DE19849111C1 - Data transmission over power lines involves using 3-phase TDD transmission method with division of total number of channels into two groups with fixed direction allocation - Google Patents

Abstract

Data is transferred between repeaters (11 - 17) using three phase TDD, available channels being subdivided into two groups and repeaters being subdivided by a modulo three division into three groups, based on their order on the line. The method involves using a repeater chain and channel-oriented data transmission. Data are allocated to channels using an access method and transmitted between repeaters in duplex mode using a transmission method. The transmission method is three-phase TDD with division of the total number of channels into two groups. The repeaters are divided into three groups. The first group is sent in a first phase to the adjacent repeaters in the up and down directions; the second group is sent in the second TDD phase to the adjacent repeaters in the up and down directions; and the third group is sent in the third TDD phase to the adjacent repeaters in the up and down directions, whereby the channel groups for each TDD phase for all repeaters are permanently allocated to a transmission direction. An Independent claim is also included for a circuit for implementing the method.

Description

The invention relates to a method and a circuit arrangement for Data transmission on powerlines.

The use of powerlines for purposes of telecommunications and Data technology is due to the existing comprehensive low voltage Power distribution networks are extremely interesting because they make a separate one Cabling for data transmission would be superfluous. The existing ones Low-voltage power distribution networks not only reach every household, but also allow almost any accessibility within the Household itself. Therefore, the existing low-voltage Power distribution networks even for narrowband point-to-point Transfer such as measured value recordings, alarm systems, Voice transmissions or baby monitors used.

For broadband transmission methods, the use of the Low voltage power distribution network, however, has some technical problems on. In addition to the 230 V AC voltage, voltage peaks of up to 1000 V occur. Furthermore, the impedances fluctuate depending on from the connection and disconnection of electrical consumers, since both the electrical cables of the network as well as the connected electrical Consumers can have different wave resistances. About that In addition, the damping on the Powerlines is relatively large and moreover frequency dependent, with a characteristic value of approximately 0.05 dB / (MHz.m). This means that at 30 MHz the attenuation is 1.5 dB / m is, so that higher transmission frequencies are no longer useful. There Each line acts as a receiving antenna at the same time RF reception of the line and due to the connected consumers too high interference levels, so that correspondingly high reception levels are required. On the other hand, the permitted transmission power is limited because the line also acts as a transmitting antenna, whose radiation is below the specified  Interference voltage limits must lie, otherwise it is too unwanted Malfunctions in other facilities can occur. The required high Reception level with low transmission power result in a very low link budget, which is also in contrast to radio transmission technology cannot be improved by profit antennas. Another technical The problem is the undefined joints.

Butt joints always occur wherever cables or facilities are involved unequal impedances (wave resistances) are interconnected. Now, for example, becomes an electrical consumer by means of a spur line connected to the power distribution network, the effective results Impedance of the stub line at the point of connection from impedance and length the stub in relation to the frequency and the switched on Input impedance of the consumer. If these impedances are not matched, this leads to reflections at the joint. Since most Consumers change the input resistance with the mode of operation, it happens to dynamic changes in the reflection conditions on such Joints. Bumps therefore have two linked disturbing Properties on:

Reflected signals can be caused by the longer path Delay arrive at the receiver and disturbing interference there cause. Due to the high attenuation in the low-voltage network Assume that reflected signals are larger with a delay approx. 200 ns have no meaning, because of the associated long path lengths are already significantly damped.

The energy transfer via the joint is punctually dampened. The physical reason is that the incoming Energy split into the reflected and the continuing energy part becomes. A strong reflection also means a high attenuation for the Retransmission via the joint. It should also be noted that in the Low-voltage cables are not fed in a directional manner and also from Low voltage cable can not be received directionally, such as Repeaters in radio networks can do this using directional antennas.  

The physical conditions shown lead for broadband Transfer procedure for the following problems or requirements. The Decoupling from the operating voltage of the low-voltage network to Protection of people and communication technology is complex and costly. The through loss of the coupling goes into the link budget twice, namely when coupling and decoupling the signals. Due to the large attenuation and the low link budget are only minor Transmission distances possible, so that repeater chains (Regeneration amplifier) are necessary. Because of the frequency dependent Attenuation predistortion in the transmitter is necessary. Furthermore are with regard to the frequency response insensitive access and Transmission method to choose. The inseparable physical Medium requires a special transmission regime, which in particular Parameter links across the entire repeater chain (send / Receive level, synchronization, predistortion) and a parasitic Avoids reception of neighboring transmission links.

DE 195 04 587 A1 describes a two-way communication system for Data transmission between a control center and substations and between Intermediate stations and consumer facilities known. The Intermediate stations are linked to the low-voltage network Node controller, whereby for data transmission between the control center and the Substations a large-scale telecommunications system, for example a Radio data network or a wired network, in particular Fiber optic network is used. The distribution network transformers assigned node controllers have standard modems as Interface between the low voltage network and the large area Telecommunications network, while as an intermediate station on the Transmission path between the node controller and the end user device a modem with repeater function is provided and the transmission in local low voltage network carried out using a spreading technique becomes.  

Data transmission in low-voltage networks takes place in Europe permissible frequency range up to 148.5 kHz instead. Is in this frequency range but the transmission quality on the one hand due to a large number of here Interference signals and strong noise are limited and on the other hand by the narrowband transmission band with regard to the number of participants and the transmission capacity per participant is limited.

