefore possible to introduce an additional parameter TL (Throughput Loss) of "relative loss of throughput" with respect to the maximum, defined simply as:

The average spectral efficiency <η>, known the form of the distribution of the additional attenuation A $ on the section, is given by the relationship:

where fi (As) represents "the efficiency tier" that the chosen physical level profile makes available.

For a cellular backhauling connection with physical level adaptivity (for example, a BER limit * = 1E-6 has been set), an excellent quality objective is for example represented by the pair of values:

POUT - 1E-5 TL - 1E- 3

The best performances deriving from the proposed invention can be fully appreciated by comparing under the same conditions and, for example, on the same length of section, the configuration object of the invention MIMO (2,2) and a conventional single antenna configuration, not already with fixed modulation, but also operated with adaptation of the physical layer on a profile using the same modulation / coding formats. STND (1,1) notation will be used for this reference configuration.

The advantages are attributable to the possibility of creating:

A higher system gain near the threshold (bad propagation occurring for very low percentages of time), tolerating a momentary reduction in throughput, and Doubling the average spectral efficiency <η> (i.e. high throughput T) under normal conditions propagation, for a very large part of the time.

It can also be argued that this combination of effects allows for a number of advantages such as:

Doubling of the throughput with the same bandwidth granted BW, which with the conventional solution would only be possible by using another RF channel (i.e., while admitting its availability, also bearing the doubling of the recurring license costs,

With the conventional solution, therefore, the cost of the hardware would also double, while with the proposed solution the hardware cost is less than double due to considerable synergies in the implementation of the MIMO solution (2,2).

Significant reduction of the PPTX power required per antenna for the same length of the covered section. This translates into savings of expensive components in the 2 RF transmitters compared to the two of the double radio solution STND (1,1), and in the associated benefits of savings in absorbed power, reduced costs of the DC / DC regulators and fewer problems related to dissipation. overall thermal (improved hardware reliability.

Reduction of the diameter of the antennas (which with the conventional double radio solution would always be 2 for each side of the connection), always with the same length of the section covered, with a consequent significant reduction of the visual environmental impact (this aspect is extremely important, for example , in urban applications).

To quantify the advantages of the proposed solution in a concrete case, the performance of MIMO (2,2) solutions with 2-state profiles in the 23GHz range is evaluated here, comparing them with the conventional STND (1,1) solution that uses the same formats modulation.

The 2-state profiles in fact lead to the maximum simplification of the MAC and do not appreciably reduce the performance compared to the 4-state profiles.

For example, it can be estimated that a 4-state profile, compared with the same extreme states (PHYMIN and PHYMAX equal) and other conditions being equal with a 2-state profile, involves about a halving of the TL value which is not very important for the overall sizing of the system (it would only allow a reduction of the spacing s between the antennas of about 10%, therefore quite marginal).

The performances of the configurations listed in Table 4 are then compared System ED Type PHYMIN PHYMAX MIMO mode

to the PHYMIN

REF STND (1,1) 1x4 (1,1) 1x64 (1,1) -

A MIMO (2.2) lx4 (l, 2) 2x64 (2.2) Simple diversity RX B MIMO (2.2) 1x4 (2.2) 2x64 (2.2) TX and RX diversity C MIMO (2, 2) 1x16 (1.2) 2x64 (2.2) Simple diversity RX D MIMO (2.2) 1x16 (2.2) 2x64 (2.2) TX and RX diversity Table 4 - 2-state profile configurations compared

For the distribution of the additional attenuation As (at 23 GHz due only to rain) the form provided in ITU-R P.1057-1 "Probability distributions relevant to radio wave propagation model ing" is used.

Setting X - (As / A0> 0i%) ratio between As and the additional attenuation at 0.01% of the time (both expressed in dB) we have:

CCDF (As) - 0.01 * exp {-14.62 0.02326 * sqrt [395145.16 ~ 99011.16 * Ln (8.33X)]} valid for latitudes greater than 30 degrees, where the CCDF notation indicates "Complementary Cumulative Distribution".

