HK1023665B - Method for transmission of data packets - Google Patents

Description

The present invention relates to a method for transmitting data packets in a network of user stations, whereby the data packets are transmitted over a set of channels in the frequency hopping method, the channels being selected according to a frequency hopping pattern in time sequential data transmission slots, and assigned to a user station as needed slots in which it is eligible to transmit.

A similar procedure is known from DE 44 07 544 A1 and further state of the art is referred to in WO-A-9620538 and EP-A-189695

The known method is to transmit data packets in an additional network, which transmits over frequency channels, at least partly already used in an existing base network for digital data and/or voice transmission. In this method, a first step is to determine a frequency channel not currently occupied by the base network, and then a second step is to transmit a data packet over the determined frequency channel. These steps are repeated cyclically in a kind of frequency hopping technique until all data packets of a broadcast have been transmitted.

The known method allows the use of the frequency hopping technique to make better use of existing channels without disrupting the respective base network.

As is common in the frequency hopping technique, the individual data packets are transmitted in time-sliced slots over different frequency channels according to a frequency hopping pattern, with the load distributed evenly over the available channels.

The underlying frequency jump table is here generated pseudo-randomly, with the destination address being included in the channel calculation.

The selection of a frequency channel not currently used by the base network is carried out in this method by first selecting one of the several frequency channels and then listening to that channel to check whether the base network is transmitting on that frequency channel.

However, if it is found that the selected channel is currently being used by a primary user, the time slot expires unused, i.e. no data packet is transmitted over the selected channel in order not to disturb the primary user.

It follows from the above that the known method assigns the broadcasting rights to the user stations so that the frequency at which a user station can broadcast within a given time period depends on the number of user stations.

The assignment of the individual channels of the frequency jump pattern to the respective user stations therefore depends on the number of user stations.

In addition to the above mentioned data throughput, which is not optimal in certain network conditions, the known method thus has the additional disadvantage that a real-time data transmission is not always possible.

In this context, the present invention is intended to develop the method mentioned at the outset in such a way as to maximize data throughput in the network of user stations and to enable real-time data transmission.

According to the invention, this task is solved by the method described in claim 1.

The problem underlying the invention is thus fully solved.

A central station for a packet-transmitting network of user stations, where the data packets are transmitted over a set of channels in a frequency hopping process, the channels being selected in sequence for data transmission according to a frequency hopping pattern, with the central station having a first transmit/receive part which processes a first frequency hopping pattern to transmit data packets to and from the user stations load-dependently, is characterised by the central station having a second transmit/receive part which processes the data packets in approximately the same time as the first orthogonal second frequency hopping pattern, in order to transmit the requested data in a pre-redactable time.

Accordingly, a user station for a packet-transmitting network of such user stations, where the data packets are transmitted over a set of channels in the frequency hopping procedure and the channels are selected for data transmission in sequence according to a frequency hopping pattern, is characterized by the invention by the fact that the user stations are set up to store two frequency hopping patterns and to call them optionally to transmit a data packet, wherein the first frequency hopping pattern transmits data packets load-dependently as specified by a central station and wherein the first orthogonal to the second frequency hopping pattern transmits data packets in a pre-sampling time.

The central station and the user station therefore process the same frequency jump pattern, with the central station granting the broadcasting authorisation as needed, i.e. load-dependent for asynchronous data transmission and with a fixed time grid for synchronous data transmission.

The time slots are allocated to the user stations according to their needs, i.e. depending on the load, which allows for flexible response to changing load conditions on the network. A user station with a high transmission demand receives more time slots than other user stations that have no or only a small amount of data to transmit per unit of time. This is surprisingly possible because, as in the state of the art, not only can channels be allocated from the frequency jump pattern, but also many more time slots can be allocated to the user stations, resulting in the frequency frequency being used for transmission.

It is also preferable to allocate at least one initial fixed time frame of periodically arranged time slots to a user station for synchronous packet transmission and to allocate at least some of the remaining time slots to the user stations dynamically depending on their data volumes.