From the post-published DE 197 14 386 a method for Data transmission in two-way communication via low-voltage networks, which are coupled to a higher-level telecommunications network, are known where the data transmission in the low-voltage network in one high-frequency range up to 30 MHz with a spread spectrum of the data signals and a transmission level below the predetermined radio and Line interference voltage limit is carried out and ensuring a Multi-user structure with different sequences of a family of Pseudo random numbers spread signals to specify a receiver-specific logical direction in the low-voltage network with a Direction coding are provided, each of which is user-specific spread and direction-specifically marked binary data sequences in the Low-voltage network in the determined by the degree of damping Distances identified with the help of the given sequences by correlation, regenerated and with a new direction detection for forwarding the Signals are evaluated. A disadvantage of the known method is that the  Prerequisite for a safe differentiation of the different spread Signals is that the received useful signals are at least the same Reception levels must have as possible parasitic reception levels. Furthermore, the reception levels from different transmitters arrive at the same time and with the same reception level. The procedure therefore requires that the regeneration amplifier or signal connection units (Repeater) are arranged at an equidistant distance from each other. In practice However, an equidistant arrangement can hardly be guaranteed, since most of the time the possible connection points for the repeaters by external specifications are set so that it is even with small deviations due to extremely large attenuations to significant differences in the Reception levels can come, one with a certain power level sending source generated by neighboring recipients or the one Receivers from two neighboring sources. If the latter occurs, so is already at small level differences of, for example, less than 0.5 dB signal detection not possible.

The invention is therefore based on the technical problem, a method and to create a circuit arrangement for data transmission on powerlines, which without complex RF components with varying distances between Repeater simple and reliable duplex transmission on Powerlines enable.

The solution to the technical problem results from the characteristics of Claims 1, 2 and 8. Further advantageous embodiments of the Invention result from the subclaims.

This is done using a 3ph-2gr or a 6ph-1gr transmission method the problem of chaining across the entire repeater path broken up so that a particularly simple control regime results, especially with regard to the setting of the transmission levels and delays. Alternatively or cumulatively, an OFDM access method is used, by means of whose frequency response compensation only by means of a  Preemphasis on the transmission side is made possible.

The invention is based on a preferred Embodiment explained in more detail. The figure shows:

Fig. 1a-d a circuit diagram of a logical transmission scheme, a physical propagation scheme and a transport scheme for a two-phase TDD scheme with four channel groups,

FIG. 2a-b level plans a simulated repeater chain for the two TDD phases,

Fig. 3a-b synchronization plans a simulated for the two TDD phases for setting a synchronous reception situation,

Fig. 4a-b frequency response diagrams for the two TDD-phase with one-side alignment by a pre-emphasis,

Fig. A circuit diagram of a logical transmission scheme, a physical dispersion and a transport scheme for a 5a-d 4PH-2gr method,

FIGS. 6a-d a circuit diagram of a logical transmission scheme, a physical dispersion and a transport scheme for a 3-ph 2gr method,

Fig. 7 is a plan for the level 3Ph 2gr method for a TDD-phase at the repeater chain according to Fig. 4a-b,

Fig. 8 is a frequency response scheme with a pre-emphasis for the TDD phase according to Fig. 7,

Fig. A circuit diagram of a logical transmission scheme, a physical dispersion and a transport scheme for a 9a-d 6PH-1gr method,

Fig. 10, a level and for a Vorverzerrungsplan 3Ph 2gr method with a pre-emphasis in an OFDM access method,

Fig. 11 is a schematic block diagram of the transmitting side processing of the OFDM,

Fig. 12 is a schematic block diagram of the receiving side processing of the OFDM,

Fig. 13 is an initialization scheme for a 3-ph 2gr method,

Fig. 14 is an initialization scheme for a 6PH-1gr method and

FIG. 15 is a scheme for the measurement procedures during initialization.

For a better technical understanding of the invention, the transmission by means of a repeater chain 10 with a two-phase TDD method with division of the total number of available channels into four channel groups 1-4 is first explained. In Fig. 1a, the circuit diagram of the repeater chain 10 is shown schematically with seven repeaters 11-17 , which are connected to a physically non-interruptible line 18 . Repeaters 11-17 are each switched alternately in space and time as transmitters or receivers, ie in one time phase all even-numbered repeaters are switched and all odd-numbered repeaters are switched as receivers and in the second time phase are reversed. The logical transmission scheme is shown in FIG. 1b, the transmission from left to right being the down-ward and from right to left the up-ward direction. Channel groups 1 and 3 are used for transmissions in the up-ward direction and channel groups 2 and 4 for transmissions in the down-ward direction. A white circle symbolizes that the respective repeater is switched in phase as a receiver. Accordingly, a black circle means that the repeater works as a transmitter. Why a division into four channel groups 1-4 is necessary can be seen from the physical propagation scheme according to FIG. 1c. Since the line 18 cannot be physically separated, the parasitic reception situations shown in broken lines also occur. This reception situation for a repeater means that all four channel groups 1-4 arrive, of which only two channel groups carry useful data. However, all incoming channels must come from the pool of mutually orthogonal signals. If this were not the case, for example for the parasitic channel groups, these would prevent the reception of the useful signals in the receiver as a disturbance variable. It follows, therefore, that the pool of orthogonal signals must be distributed across these four channel groups. The repeater 13 must therefore receive, for example, a broadcast from up-ward (from repeater 14 ) and one from down-ward (from repeater 12 ). At the same time, however, the parasitic upward transmission of repeater 12 (which is intended for repeater 11 ) and the downward transmission of repeater 14 (which is intended for repeater 15 ) arrive. The corresponding transport scheme is shown in FIG. 1d, the thick black arrow representing the transport of a commercial item down-ward and the thick dashed arrow representing the transport of a commercial item up-ward. In a first phase, the useful broadcast is sent downward by a telecommunication device in channel group 4 to the repeater 11 and received by it. In the next phase, the repeater 11 sends the useful broadcast via the channel group 2 to the repeater 12 . The repeater 12 also sends in the next phase via channel group 2 and so on. In the ideal case, ie with a constant distance between the repeaters 11-17 and a constant, frequency-independent attenuation, this transmission method would represent a satisfactory solution. Under real conditions, however, this transmission method leads to considerable problems, which will be explained in detail using an example.