As regards the characteristic system parameters, the following assumptions are made:

Segmentation / Coditka

For BW = 3.5MHz and PHYMIN 4QAM: frame length = 86 bytes of 64Kbps,

with Gross Throughput equal to 86 * 64 - 5504 Kbps and F3⁄4 = 2752 KHz (*) Roll-Off = 100 * (3500/2752 - 1) = 27.2%

Shortened RS codec (86.74), triple error corrector

Threshold BER: Uncoded = 1E-3, Coded? 1E-7

Overhead bytes = (6 for FEC correction, 2 for UW U, 2 for UW A, 1 for EOW, 1 for MAC) RF parameters

Gamma = 23 GHz, with channel band BW = 3.5 MHz (Fg = 2.752 MHz)

K zone (42mm / h) and vertical polarization (Κγ = 0.094, αγ - 1.043)

30 cm antennas (Gant = 34.5 dB)

NF = 8 dB

PPTX = variable in the range from 10 to 20 dBm (to consider that the HPA saturation power must be about 4/5 dB higher than the PPTX).

Quality objectives

POUT = 1E-5

TL = 1E-3

(*) The PHYMIN 4QAM [options A) and B)] would therefore carry a capacity equal to that of the products currently available (2xEl in 3.5MHz band), while options C) and D) would carry double.

The MIMO configuration (2,2) has an additional parameter to be sized with respect to the conventional configuration and that is the separation s between the antennas. As can also be guessed from Figure 4 (even if it relates to a 4-state profile) it is possible to vary s (i.e. Φο, the phase shift between crossed and direct paths) without appreciable directly affect the performance of the PH YMAx (hence the value of TL).

Naturally interested in minimizing s for obvious reasons of simplicity of installation, it is natural to adopt for Φο the minimum value compatible with the TL ™ lE-3 lens.

The performance of the various profiles examined can therefore be easily compared by plotting the minimum PPTX power trends required to meet the POUT = 1E-5 goal as a function of the length L of the connection, and

- Minimum Φο angle to meet the objective TL = lE-3 Figures 10 and 11 provide precisely these results for the various configurations examined.

We therefore observe the advantage offered by MIMO solutions with PHYMIN <=> 4QAM [options A) and B)] compared to the conventional reference system in terms of PPTX required at the same length of section, as well as of course the greater advantage of doubling the throughput .

Option D) is also very interesting with PHYΜΓΝ = 16QAM and combined TX and RX diversity, which, while behaving almost like the reference system in terms of PPTX, already provides a doubling of the minimum throughput.

A precise comparison of the limit performances for the various options considered is presented in Table 5 with the same PPTX.

PPTX = 15dBm and BRF = 3.5MHz Fs (KH /) = 2752 Configuration (1,1) STND <■■■> L; ·: <·> Y-AJ. V; - -j / B):; -;

PHY MIN lx4QAM (l, l) lx4QAM (l, 2); lx4QAM (2,2): lxl6QAM (l, 2) lxl6QAM (2.2) η ΜΑΧ (bit / Hz) 5.16 10.33 10.33 10.33 10.33 η MIN (bit / Hz) 1, 72 1.72 1.72 3.44 3.44 MAX net profit (MBps) 14.21 28.42 28.42 28.42 28.42 MIN net profit (MBps) 4.74 4.74 4.74 9.47 9.47 LMAX (Km) 5.25 5.65 6.48 4.38 5.11 Φο (degrees) NA 10jl8 12.91 6.35 8.04 Antenna separation s (m) NA 2.04 2.46 1.42 1.73 Table 5 - Limit performances with PPTX = 15 dBm

The necessary antenna separations are not particularly burdensome as Figure 12 shows not even for the less interesting option [option A) with PHYMIN <=> 4QAM (1,2)].