The advantage here is that the same frequency jump pattern now allows both synchronous and asynchronous data transmission. A user station that wants to transmit language, for example, is assigned a fixed grid of time slots, whereby the assigned channels are again derived from the respective current frequency jump pattern. The still free time slots are still assigned dynamically, i.e. load-dependently.

On the other hand, it is preferable to transmit data packets as required in a first frequency jump pattern and at least in a second frequency jump pattern orthogonal to the first frequency jump pattern in a predefined time frame synchronously in the network.

The inventors of the present application have recognised that, between the same user stations and over the same set of channels, two surprisingly different types of data transmission can be carried out, thus maximising data throughput and enabling real-time data transmission.

Err1:Expecting ',' delimiter: line 1 column 1022 (char 1021)

The present inventions have, however, recognised that the inventive use of the second frequency jump pattern nevertheless makes synchronous traffic possible in parallel with the load-dependent allocation of the broadcasting authorisation, whereby the data packets are transmitted in a fixed time frame, i.e. the distance between the allocated time slots is constant; this distance is predefined and can therefore be adjusted to the amount of data to be transmitted or to their frequency, i.e. two networks operate, so to speak, between the same user stations and over the same channels, while measures to be described ensure that the load-dependent network optimised for synchronous data traffic does not interfere with the other network.

The orthogonal frequency jump patterns also prevent mutual interference by interference, and orthogonality is understood to mean that within each time slot the two frequency jump patterns always have different channels, so that even if both networks transmit together within a time slot, no mutual interference can occur.

Err1:Expecting ',' delimiter: line 1 column 222 (char 221)

The new method can be implemented, for example, by each user station having a transmitter/receiver part for the first frequency jump pattern and a second transmitter/receiver part for the second frequency jump pattern. However, a user station can only handle one data packet, e.g. in the event of a collision, the data packet transmitted over the channel from the second frequency jump pattern must be given priority in order to avoid loss of synchronization. However, it may happen that data packets to be transmitted over a channel of the first frequency jump pattern or by the user who is sending or receiving a repeated synchronized signal are lost.

A user station may also process both asynchronous and synchronous traffic in time-neutral mode, for example, where the time slots between the fixed time slots for synchronous traffic are used by other user stations depending on the load.

While the channels in the first frequency jump pattern are dynamically assigned to the user stations to respond to load changes, the channel assignment in the second frequency jump pattern can be static or variable to also handle real-time data transmissions with different data volumes.

Accordingly, the new central station is preferable to have a second transmit/receive part that processes a second frequency jump pattern approximately simultaneously to the first orthogonal to transmit data packets in a predefined time frame.

Therefore, the new user station is preferably configured to store two frequency jump patterns and to optionally call to transmit a data packet, where the first frequency jump pattern transmits data packets load-dependently and the first orthogonal second frequency jump pattern transmits data packets in a predefined time frame.

The advantage here is that the new user station requires only a few further changes, it only needs to store two frequency jump patterns and switch from one to the other. The new central station, on the other hand, is simply composed of two preferably identical transmitting/receiving parts, one of which constantly follows the first and the other constantly follows the second frequency jump pattern. Since both frequency jump patterns are orthogonal to each other, a user station can transmit synchronous and the other asynchronous data, the central station then gives the data transmitted to the synchronous data the advantage in the forward processing, the data transmitted asynchronously are lost, the data transmitted asynchronously are transmitted in a subsequent split-second transmission.

In general, it is preferred here that the set of channels is in a frequency band used by primary users, whereby a channel selected by the network of user stations is checked before any possible data transmission, for example, to see if a primary user occupies that channel, and depending on this check either a data packet is transmitted over the selected channel or the next channel is selected for checking and possible data transmission in the respective frequency jump pattern.