A CDMA access procedure is assumed, assuming that the CDMA signal has a center frequency of 20 MHz with a bandwidth of 10 MHz. With a real attenuation of approx. 0.05 dB / m.MHz results At a length of 20 m there is an attenuation of 15 for the lower band limit dB and for the upper band limit an attenuation of 25 dB, which is one Attenuation difference of 10 dB corresponds. At a distance of 40 m there is an attenuation difference of 20 dB. such Frequency responses are for common modulation methods such as QPSK (Quatro Phase Shift Keying) and the detection of CDMA signals not permitted. With point-to-point transmissions, it is known to insert corresponding equalizations in the transmission path. this happens mostly on the transmitter side in the form of pre-emphasis. Provided one repeater simultaneously to two or more neighboring repeaters in must send different distances, its predistortion can only can be optimized for a receiving repeater. Provided a repeater receives signals from two or more neighboring repeaters simultaneously and the residual distortions of the input signals are not identical, one can  Equalization does not take place at the receiver input as this applies to all Input signals acts the same. Another disadvantage of using CDMA is that both the pre-emphasis and the input equalization in the RF Level using frequency-selective components. The Effort is particularly increased if automatic adaptability is available the application with a high setting range is required because machining in the signal level is not possible because the CDMA signals in the code level available. This disadvantage is also not avoided, if by means of a other transmission methods, the residual distortions are avoided.

The transmission levels are dimensioned in such a way that a reception level is generated at all neighboring stations to which transmission is carried out, which is the same as the other reception levels present there, but is at least equal to the minimum reception level Rxmin. This creates a link across the entire repeater chain 18 , according to a scheme as shown in Table 1 below for a time phase.

Table 1

The table shows in the column transmit level that with an arbitrarily assumed transmit level P0 of a first repeater the transmit level of the third repeater is already fixed if both are to produce the same receive level at repeater 2 . In this way, the chaining continues across the entire repeater chain. The transmission levels generally have the form P0 + Σ (Ai), where Σ (Ai) is the addition of the line attenuation with alternating signs according to the scheme in the table, column transmission levels. Then follows for the minimum and maximum transmission level in a repeater chain

Pmin = P0 + Σ (Ai) min ≦ P ≦ Pmax = P0 + Σ (Ai) max,

which is the span of the transmit level with

Pmax - Pmin = Σ (Ai) max - Σ (Ai) min

results. Simulations with a chain of 20 repeaters have shown that the Differences occurring can assume values of up to 100 dB.

The table also shows in the Reception level column that with an arbitrarily assumed transmission level P0 of a first repeater and the subsequent linking of the transmission level over the entire repeater chain, the reception level for the even-numbered repeaters is generally in the form P0 + Σ (Ai), whereby Σ (Ai) is the addition of the line attenuation with alternating sign according to the scheme in the table, column Reception level. Then follows for the minimum and maximum reception level in a repeater chain:

RXmin = P0 + Σ (Aj) min ≦ P ≦ Rxmax = P0 + Σ (Aj) max,

which is the span of the reception level with

Rxmax - RXmin = Σ (Aj) max - Σ (Aj) min

results. Simulations with a chain of 20 repeaters have shown that the Differences occurring can assume values of up to 100 dB.

In addition, there is the task of dimensioning P0 so that Rxmin assumes a value that corresponds to the actual minimum input level. The associated control regime is complex and critical, especially when automatic self-balancing is required. Since P0 is unknown when it is started up for the first time, the entire repeater chain 10 must always be readjusted if the top-down setting chain for a repeater falls below Rxmin. Furthermore, changes during operation can lead to the repeater chain having to be adjusted again. Additional problems arise with branches. Make-ups are not possible because this contradicts the exclusive nature of the top-down comparison. The link budget is further restricted by the dynamics of the transmission and reception levels, with Rxmin being determined by the receiver sensitivity or the channel noise. Pmax, on the other hand, is limited by the hardware, power supplies or other specified restrictions, in particular by the requirement to comply with the radio and line interference voltage limits. For the link budget, only either Pmax - RXmin - (Σ (Ai) max - Σ (Ai) min) or Pmax - RXmin - (Σ (Aj) max - Σ (Aj) min) is available, depending on the dynamic range of the is bigger. This can lead to the impossibility of dimensioning.

A DS-CDMA (Direct Sequence CDMA) requires that the broadcasts of neighboring repeaters with a receiver repeater with high Accuracy arrive in sequence. Depending on the parameter of the CDMA, accuracy requirements on the order of 20, for example ns occur. Due to the different distances between the Repeaters is the synchronization to a network-wide uniform Phase start time not sufficient. The need arises Starting times of all sending repeaters with delays, based on the Start time, to ensure that the above condition is met becomes. This is for the two TDD phases in the two following Tables 2 and 3 are shown.  

Table 2

Table 3

Table 2 shows in the column transmission time that with an arbitrarily assumed transmission time V0 of a first repeater, which should be identical to the phase start time, the transmission time of the third repeater is already fixed if the same reception time is to result at repeater 2 . In this way, the chaining continues across the entire repeater chain. This is shown as an example in Tables 2 and 3 for the two time phases of TDD operation. If the track lengths cannot be freely selected, no influence can be exerted on the amounts of the sum of the signal transit times Ti. At v ≈ 10 ns / m, these amounts can assume considerable values even with small differences in route lengths. The transmission times generally have the form V0 + Σ (Ti), where Σ (Ti) is the addition of the running times with alternating signs according to the scheme in Tables 2 and 3, column transmission time. Table 2 also shows in the Receiving column that, with an arbitrarily assumed transmission time V0 of a first repeater and the subsequent linking of the transmission times over the entire repeater chain, the reception times for the even-numbered repeaters generally have the form V0 + Σ (Ti). In Table 3, the same consideration is made for the other time phase of the TDD, the time V0, which should be identical to the phase start time, being arbitrarily assumed for the reception time of the first repeater, so that this repeater both transmits and synchronously with the phase start times also receives. It follows that the transmission and reception times of each repeater are offset from V0 by + Σ (Ti) and -Σ (Ti). So the difference goes 2. | Σ (Ti) | on the protection time between the TDD phases and must be taken into account when dimensioning them. The control regime for synchronization is relatively complex since a top-down setting chain has to be implemented. With every change during operation, a new adjustment from the event location downward over the rest of the chain is required. Meshes are not possible as this contradicts the exclusivity of the top-down comparison.