For PPTX values in the range from 10 to 20 dBm the separation never exceeds about 2.3m, even remaining consistently below, for values of L lower than the maximum length LMAX

Figure 12 shows two families of curves, one in the hypothesis of adopting a dimensioning at constant Φο in the range of variation of L, the other in the hypothesis of adopting in any case for each L (even lower than LMAX) the minimum sufficient separation to satisfy TL = 1E-3.

By way of example, the description of a possible realization of the invention is provided below.

The described realization is to be considered "typical", in the sense that, even if similar solutions, perhaps in part alternatives, are naturally possible, they must necessarily guarantee the practical realization of the process with the same functionalities and characteristics that are commented here.

Reference is made to the block diagram of Figure 13, in which the blocks that manage the operation of a communication direction (the GO direction) are shown numbered from 201 to 208, while the analogues are shown in broken lines and numbered from 201b to 208b. blocks for the reverse direction (RETURN direction), which is completely reciprocal.

The operation of the system is described in relation to a single direction (the GO direction), the other being completely similar.

As already discussed, the transmitted signal consists of the combination of the actual information signal (the payload) to be transferred from one terminal to the other of the section, and of a Services / Telemetry channel (obviously at a much lower speed than the payload). to be used for the transfer of one (or more) service channels (EOW = Engineering Order Wire) and remote commands / controls from one terminal to another. Ai obtains this combination of signals by means of well known techniques of numerical multiplexing, realizing a segmentation of the signal transmitted in the antenna in frames of equal duration and successive in time. It has been seen that the signal formatting is such that a very small portion (portion) of the frame is reserved for the two unique words UW, in the example U and A, a subsequent portion for the FEC error correction and finally one for the SRV services, i.e. the channel EOW and, in fact, to the telemetry channel. The remaining and predominant portion is reserved for the actual payload transfer.

Overall, the portions dedicated to services represent an "overhead" (OH) with respect to the payload (PD) and the ratio between the net payload capacity and the total capacity, given by the sum of the payload and overheads, is of interest: Frame Efficiency - PD / (PD + OH)

The detection of the single word U allows to synchronize the receiver on the frame timing. The creation of the frame "in the air" or, as it is also called, of the insertion ("bit insertion") of the services and of the telemetry in transmission, is carried out by the transmission processor of physical level 201, which also includes the functionalities coding, double modulation and distribution for the realization of the physical levels in transmission. The processor 201 generates two signals modulated towards the double radio frequency transmitter 202.

The selection in real time of the physical layer and of the MIMO mode is carried out under the control of the receiver 203 of the telemetry channel 208 adapted to activate the activity in transmission.

The two modulated signals are then translated to the RF transmission frequency, by IF / RF conversion and suitably amplified by the transmitter 202 comprising the two RF transmitters which feed the relative antennas.

At the receiving side, the signals collected by the two receiving antennas are amplified at low noise and translated in frequency, by means of RF / IF conversions by the double radio frequency receiver 204.

Subsequently, the signals are processed by the receiving combiner / canceller and receiving processor of the physical layer 205 having double demodulation, decoding and telemetry extraction functions.

For a correct operation of the connection, avoiding errors in the switching transients between the various physical levels, it is necessary to dynamically ensure, in real time and in a fine manner, the congruence between the physical layer selected in transmission and the predisposition of the combiner 205 upon reception to receive the same physical level.

This physical layer TX / RX synchronization feature is ensured by keeping the two physical layer processors 201 and 205 in mutual tracking, via the remote telemetry channel.

The propagation conditions on the section are evaluated in real time by measuring the quality of reception (in practice the BER error rate that is being committed), inferred from the value of an indicator linked to the BER itself by a monotonic dependence, in the following called for brevity CSI (Channel State Indicator).

The following criteria can be used as CSI which can be obtained from real-time measurements carried out in reception on signals available in block 205:

- Ratio (S / N) in reception, obtained from the simple measurement of the two received powers S, since the noise N is known and in any case constant for each BW channel band option

- Average number of corrections in the unit of time made by the reception decoding (FEC), if available as a component of the physical layer

- MSE (mean square error) to the decision, ie downstream of the receiving AGCs which bring the level of the received signals S back to a constant reference value.