Err1:Expecting ',' delimiter: line 1 column 775 (char 774)

Err1:Expecting ',' delimiter: line 1 column 253 (char 252)

It is of course possible to adapt two frequency jump patterns to the channel occupancy situation by primary users adaptively using the procedure described there, while also ensuring that the changed frequency jump patterns are still orthogonal to each other.

Generally, it is preferred to dynamically allocate the time slots of the first frequency jump pattern to the user stations according to their data input.

The advantage here is that a simple, load-dependent allocation of time slots and the channels assigned results: a user station with a higher data volume receives more time slots and thus also the corresponding channels from the first frequency jump pattern assigned than a user station with little or no data volume.

Furthermore, it is preferable for the network of user stations to have a common system time and to process both frequency jump patterns cyclically and simultaneously so that in a given time slot one channel from the first and second frequency jump patterns is available for data transmission, preferably with fixed periodic time slots allocated to each user station in the second frequency jump pattern or with a set of fixed periodic time slots allocated to a user station in the second frequency jump pattern if necessary.

The advantage here is that the two frequency jump patterns are processed in synchrony with each other so that in each time slot a user station with a right to transmit can either transmit an asynchronous data packet via the current channel from the first frequency jump pattern or a synchronous data packet in the current channel from the second frequency jump pattern.

In other words, the user stations usually process the channels from the first frequency jump pattern and only if the current time slot matches the fixed set of channels assigned to this user station in the second frequency jump pattern can a synchronous data packet be transmitted.

In the first alternative, with fixed allocation of time slots from the second frequency jump pattern, the respective time slot expires unused if the user station does not have a synchronous data packet to transmit at the moment; in the second alternative, i.e. the variable allocation, this allocation does not take place at all if the respective user station does not want to transmit synchronous data packets, so that this user station only processes channels from the first frequency jump pattern.

Err1:Expecting ',' delimiter: line 1 column 145 (char 144)

Furthermore, it is preferable if a user station is a central station with two transmit/receive parts, one of which is set to the channel assigned to the first frequency jump pattern and the other to the channel assigned to the second frequency jump pattern in a given time slot.

This is particularly advantageous in a hierarchical network, since the central station can thus send and receive both asynchronous and synchronous data packets at any time of possible data transmission. In a hierarchical network, data traffic between two user stations is usually always transmitted via the central station, so that a user station with the right to transmit transmit transmits its data packets to the central station.

In this context, each user station preferably has a transmit/receive component that is set to either the first or second frequency jump pattern channel.

The advantage here is that the user stations can be set up much more simply than the central station, which only needs to be able to manage both frequency jump patterns.

Furthermore, it is preferable for the user station to follow the first frequency jump pattern and only switch to the second frequency jump pattern if it is authorised to do so.

The advantage here is that the user station switches its transmitting/receiving part to a channel from the second frequency jump pattern only if it is also authorized to transmit over this channel. As mentioned above, this broadcasting authorization can be fixed or optionally assigned.

It is also preferable for the user station to switch to the second frequency spectrum only if it is to transmit synchronous data.

The advantage here is that the number of unused time slots is further reduced only if the user station itself wants to send a synchronous data packet or expects a synchronous data packet, it switches to a channel from the second frequency jump pattern.

In this case, it is preferable for the central station to announce a synchronous data transmission on a channel from the first frequency jump pattern to the user station.

The advantage here is that the user station stays in the first frequency jump pattern as long as it does not want to start synchronous data transmission itself.

In summary, the new method is based on the transfer of data packets between the user stations and the central station, depending on the load, whereby the central station allocates channels and thus time slots from the first frequency jump pattern to each user station depending on the data volume.

If a user station is to transmit data simultaneously, e.g. voice, it must repeatedly use a channel from the first or second frequency jump pattern in a given or to be given equidistant pattern of time slots.

The central station is preferably designed to follow both the first and second frequency jump patterns, i.e. to be able to transmit and receive both synchronous and asynchronous data packets in each time slot. This is necessary because it is not known in advance whether a synchronous data packet is actually transmitted over a channel from the second frequency jump pattern.