To illustrate this chaining problem, the level diagram for the two TDD phases of a chain consisting of 19 repeaters is shown in FIGS . 2a-b, wherein an attenuation of 1 dB / m was assumed, so that the distance in meters corresponds to the attenuation and Distances between the repeaters were randomly chosen between 16-60 m. The levels are also standardized to Rxmin, ie P = 0 in the illustration corresponds to the minimum reception level Rxmin. If, for example, about 78 dB is selected on the first repeater P0, the reception level P0 - 28 dB is set on the second repeater. Accordingly, the sending third repeater transmission level with P0 - 28 + 37 dB must be selected so that the reception level at the second repeater switched as the receiver is the same. From the level plans you can now see that changes at the beginning of a chain have a direct impact on the dimensions of the other repeaters. It can also be seen that, for example in FIG. 2a, the transmission power can fluctuate from over 100 dB to 40 dB.

In FIGS. 3a-b, the synchronization scheme for the two TDD repeaters phases of the chain is shown in FIG. 2a-b shown, which were normalized relative to a start time point 0.. The values of the start times for sending are each shown as white bars. The values of the start times for reception are shown as black bars. The start times for sending are dimensioned such that the signals arrive synchronously from both directions in the receivers. The relative starting times of the two phases are shifted from each other by the duration of one phase. It becomes clear that for the individual repeaters the amounts of the delays for sending and receiving are in each case the same size in the two phases, but have opposite signs. The diagram also shows that the difference in the delays can be up to 1400 ns. However, this time goes directly into the protection times between the TDD phases. As can additionally be seen from FIGS . 3a-b, as with the levels, there is a direct connection between the individual repeaters in the dimensioning of the individual delays, so that local changes always affect the entire repeater chain.

In FIGS. 4a-b show the frequency response compensation is shown for repeater chain in the two phases. The preemphasis values are shown as black bars. The residual distortion for downward is shown as white bars and that for upward as gray bars. As can be seen from FIG. 4a, the residual distortion can be corrected to zero in some places, but considerable residual distortion remains in some places, which cannot be tolerated.

In general, it can thus be stated that such a transmission scheme with 2-TDD phases and four channel groups very structured at first glance and looks simple, but especially in applications where the Lengths of track cannot be freely selected, which can in some cases cause insoluble problems can lead. This method is for universal use on Powerlines not suitable with CDMA.

Furthermore, a transmission method with four TDD phases with division of the total number of available channels into two channel groups 1 , 2 can also be considered, which is shown in FIGS. 5a-d. The repeaters 11-17 are again spatially and temporally alternately switched as a transmitter and sometimes as a receiver, a repeater being connected for two TDD phases, for example as a transmitter, and transmitting downward in one phase and upward in the other phase. The transmission levels are dimensioned in such a way that a minimum reception level Rxmin is generated at all neighboring stations to which user data is transmitted. Furthermore, it must be ensured that the received level of the useful signal is greater than / equal to that of the parasitic signal received. The reception level of the parasitic signal is predetermined by these two conditions. If a reception level is set higher than the minimum reception level, then there is a chain again. If, on the other hand, the minimum reception level happens to fall below, the chaining is broken at this point. In principle, however, both transmission methods have the same problems, since each receiver receives useful and primary signals from different transmitters, so that with regard to the problems of the reception level, synchronization and frequency response compensation, reference can be made to the previous explanations regarding the 2ph-4gr method.

A significant improvement results from a transmission method with three TDD phases and two channel groups, which is shown in FIGS. 6a-d. As shown in Fig. 6b, the logical transmission scheme is that the connected repeaters 11-17 according to their order on the line 18 modulo 3 in three groups 11 , 14 , 17 ; 12 , 15 ; 13 , 16 ; are divided. In a first TDD phase, the repeaters of the first group 11 , 14 , 17 send to the repeaters 12 ; 13 , 15 ; 16 . In the subsequent TDD phase, the repeaters of the second group 13 , 16 send to the repeaters 12 , 14 ; 15 , 17 . In the subsequent TDD phase, the repeaters of the third group 12 , 15 send to the repeaters 11 , 13 ; 14 , 16 . There is a fixed assignment of transmission direction and channel group 1 , 2 in each TDD phase.

In each TDD phase, each transmitting repeater and its immediately adjacent repeaters form a kind of transmission island, so that the chaining problem of the repeater chain 18 is broken up. The dimensioning of the transmission levels takes place in such a way that a reception level greater than or equal to the minimum reception level Rxmin is generated at all neighboring stations. For this purpose, the transmission level is selected such that the condition of Rxmin is still met at the receiving station with the greatest distance. Accordingly, a reception level is set at the neighboring station in the shorter distance, which is above the Rxmin by the difference in the attenuations of the two transmission directions. Due to the very high attenuation, the parasitic transmissions from the next transmission repeater can be neglected. In the case of non-freely selectable line lengths, the problems of frequency response compensation can be solved better with this transmission scheme than with the two aforementioned methods, although not yet optimally. A transmitter preemphasis can also only be optimized for one receiver at a time. On the other hand, this method allows an equalizer amplifier arranged on the receiving side, since within a TDD phase, a repeater connected as a receiver always receives signals from only one transmitter, which enables perfect equalization.