The CSI measurement is performed in the CSI 206 meter and supplied to the adaptability controller 207. The latter which incorporates the functionality of link adaptivity controller, receive adaptivity actuator and telemetry channel transmitter, is able to manage the adaptability of the connection by checking:

remotely the far transmission side through the telemetry channel, e

- locally the receiving side near.

The controller 207 is also supplied with external parameters, normally configurable via software, which represent appropriate threshold values with which to compare the current measurement of the CSI to derive dynamically and in real time the criterion for choosing the physical layer and consequently command its switching on the section. It should be remembered that it is necessary to synchronize the implementation of the physical layer switching on the transmission side, carried out by the processor 201, and the corresponding setting on the reception side, carried out through the combiner 205, in order to avoid committing error packets at each switching. This requires the use of an "ad hoc" communication protocol on the telemetry channel that appropriately controls the means of access to the transmission medium (MAC = Medium Access Control), represented precisely by the synchronized and congruent choice of the physical layers on both sides of the link.

The requirement of synchronization in the physical layer change can be satisfied, for example, by locating the instants of possible switching exactly at the beginning of each frame, cyclically numbering the frames and communicating to the processor 201 the frame number chosen for the subsequent switching. For added security, a backward confirmation message can be provided.

For greater clarity, Figure 14 shows a typical physical layer adaptation process for a two-state profile on a hypothetical link having KK = 50dB. The characteristics of the profile are summarized in Table 6.

Order Type of Net efficiency RSG AS = KK + RSH of PHY Level PHY Level η (bit / Hz) (dB) (dB)

1 (PHY MIN) lx4 (2.2) 1.72 -2.70 47.3

2 (PHYMAX) 2x64 (2.2) 10.33 -31.94 18.06 Table 6 - Parameters of a 2-state profile on a hypothetical section (KK = 50dB) Figure 14 shows in particular the trend of the ratio Reception S / N, as a function of the propagation condition on the section and more particularly as a function of the additional attenuation As (dB) that is determined on it.

As As increases with the worsening of propagation, the S / N ratio available at reception decreases linearly in dB, which is supposed to be monitored in real time and sufficiently precisely by the CSI 206 meter. As soon as the S / N value drops below a predetermined threshold s2, with s2 = [p2 (avg) m) dB

where

- m represents an appropriate safety margin before reaching the threshold BER e

- p2 (avg) is the value of (S / N) avg of the maximum physical level PHYMAX,

the controller 207 decides the switching of the physical level towards the more robust level (PHYMJN), communicating to the remote terminal 201 the first progressive frame number at which to implement it and similarly setting up the local receiving terminal205 for the same instant (beginning of the same frame) .

The switching occurs when As exceeds the C value (about 20 dB in the example), the instantaneous throughput is drastically reduced but the link threshold is lowered and is reached for a value of As (about 47 dB in the example) so the available S / N avg value becomes equal to pl (avg), ie the minimum level threshold.

It is assumed to operate at constant peak power PPTX on transmitters 202, therefore the value of (S / N) avg increases instantaneously in the passage to PHYMIN of the "static" ratio ΔΡΤΧ between the average powers of the two modes (in this case ΔΡΤΧ = 3.68dB, ratio between 4QAM and 64QAM).

In the inverse passage, i.e. when As decreases with the return to better propagation conditions, it is appropriate to provide for the new mode switching, i.e. the return to PHYMAX for a different and higher value of (S / N), creating an appropriate hysteresis in the execution physical level switching in both directions (As increasing or decreasing). This solution makes it possible to avoid possible and useless oscillations between the two levels due to “almost stationary” or very slowly variable propagation conditions.

It should also be noted, as shown by way of example in Figure 14, that it is possible to adopt the adaptivity of the physical level together with the automatic control of transmitted power (ATPC = Automatic Transmit Power Control).