This greatly increases the total data throughput through the user station network, because the channels from the second frequency jump pattern are actually used only if a transmission of synchronous data packets takes place.

This new method thus allows for the optimum use of available channel capacity even in networks where the user stations must be very safe to avoid interference with the primary users.

Further advantages are derived from the description and the accompanying drawing.

An example of the invention is shown in the figure and is described in more detail in the following description. Fig. 1a schematic example of a network of user stations;Fig. 2a schematic example of a frequency jump pattern used by the network from Fig. 1;Fig. 3a schematic example of a frequency spectrum of a base network to which the network from Fig. 1 is superimposed;Fig. 4the division of a time slot of the network from Fig. 1 into the different operations;Fig. 5a more detailed representation of some user stations of the network from Fig. 1; andFig. 6a schematic example of a frequency jump pattern used by the network from Fig. 5.

Figure 1 shows a schematic network 10 comprising a central station 11 and several user stations 12, 13, 14 and 15.

The network 10 is hierarchically structured, with user stations 12, 13, 14, 15 only communicating with each other via central station 11 and contacting other external stations via central station 11.

The network 10 uses for data transmission channels of a frequency spectrum described in detail in connection with Figure 3 in the frequency hopping procedure, for which purpose a frequency hopping pattern 21 shown in Figure 2 is stored in both the central station 11 and the user stations 12, 13, 14, 15 and cycled through as indicated by the arrow 22.

In the example shown, in the frequency jump pattern 21, a set K (k=1...80) of 80 channels is arranged statistically so that there is the greatest possible jump distance between two consecutive channels K (k). This ensures that the phase loops of two consecutive channels are spaced so that potentially disturbing interference on one channel is avoided with high confidence by multiple reflections on the next channel.

The network 10 operates with a system time which defines successive time slots, in which each time slot transmits data via the respective valid channel between central station 11 and one of the user stations 12, 13, 14, 15.

The frequency jump pattern 21 is cyclically passed through so that the time slot t = 79 is reconnected to the time slot t = 0. The user station 12, 13, 14 or 15 for which a transmission from central station 11 is intended is indicated by an address that the central station 11 pre-dates the data packet to be transmitted in the respective time slot. In other words, while the system time through the time slot determines the channel to be used, central station 11 indicates the addressee when transmitting data. Furthermore, during this data transmission it is indicated which user station 12, 13, 14, 15 in the next time slot receives the slit authorization. The time slot of the next release is indicated by the source channel 21 from the pattern Figure 2.

In addition to their respective data packets, the user stations 12, 13, 14, 15 of central station 11 also transmit information on the number of data packets still to be transmitted, so that central station 11 can allocate the broadcasting authorization according to the load.

The network 10 thus allows for asynchronous, load-dependent transmission of data packets, optimising data throughput by the hierarchical allocation of broadcasting authorisations.

However, the network 10 described here does not operate in an exclusive frequency range, but is rather an additional network overlapping an existing network, as will now be shown in Figure 3.

Err1:Expecting ',' delimiter: line 1 column 202 (char 201)

In addition to the very narrowband primary users 24, 25 and user station 13, there is another primary user 26, whose mid-frequency is f25. This primary user 26 is, however, very broadband and has a spectral bell curve 27 extending from f20 to f30. However, in the peripheral areas of this bell curve, above f20 and below f30, the transmitting energy S of the primary user 26 is so low that it is below a detectable threshold W that a user station, for example on channels f21 and f29, would not be able to receive a signal and would therefore consider this channel free.

As mentioned above, the user stations use channels f1 to f80 for data transmission using the frequency hopping technique, whereby, to avoid interference to the primary users, at the beginning of any data transmission, the user checks whether the frequency channel just selected is occupied by primary users.