In FIG. 7, the level plan for one of the three phases of TDD is shown. The transmission levels are shown as black bars and the reception levels as white or gray bars. In addition, the parasitic reception levels based on the next but one transmitter are shown. In FIG. 9, the first transmission level P0 is set to 37 dB above Rxmin, so that the repeater downward receives a reception level equal to Rxmin from the transmitter repeater, so that no bar can be seen. The upward repeater receives a received signal of 37 - 28 = 9 dB. The next transmission level is again only designed for its two receivers, with the greater path loss being downward. Therefore the transmission level P1 = RXmin + 42 db is selected. As a result, its downward receiver receives a reception level Rxmin and its upward receiver a reception level of Rxmin + 42 - 21 dB. The first downward receiver therefore receives a parasitic signal of RX + 42 - 21 - 53 = RXmin - 32 dB. On the one hand, it will be clear that three repeaters together form a transmission island and that their transmission levels can be dimensioned completely independently of neighboring transmission islands. Furthermore, the parasitic reception levels are attenuated in such a way that they are negligible. The same applies to the other two TDD phases, so that a separate presentation has been omitted. Due to this breaking of the chaining over the entire route, the synchronization problem no longer arises, since it is always only a point-to-point connection, since a receiver always receives useful and parasitic signals from one transmitter, which is easy to see from Fig. 6c. FIG. 8 shows a level diagram for the frequency response compensation for the TDD phase shown in FIG. 7. The pre-emphasis in the transmitter is set such that complete equalization occurs in one of the two receivers. The distortion remaining at the other receiver is shown as a white bar and can be compensated for by a corresponding receiver-side equalizer amplifier. With this 3ph-2gr transmission method, all major problems with duplex transmission to Powerlines with CDMA can be solved even at different distances between the repeaters. Another special feature of the transmission method is that the transport speed to upward is twice as high as to downward or vice versa. This is very easy to see if you look at the thick solid and the thick dashed arrow in the transport scheme according to Fig. 6d. It should also be noted that the existing total number of channels can also be divided unevenly between the two channel groups 1 , 2 .

In FIGS. 9a-d, an alternative solution in the form of a transmission method is shown with six TDD-phase and a channel group. As with the 3ph-2gr transmission method, it is again a modulo-3 method, ie every third repeater in the chain behaves the same. In a first TDD phase, the repeaters of the first group 11 , 14 , 17 send to the neighboring repeaters 12 , 15 in the down direction. In the second TDD phase, the repeaters of the first group 11 , 14 , 17 send to the neighboring repeaters 13 , 16 in the up direction. In the third TDD phase, the repeaters of the second group 12 , 15 send to the neighboring repeaters 13 , 16 in the down direction. In the fourth TDD phase, the repeaters of the third group 13 , 16 send to the neighboring repeaters 12 , 15 in the up direction. In the fifth TDD phase, the repeaters of the third group 13 , 16 send to the neighboring repeaters 14 , 17 in the down direction and in the sixth TDD phase, the repeaters of the second group 12 , 15 send to the neighboring repeaters 11 , 14 in up direction.

The dimensioning of the transmission level is such that the receiver is received in the direction of use Rxmin. Since the process is only point-to- contains point connections, the chaining problem is complete broken up and the dimensioning of the transmission level can be independent of  the other repeaters. The same applies to the Synchronization, since only one transmitter uses signals at a time synchronization is unnecessary. An advantage over The 3ph-2 gr transmission method is that frequency response compensation can only be realized with a preemphasis, so that on receiver-side equalizer amplifier can be omitted. Just like the 3ph-2gr transmission methods are possible because the Chaining has broken. If necessary, the The number of repeaters in the mesh should be added so that the modulo 3 Classification is guaranteed.

It should also be noted that the transmission capacity in the latter two methods is somewhat larger than in the 2ph-4gr or 4ph-2gr transmission methods. The total capacitance K is defined as K = k. f _ph . D, with k-number of channels, f _ph -phase frequency and D-bit per phase and channel, whereby phase frequency and number of bits per phase and channel are used for the same consideration for all methods. The available capacity per direction results with K _R = a. (K / R) and a = k _R / k, where k _{R is} the number of channels per direction and repeater and R is the number of phases. If the number of channels k is divided equally into groups of equal size, a = 1 / G and thus K _R = K / (G. R). The transmission capacity is thus larger in the ratio 116 to 1/8.

The delay times result from T _V = V. T _ph with V delay per repeater, ie after how many phases a received data packet is retransmitted and T _ph duration of a phase. Assuming the phases have the same duration, the following values result from the transport schemes in FIG. 1d for the 2ph-4gr method, in FIG. 5d for the 4ph-2gr method, in FIG. 6d for the 3ph-2gr method and can be inferred in Fig. 9d 6PH-1gr for the method, with V = 1 for both directions in the process 2ph-4gr, V = 2 for both directions in the methods 4PH-2gr and 6PH-1gr and V1 = 1, V2 = 2 , ie different for both directions in the 3ph-2gr method.

It is easy to see that the 2ph-4gr- Transmission method advantages over the others Has transmission method.

With all previous considerations it has been assumed that a CDMA method is used. As with the disadvantages previously A particular disadvantage of the CDMA is that the signals in the code and not in the frequency plane, which in particular is disadvantageous in terms of frequency response compensation.

A significant simplification of frequency response compensation and others Advantages arise if an OFDM / OFDMA instead of a CDMA method (Orthogonal Frequency Division Multiplex / Orthogonal Frequency Division Multiple Access) method is used, regardless of which the aforementioned transmission method is used. The OFDM The method uses orthogonal frequencies for the transmission. These are characterized in that it is an integer multiple of one Are base frequency, have the same constant amplitudes and with the Symbol duration are correlated so that they begin with a symbol duration same start phase ωt = 0 and end with the symbol duration with Phase ωt = 2π end. One of these frequencies becomes a channel assigned. The signals of all channels are superimposed on one another, so that in a broadband transmission signal a variety of channels (Multi-carrier process) can be realized. An advantage of this procedure consists in the fact that a simple complete processing can be realized using computing technology, for example using known digital signal processors DSP. there the digital symbols of all parallel channels are transmitted as Signals viewed in the frequency plane. Using an inverse Fourier transforms this into the signal in the time plane for each channel calculated. The signals obtained in this way are in a summing stage summed up. This sum signal can additionally, for example through a mixer, into a different frequency range, usually through an up-  Conversion. At the receiving end, the signal if necessary after the necessary down-conversion by means of a Fourier transformation from the time level back to the frequency level converted back, so that the digital symbols of all channels again parallel to To be available.