The ATPC technique is well known and allows to dynamically reduce the value of the transmitted power PPTX, for particularly low values of As, i.e. in good or normal conditions of propagation (conditions that occur for the most part of the time), with the large benefit of reducing the level of interference generated on other links reusing the same RF frequencies in nearby geographic areas. It is also possible to create a "joint" control of adaptivity on the section (i.e. of ATPC combined with the adaptivity of the physical layer) by assigning its integrated management to the controller 207, and creating an appropriate MAC protocol to manage the double functionality through the same channel 208 telemetry.

With this form of “Combined Adaptivity”, it is thus possible to distinguish in the most general way various automatic operation zones on the section, as shown in Figure 14.

1) ATPC zone: for low values of As (from 0 to point A) the value of PPTX increases proportionally with As- The physical layer is the most complex, PHYMAX, and the value of PPTX is controlled in a loop through the telemetry in order to keep the value of (S / N) in reception constant and equal to a value β * = [32 (avg) m *, greater with an appropriate margin than P2 (avg).

2) Zone at maximum physical level PHYMAX: for values of As between A and B.

In practice, it is a separation zone between the ATPC zone and the subsequent zone 3), and is operated at the maximum physical level PHYMAX with constant PPTX and equal to the maximum available power.

3) Hysteresis zone: for values of As between B and C, in which the physical level switching takes place for suitably different values of (S / N), depending on the direction of variation of As, increasing or decreasing.

4) Zone at minimum physical level PHYMIN: for values of As beyond point C, in which the physical level is now PHYMIN and the PPTX is the maximum available. As can continue to grow until it enters the next zone 5).

5) Zone of unavailability: for As beyond point D, that is for (Y / N) that has now fallen below the threshold of the minimum physical level PHYMIN = pl (avg). In the example of Figure 14 this area begins with As greater than about 47 dB, so it has a very low probability.

As can be appreciated from what has been described, the point-to-point radio system and the method for operating the radio system according to the present invention allow to meet the needs and to overcome the drawbacks referred to in the introductory part of the present description with reference to the known art.

Obviously, a person skilled in the art, in order to meet contingent and specific needs, will be able to make numerous changes and variations to the system according to the invention described above, all however contained within the scope of protection of the invention as defined by the following claims.

Claims (24)