The data is transmitted by a user station, which is then able to determine the channel through which the transmission is to be made, and to ensure that the transmission does not disturb a primary user, the user station checks whether the selected channel is being used by a primary user before sending a data packet in the respective time slot.

Fig. 4 shows a schematically indicated time slot 29 of T = 8 ms on the time axis. At the beginning of this time slot 29, the user station's transmitter is first set to the selected channel, which is done during the time T1 = 50 ns.

The channel is then queried for occupancy during the time T2 = 500 μs. If the channel is free, a data packet is transmitted during the time T3 which connects to T2, although not directly, for which 4 ms are available.

After the data packet has been transmitted, T4 waits for a certain time to confirm the receipt of the data packet by the recipient. If this confirmation is received, the next data packet is transmitted in the next time slot, the confirmation is not received, the same data packet is transmitted again in the next time slot.

If the check in T2 shows that the selected channel is already occupied by primary users, the remainder of the time slot will be unused and the operations described above will start again at the beginning of the next time slot.

The network 10 described in Figure 1 is able to adapt to different load states by means of an asynchronous transmission of data packets with load-dependent allocation of the broadcasting authorization by the central station 11.

However, this can be remedied by allocating to the central station 11 of a user station 12, 13, 14, 15 a fixed time frame of periodically arranged time slots in which it can transmit language.

The user station transmitting the language may not always need the specified time slots, however, and may lose them for data transmission. In the case of voice breaks, the user station does not transmit, so that other asynchronous data could be transmitted.

It is of course possible to ensure that such an unused time slot is available for asynchronous data transmission by exchanging information between central station 11 and the respective user station 12, 13, 14, 15. However, the coordination effort required can be very high, especially if the total data throughput through the network is very high, so that central station 11 has to spend a large part of the available processing time on data traffic.

For this reason, in an alternative design of the invention, between the user stations 11, 12, 13, 14, 15 of the network 10 of Fig. 5 another synchronous data network is laid, following a frequency jump pattern orthogonal to the frequency jump pattern for asynchronous data transmission, by both the central station 11 and the user station 12, 13, 14, 15 processing both frequency jump patterns as described in Fig. 5.

Central station 11 has a control unit 31 and two transmitting/receiving units 32 and 33. Each transmitting/receiving unit 32, 33 has a transmitting/receiving antenna 34 and 35 respectively. Data lines 36 connect the transmitting/receiving units 32 and 33 to the control unit 31 and a control line 37. Central station 11 may also include two central stations of the type intended to operate a network that handles only one frequency jump pattern.

The user stations, of which stations 12, 14 and 15 are shown in Fig. 5, also have transmitting and receiving antennas 38, 39, 40 through which they communicate with central station 11.

Err1:Expecting ',' delimiter: line 1 column 533 (char 532)

The transmitting/receiving part 33 shall process a frequency jump pattern 44 accordingly by means of a pointer 45.

While the frequency jump pattern 41 is used for asynchronous data transmission as described in connection with the frequency jump pattern 21 in Figure 2, the frequency jump pattern 44 is used for synchronous data transmission, e.g. for voice transmission, in a manner to be described.

The user station 12 (S1) uses its transmitting/receiving part 46 to process a frequency jump pattern 47 with the help of a pointer 48. This frequency jump pattern 47 is associated with the asynchronous frequency jump pattern 41 of the central station.

Similarly, the user station 14 also includes an asynchronous frequency jump pattern 51 processed by a pointer 52 and a synchronous frequency jump pattern 53.

Both user stations 12, 14 are equipped with an off-line data connection 55 and an on-line voice connection 56.

The frequency jump patterns 41, 47, 51 are identical with the hands 42, 48, 52 moving synchronously from time slot to time slot, which defines for each time slot exactly which channel the central station 11 and the user stations 12, 14 can communicate with each other.