For the transmission to power lines, the property of the OFDM method is used that all signals (carriers) are not overlapping in the frequency level. This also applies if multi-phase modulation is applied to the individual carrier. The advantage is that only a pre-emphasis on the transmission side is sufficient for the channel correction and that this can be carried out at the signal level. For the predistortion, the signals are provided with an amplitude weighting which correlates with the signal frequency before the inverse Fourier transformation, as a result of which elaborate solutions in the RF level, as in CDMA, are dispensed with. This will be explained in more detail below with reference to FIG. 10, in which the conditions for a 3ph-2gr transmission method are shown, in which a repeater transmits simultaneously to two neighboring repeaters at different distances. The attenuation on the transmission link is 0.05 dB / MHz.m. For a short distance I1 = 20 m the frequency range of 20-25 MHz and for a larger distance I2 = 40 m the frequency range of 15-20 MHz are used as an example. The illustration shows the transmission level, taking into account the required pre-distortion, whereby it is possible to comply with the following conditions:

• - The received useful signal has a leveled frequency response and
• - All parasitic signals have a reception level less than / equal to that Useful signal. It can be shown that for every distance ratio a setting can be found between the neighboring repeaters that these two conditions are sufficient. For exemplary dimensioning of an OFDMA-based transmission method were the following in the Table 4 shows initial data relationships, where with regard to their respective physical meaning towards the literature  OFDM is referenced.

Table 4

The stipulations listed in Table 4 ensure that the symbol duration is sufficiently greater than the maximum detour delay, which is achieved by the factor n and its a priori stipulation of n ≧ 10. In conjunction with the OFDM protection time T _G between the symbols, this ensures that the resistance to multipath reception is increased. The factor n also has the effect that when n is increased, the proportion of the OFDM protection time T _G in the OFDM block duration T _B becomes ever smaller, and this results in a higher effectiveness in terms of utilization of the available bandwidth. This effect is bought with an increase in the number of OFDM channels and a narrower OFDM channel grid. If a higher quality modulation such as QPSK is used, the net data rate per channel and thus the capacity per direction increases to twice the value with the same symbol rate. The sensitivity of QPSK to frequency response distortions in the transmission channel mentioned elsewhere is not relevant here, since QPSK is applied to very narrow-band carriers. Applications with 8PSK are also conceivable, which would result in a tripling of the capacity values. Possible physical realizations depending on n and based on the relationships according to Table 4 are given in Table 5, where Sps stands for symbols per second.

Table 5

The transmission capacity per direction available in a repeater chain is further determined by the TDD method used in each case, with the methods 2ph-4gr and 4ph-2gr a reduction factor of 118 and a reduction factor of 1/6 for the 3ph-2gr and 6ph-1gr processes is taken into account.

In FIG. 11, the processing of the transmitting signals is shown. The signals of each channel are fed separately to a serial-to-parallel converter 20 . In the case of m-valued modulation, this converts the serial bit stream in groups to m parallel bits. A respective downstream encoder 21 converts each bit group into the modulation status information appropriate for the bit group. These signals are then subjected to a band limitation 22 . In a subsequent modulator 23 , the carrier f _i is thus modulated and the modulated signal is subjected to an amplitude weighting 24 to compensate for the frequency response of the transmission channel.

The signals of all channels are then summed up in a summer 25 and fed to the actual transmission part. In the case of a two-value modulation, serial-to-parallel converter 20 and encoder 21 are omitted, since encoder 21 would only undo what was previously carried out by the serial-to-parallel converter. In this case, the channel signals are therefore fed directly to the modulator. In the case of a four-valued modulation, the serial-to-parallel converter 20 and the encoder 21 would correspond to the module for processing the I / Q signals. Furthermore, it should be noted that the amplitude weighting 24 can also be carried out before the modulator 23 . Preferably, parallel-to-parallel converters 20 , encoders 21 , modulators 23 , amplitude weighting 24 and summer 25 are implemented by means of a DSP module, the modulation being carried out as an inverse Fourier transformation and thus no carrier generation is necessary anymore, since the carrier information is used as a parameter in the Inverse Fourier transform algorithm is received.

In FIG. 12, the processing of the OFDM receiving signals is shown. The received signals are fed in parallel to the channel demodulators 26 . By adding the carriers f _i , demodulation takes place channel by channel. The signals at the output of each demodulator are subjected to low-pass filtering 27 and fed to a clocked threshold switch 28 . The clock T is obtained from the received signal by a conventional clock recovery. The decoders 29 convert the modulation status information into the bit groups adequate for this information. Subsequently, the conversion into a serial data stream takes place again in downstream parallel-serial converters 30 . The demodulators 26 , low-pass filters 27 , threshold switches 28 , clock recovery, and decoders 29 and parallel-serial converters 30 are preferably implemented in a DSP module, the demodulators 26 being implemented as a Fourier transform and therefore no carrier generation is required any more, since the carrier information as Parameters in the Fourier transform algorithm.

The OFDM access method can be used to improve the bit error rate  a frequency hopping can be combined. The basic principle is that the Data of the individual channels cyclically with a changing OFDM frequency be assigned so that in the event of failure or poor reception on a OFDM frequency the information of the whole channel is not lost, but only one bit of all channels, for example, which then if necessary, using suitable error-correcting procedures is reconstructable. In a further preferred embodiment, the existing channels are not fully used for user data transmission and for example at least one for each direction of transmission Reserve channel kept in stock. Then one of the repeaters determines that this a channel cannot receive, so its neighboring repeaters initialized to send on a reserve channel or if necessary to receive this, for example because the corresponding transmission part at Repeater has failed. This enables an unlimited Data transmission procedure in the event of external local disturbances or Defects of a repeater due to the very different external Circumstances can never be completely ruled out. Below is the practical one Implementation explained in more detail.