CLAIMS 1. Point-to-point radio system comprising a plurality of transmission antennas and a plurality of reception antennas, said antennas being usable with an operating mode selected from a plurality of operating modes available and being intended to transmit and receive single signals. carrier and transmission frequency greater than 6 GHz, said signals having a modulation and coding format chosen from a plurality of modulation and coding formats available, said antennas operating mode and said modulation and coding format jointly identifying a physical transmission layer , said system being characterized in that it comprises, in transmission and reception, signal processing means coupled respectively to the input of the transmission antennas and to the output of the reception antennas and adapted to switch the physical level of transmission as a function of the quality for receiving the signal received by said plurality of receiving antennas. 2. Point-to-point radio system according to claim 1, in which said signal processing means switch the physical layer between two or more transmission formats with the same channel bandwidth as the signal reception quality varies. Point-to-point radio system according to claim 1 or 2, wherein said plurality of antenna operation modes comprises simple diversity mode in reception, in which a transmitting antenna is off, combined diversity mode in transmission and reception and spatial multiplexing modes. 4. Point-to-point radio system according to any one of claims 1 to 3, wherein said plurality of modulation and coding formats comprises different modulation formats with the same coding / decoding schemes (FEC) used, modulation formats same but different coding / decoding schemes (FEC) used, and different modulation formats and simultaneous diversity of coding / decoding schemes (FEC) used. 5. Point-to-point radio system according to any one of claims 1 to 4, operating in propagative environments that mainly cause non-selective attenuation in the band, 6. Point-to-point radio system according to claim 5, wherein said propagative environments comprise high frequency links with rain attenuations or attenuations due to multiple paths with echo delays of small duration with respect to the inverse of the channel band . 7. Point-to-point radio system according to any one of claims 1 to 6, in which said processing means switch the physical level by measuring in real time a quantity linked in a monotonic way to the instantaneous error rate of the signal received at the antennas recipients. 8. Point-to-point radio system in accordance with any one of claims 1 to 7, in which the channel quality is estimated by measuring at least one variable chosen between the signal to noise ratio in reception, obtained from the measurement of the two received powers, number mean of corrections in the unit of time carried out by the reception decoding complex, mean square error at the decision for each of the demodulators coupled to the receiving antennas. Point-to-point radio system according to any one of claims 1 to 8, in which the switching between the different operating modes of the antennas is carried out using a method based on several independent unique words. 10. Point-to-point radio system in accordance with claim 9, in which the use of several independent single words is aimed at maintaining the continuity of control over the various automatic adjustment loops of the variables necessary for the correct functioning of the various operating modes of the antennas, and realize the mutual independence of these regulations and their total insensitivity to physical level switching. 11. Point-to-point radio system according to any one of claims 1 to 10, in which said antennas are also used on orthogonal polarizations. Point-to-point radio system according to any one of claims 1 to 10, wherein said system is used together with a second point-to-point radio system operating on a polarization orthogonal to the polarization of said point-to-point radio system to quadruple the transmission capacity. 13. Method for operating a point-to-point radio system comprising a plurality of transmission antennas and a plurality of reception antennas, said antennas being usable with an operating mode selected from a plurality of operating modes available and being intended for transmit and receive signals with a single carrier and transmission frequency greater than 6 GHz, said signals having a modulation and coding format chosen from a plurality of modulation and coding formats available, said antennas operating mode and said modulation and coding format identifying jointly a physical layer of transmission, said method comprising the switching, in transmission and reception, of the physical transmission level as a function of the reception quality of the signal received by said plurality of reception antennas. 14. Method according to claim 13, wherein the physical level is switched between two or more transmission formats with the same channel bandwidth as the signal reception quality varies. Method according to claim 13 or 14, wherein said plurality of modes of operation of the antennas comprises modes of simple diversity in reception, in which a transmitting antenna is off, mode of combined diversity in transmission and reception and mode of space multiplexing. Method according to any one of claims 13 to 15, wherein said plurality of modulation and coding formats comprises different modulation formats for the same coding / decoding schemes (FEC) used, same modulation formats but different schemes of coding / decoding (FEC) used, and different modulation formats and simultaneous diversity of coding / decoding schemes (FEC) used. 17. Method according to any one of claims 13 to 16, operating in propagative environments which mainly cause nonselective attenuation in the band, 18. Method in accordance with claim 17, in which said propagative environments include high frequency connections with rain attenuations or attenuations due to multiple paths with echo delays of small duration compared to the inverse of the channel band. 19. Method in accordance with any of claims 13 to 18, in which the physical level is switched by measuring in real time a quantity linked in a monotonic way to the instantaneous error rate of the signal received at the receiving antennas. 20. Method according to any one of claims 13 to 19, in which the carial quality is estimated by measuring at least one variable chosen between the signal to noise ratio in reception, obtained from the measurement of the two received powers, average number of corrections in the unit of time carried out by the reception decoding complex, mean square error at the decision for each of the demodulators coupled to the receiving antennas. Method according to any one of claims 13 to 20, in which the switching between the different operating modes of the antennas is carried out using a method based on several independent unique words. 22. Point-to-point radio system in accordance with claim 21, in which Γ the use of several independent single words is aimed at maintaining the continuity of control over the various automatic adjustment loops of the variables necessary for the correct functioning of the various operating modes of the antennas , and realize the mutual independence of such regulations and their total insensitivity to physical level switching. 23. Method according to any one of claims 13 to 22, in which said antennas are also used on orthogonal polarizations. Method according to any one of claims 13 to 23, wherein said system is used together with a second point-to-point radio system operating on a polarization orthogonal to the polarization of said point-to-point radio system to quadruple the transmission capacity.