The frequency jump patterns 44, 49 and 53 are also identical, but the allocation of the broadcasting authorization is static. This means that at a very specific time slot only a very specific user station can transmit or receive over the channel resulting from the frequency jump pattern 44, 49, 53. However, as long as the user station 12 or 14 does not have a transmit/receive need, it operates the asynchronous frequency jump pattern 41, 47, 51. Only if 56 data are on its line or the central station 11 has received a signal from a channel from the first frequency sharing that language data is to be received, the respective user station 12 or 14 switches to the second frequency jump pattern only for the allocated time slots.

Figure 6 shows how this is done.

Fig. 6 shows some of the time slots t in the upper row. The following rows show the frequency jump pattern for voice transmission from user station 12 (S1V), the frequency jump pattern for asynchronous data transmission from user station 12 (S1D), the frequency jump pattern for voice transmission from user station 14 (S3V) and the frequency jump pattern for asynchronous data transmission from user station 14 (S3D). For the sake of clarity, the channel assignments found in the respective frequency jump patterns have been omitted, but only by indicating which frequency is transmitted by which user at the respective frequency jump pattern.

It is clear that the frequency jump patterns S1V and S3V assign the broadcasting privileges at equidistant intervals. The user station 12 can transmit language in time slots 0, 10, 20 ... for example, while the user station 14 (S3V) can transmit language in time slots 5, 15, 25 ... The equidistant intervals can also be assigned differently as required, either when the network is configured or dynamically changed in operation.

The frequency slots in the asynchronous frequency jump patterns S1D and S3D are load-dependent and assumed to be arbitrary.

In time slot t = 0 the user station 12 transmits language, while in time slot t = 5 the user station 14 transmits language. For this time slot the user station 12 has a broadcasting authorization in the asynchronous frequency jump pattern S1D. This means that at time slot t = 5 the user station 12 transmits data and the user station 14 transmits language to central station 11. Central station 11 receives both data packets with its two transmit/receive parts 32, 33 but in control unit 31 only the synchronous data package, as this has priority. At the end of this time slot t = 5 the user's flash confirms the reception of the speech packet, while the user's flash 12 remains in front of the central station without confirmation, so that it is placed on the user's data packet.

In the time slot t = 20, the user station 12 has both a voice and a data transmission authorization. If no voice traffic needs to be processed, the user station 12 transmits asynchronous data, but otherwise synchronous voice data.

The time slots in Figure 6 which have not been crossed are assigned to other user stations which are not shown in this figure for the sake of clarity.

Of course, the user stations 12, 14 process the frequency jump patterns S1V and S3V only if synchronous data transmission is actually required, otherwise switching to the frequency jump pattern for synchronous data transmission is omitted.

Claims (14)