A word-oriented virtual ring memory is built up for k channels, which has exactly k memory locations with a word length w. For the parallel transmission of the data of the k channels, a data word is written into the ring memory for each channel, the channel number corresponding to the memory location number. This is followed by w read cycles, with the ring memory being read modulo k starting with each read cycle with changing start positions. A specific bit position is read with each read cycle. The bit position is increased by one digit after each read cycle. The start position can be shifted using various algorithms, in the simplest case the shift v has a constant value. The following read cycles result for this:
1st reading cycle: Starting at start position a1, reading the 1st bit of the k words
2nd reading cycle: Starting at start position a1 + v, reading the 2nd bit of the k words
w. Reading cycle: At start position a1 + (w - 1). v starting reading the wth bit of the k words.

The ring buffer is then rewritten and the start position a2 the 1st read cycle is a2 = a1 + w. v. The bits read are in the Order of reading assigned to the OFDM frequencies. Possibly frequencies that are marked as disturbed are skipped in the assignment are. This means that the data of a channel are constantly changing Transmit frequencies, the change takes place bit by bit and the Jump distance with the shift of the starting position in the reading cycles matches if hidden jamming frequencies are not in the count be included. The dimensioning of v is to be chosen so that the Difference between the sequentially assigned OFDM frequencies if possible is great. It is not only the 1st order difference that has to be considered, since a Frequency not only disturbed for one, but for several cycles is. The smaller of the values D1n = n applies to the difference of the nth order. v and D2n = k - n. v, where the values n. v modulo k are to be calculated. For k = 64 are in the table below for different shifts v Differences of different orders are shown.

Table 6

Is it important that the 4th and 5th order difference also have another Sufficient protection distance to a disturbed over 4 to 5 bit periods Frequency, the values v = 11 come for the selected example k = 64 or v = 17 into consideration.

The recipient of the data transfer must after which the data is processed in the transmitter, be known and  be synchronized to the start position and the bit position of the read cycles. For this purpose, the first transmitted bit position of each burst is not in TDD mode with channel information, but for the transmission of the for the Synchronization crucial start position of the next read cycle used. The bit position does not have to be transmitted if the preferred embodiment a word-oriented transmission takes place and always starting with bit position 1. A different one Realization that transmits bit-oriented and in this case the additional one Transfer of the bit position requires, but is in principle also possible. The The receiver is then set up reversely to the transmitter.

After the repeater chain has been set up in terms of hardware, it must first be initialized. The initialization is carried out as a top-down procedure, in that each initialized repeater initializes the repeater or successive repeaters in each case, the majority being intended to express that branch branches can also exist. In the following, however, an unbranched repeater chain is assumed for the sake of simplicity. In FIG. 13, the flowchart is shown for the 3-ph 2gr transmission method. The first repeater 11 of the chain is a gateway to a network that is not described in detail here. The repeater 11 is the only repeater in the chain that has access to a responsible network management system NMS and receives from it its initialization data, the TDD configuration and the information about its network environment, such as the IDs of the repeater or repeaters connected. First, the measurement of the transmission path to the nearest repeater 12 is initiated by the repeater 11 . The results of the measurement are stored in the repeater 11 as parameters down and in the repeater 12 as parameters up. Then the downward transfer agreements are made. An agreement on the upward transmission cannot yet be made, since the repeater 12 must still know the parameters of its downward transmission path for this purpose. The same procedures are then carried out by the repeater 12 in the direction of the repeater 13 . In addition to the agreement on the downward transmission to the repeater 13 , the repeater 12 then makes the agreements with the repeater 11 on the upward transmission. Both agreements are made taking into account the necessary transmission levels and the pre-distortion. Only the repeater 12 knows the parameters of the two transmission links (parameter up, parameter down) which it has to operate and can derive the assignment of the frequency ranges, transmission level and predistortion from this. In addition, the results of the channel test, which will be explained later, are taken into account in the agreement, by specifying channels to be hidden separately for each direction. The procedures are then carried out successively in the same way by all subsequent repeaters.

Fig. 14 shows the top-down initialization procedure for the 6ph-1gr transmission method. The procedure is largely identical to the top-down initialization procedure for the 3ph-2gr transmission method. The only difference is that the agreement down and the agreement up take place in one step. This results from the fact that each repeater sends only one repeater at a time and therefore no need to consider a second transmission path as with the 3ph-2gr transmission method.

The measurement taking place during the initialization is shown schematically in FIG. 15, whereby this applies both to the 3ph-2gr and the 6ph-1gr transmission method. An already initialized repeater with ID1 knows its network environment and addresses the subsequent repeater with ID2. The info also contains information about the TDD regime (number of phases, number of the current phase, transmission phase of the source, reception phase of the source). The transmission is repeated with varying parameters until the subsequent repeater answers with ID2. The repeater then acknowledges receipt with ID2 and the repeater with ID1, whereupon the level / attenuation measurement is carried out. The level / attenuation measurement is a multi-stage process and is carried out at at least three points in the transmission spectrum (middle, upper and lower edge zone). The result is the transmission level required to generate the minimum reception level Rxmin at the remote site. Due to the reciprocity of the transmission channel, the result applies to both directions. Then the channel check is carried out down. With transmission from the repeater with ID1 and reception by repeater with ID2, whereby all channels may be used in several sub-packets, the receiving repeater checks whether individual channels are constantly disturbed at the receiving location and are therefore to be hidden. For the channel test up, the same test is carried out in the opposite direction. With regard to the detection of the interference of channels, a separate test is required for the transmission directions, since the reciprocity condition does not apply to the occurrence of interference. The status is then sent to the NMS. With the provision of the physical layer, the repeater can use ID2 to log on to the NMS. The NMS confirms the registration and transfers further configuration data and information about the network environment.

LIST OF REFERENCE NUMBERS

1

channel group

2

channel group

3

channel group

4

channel group

10

repeater chain

11

repeater

12

repeater

13

repeater

14

repeater

15

repeater

16

repeater

17

repeater

18

management

20

Serial-to-parallel converter

21

encoder

22

bandlimiting

23

modulator

24

Amplitude weighting

25

summing

26

channel demodulators

27

low-pass filtering

28

Threshold switch

29

decoder

30

Parallel to serial converter

Claims (14)

1. Method for data transmission on powerlines or other physically inseparable lines by means of a repeater chain, the data transmission being channel-oriented and the data being assigned to the channels by means of an access method and being transmitted between the repeaters in duplex mode by means of a transmission method,
characterized in that
the transmission method is organized as a three-phase TDD with division of the total number of available channels in two groups of channels (1, 2) and the turned-on to the conduit (18) repeaters (11-17) according to their order on the line (18) modulo 3 in three Groups are divided, comprising the following process steps:

• a) sending the first group ( 11 , 14 , 17 ) in the first TDD phase to the repeaters ( 12 ; 13 , 15 ; 16 ) adjacent in the up and down direction,
• b) sending the second group ( 13 , 16 ) in the second TDD phase to the repeaters ( 12 , 14 ; 15 , 17 ) adjacent in the up and down direction,
• c) send the third group ( 12 , 15 ) in the third TDD phase to the repeaters ( 11 , 13 ; 14 , 16 ) which are adjacent in the up and down direction, the channel groups ( 1 , 2 ) per TDD Phase for all repeaters ( 11 - 17 ) are permanently assigned to a transmission direction.

2. Method for data transmission on power lines or other physically inseparable lines by means of a repeater chain, the data transmission being channel-oriented and the data being assigned to the channels by means of an access method and being transmitted between the repeaters in duplex mode by means of a transmission method,
characterized in that
the transmission method is as sechsphasiges TDD with a channel group (1) which includes the total number of available channels organized and the turned-on to the conduit (18) repeaters (11-17) according to their order on the line (18) modulo 3 in three Groups are divided, comprising the following process steps:

• a) sending the first group ( 11 , 14 , 17 ) in the first TDD phase to the neighboring repeaters ( 12 , 15 ) in the down direction,
• b) sending the first group ( 11 , 14 , 17 ) in the second TDD phase to the neighboring repeaters ( 13 , 16 ) in the up direction,
• c) sending the second group ( 12 , 15 ) in the third TDD phase to the neighboring repeaters ( 13 , 16 ) in the down direction,
• d) sending the third group ( 13 , 16 ) in the fourth TDD phase to the neighboring repeaters ( 12 , 15 ) in the up direction,
• e) send the third group ( 13 , 16 ) in the fifth TDD phase to the neighboring repeaters ( 14 , 17 ) in the down direction and
• f) send the second group ( 12 , 15 ) in the sixth TDD phase to the neighboring repeaters ( 11 , 14 ) in the up direction.

3. The method according to claim 1, characterized in that the Transmission levels of the repeaters are dimensioned such that at the Repeater in the direction of transmission with the greatest attenuation itself sets the minimum reception level Rxmin. 4. The method according to claim 1 or 3, characterized in that for Frequency response compensation on the transmitter side to a receiver Repeater preemphasis is made and the residual distortion on the other receiver repeater using an equalization amplifier is compensated. 5. The method according to any one of claims 1, 3 or 4, characterized in that the proportions of the two channel groups ( 1 , 2 ) in the total number of channels k can be set variably. 6. The method according to claim 2, characterized in that the Transmission level of a repeater in each case on the receiver Repeater of the respective TDD phase is set that on Receiver repeater sets the minimum reception level Rxmin. 7. The method according to claim 2 or 6, characterized in that for Frequency response compensation only on the transmission side to the adapted preemphasis for each receiver repeater becomes. 8. Procedure for data transmission on Powerlines or others lines that cannot be physically separated, using a repeater chain, whereby the data transmission takes place channel-oriented and the data by means of an access method assigned to the channels and by means of a Transmission procedure between the repeaters in duplex mode be transmitted, characterized in that an orthogonal frequency division multiple as the access method Access procedure OFDMA is used. 9. Procedure for data transmission to powerlines according to one of the Claims 1 to 7, characterized in that the access method according to claim 8 applies. 10. A method for data transmission according to claim 8 or 9, characterized characterized in that the frequency response compensation of Transmission channel takes place on the transmission side in the signal level by the Individual signals of the OFDM before an inverse Fourier transformation with an amplitude weighting correlating to the signal frequency be, with a three-phase method for the transfer to the most distant repeater the lowest frequency range the transmission band and to the closest repeater the highest frequency range of the transmission band can be used  and that in the sub-bands thus formed, the predistortions are proportional to the respective distance and their absolute amount is chosen so that each neighboring repeater its Usable area undistorted. 11. A method for data transmission according to claim 8, 9 or 10, characterized characterized in that the OFDM access method with a frequency Hopping and one related to the data channel Error correction method combined is used. 12. A method for data transmission according to claim 11, characterized characterized in that constantly disturbed in the transmission channel Frequencies in the OFDM scheme are not assigned channel information and are omitted from frequency hopping. 13. Circuit arrangement for performing the method according to one of claims 8 to 12, characterized in that the circuit arrangement comprises a plurality of repeaters ( 11-17 ), each repeater ( 11-17 ) having a transmitting part and a receiving part for transmitting and receiving comprises k channels, the channel processing in the transmitting part for each channel comprising a band limiter ( 22 ), a modulator ( 23 ) and an amplitude weighter ( 24 ) and on the output side a summer ( 25 ) for the k individual signals, and the baseband signal processing in the receiving part for each channel comprises a demodulator ( 26 ), a low pass ( 27 ) and a clocked threshold switch ( 28 ). 14. Circuit arrangement according to claim 13, characterized in that the transmitting and receiving part of a repeater ( 11-17 ) are integrated in at least one DSP module.