1. Process for transmitting data packets in a network (10) of user stations (12, 13, 14, 15), wherein the data packets are transmitted via a set of channels in the frequency hopping process and the channels are selected in this process for data transmission according to a frequency hopping pattern (21; S1V, S1D, S3V, S3D) in time slots (t) following one another in time sequence, and time slots (t) are allocated as necessary to a user station (12, 13, 14, 15), in which it is authorised to transmit, characterised in that in a first frequency hopping pattern (S1D, S3D) data packets are transmitted as necessary in a presettable time grid synchronously in the network (10) and at least in a second frequency hopping pattern (S1V, S3V) orthogonal to the first frequency hopping pattern (S1D, S3D), and in that each user station (12, 13, 14, 15) has a transmitter/receiver part which can be adjusted as desired to a channel from the first or second frequency hopping pattern (21; S1V, S1D, S3V, S3D).
2. Process according to claim 1, characterised in that the set of channels is in a frequency band that is used by primary users (24, 25, 26), wherein one channel selected, in each case, from the network (10) of user stations (12, 13, 14, 15), prior to a possible data transmission, is checked for whether a primary user (24, 25, 26) is currently occupying this channel, and depending on this check, either a data packet is transmitted via the selected channel or else the next channel in the respective frequency hopping pattern (21; S1V, S1D, S3V, S3D) is selected for checking and possible data transfer.
3. Process according to claim 1 or 2, characterised in that the time slot (t) of the first frequency hopping pattern (S1D, S3D) is allocated to the user stations (12, 13, 14, 15), dynamically depending on their data supply.
4. Process according to any one of claims 1 to 3, characterised in that the network (10) of user stations (12, 13, 14, 15) has a common system time (t) and both frequency hopping patterns (21, S1V, S1D, S3V, S3D) are serviced cyclically and synchronously in such a way that one channel from the first and second frequency hopping pattern (21; S1V, S1D, S3V, S3D) is always available for data transmission in a specific time slot (t).
5. Process according to any one of claims 1 to 4, characterised in that fixed, periodically arranged time slots (t) in the second frequency hopping pattern (S1V, S3V) are allocated to each user station (12, 13, 14, 15).
6. Process according to any one of claims 1 to 4, characterised in that a set of fixed, periodically arranged time slots (t) in the second frequency hopping pattern (S1V, S3V) are allocated as necessary to a user station (12, 13, 14, 15).
7. Process according to any one of claims 1 to 6, characterised in that a user station is a central station (11) which has two transmitter/receiver parts (32, 33), of which, in a specific time slot (t), the one is adjusted to the allocated channel of the first frequency hopping pattern (S1D, S3D) and the other to the allocated channel of the second frequency hopping pattern (S1V, S3V).
8. Process according to any one of claims 1 to 7, characterised in that the user station (12, 13, 14, 15) follows the first frequency hopping pattern (S1D, S3D) and only switches to the second frequency hopping pattern (S1V, S3V) on its own transmission authorisation.
9. Process according to claim 8, characterised in that the user station (12, 13, 14, 15) only switches to the second frequency hopping pattern (S1V, S3V) if synchronous data transmission is waiting for it.
10. Process according to claim 9, characterised in that the central station (11) of the user station (12, 13, 14, 15) signals a synchronous data transmission via a channel from the first frequency hopping pattern (S1D, S3D).
11. Central station for a network (10) of user stations (12, 13, 14, 15) transmitting data packets, wherein the data packets transmit via a set of channels in the frequency hopping process and the channels are selected for data transmission one after the other in time sequence according to a frequency hopping pattern (21; 41, 44), wherein the central station (11) has a first transmitter/receiver part (32) which services a first frequency hopping pattern (41) to transmit data packets to and from the user stations (12, 13, 14, 15) depending on load, characterised in that the central station (11) has a second transmitter/receiver part (33) which approximately synchronously services a second frequency hopping pattern (44) orthogonal to the first to transmit data packets in a presettable time grid.
12. Central station according to claim 11, characterised in that it is equipped to transmit data packets via a set of channels which is in a frequency band which is used by primary users (24, 25, 26), wherein a channel selected, in each case, from the network (10) of user stations (12, 13, 14, 15) is checked before a possible data transmission as to whether a primary user (24, 25, 26) is currently occupying this channel, and, depending on this check, either a data packet is transmitted via the selected channel or else the next channel in the respective frequency hopping pattern (21; 41, 44) is chosen for checking and possible data transfer.
13. User station for a network (10) of user stations (12, 13, 14, 15) transmitting data packets, wherein the data packets are transmitted via a set of channels in the frequency hopping process and the channels are selected in this process for data transmission one after the other in time sequence according to a frequency hopping pattern, characterised in that the user station (12, 13, 14, 15) is equipped to store and to call in as desired two frequency hopping patterns (47, 51; 49, 53) to transmit a data packet, wherein in the first frequency hopping pattern (47, 51), data packets are transmitted, depending on load, according to the requirement of a central station (11) and in the second frequency hopping pattern (49, 53) orthogonal to the first, data packets are transmitted in a presettable time grid.
14. User station according to claim 13, characterised in that it is equipped to transmit data packets via a set of channels which is in a frequency band which is used by primary users (24, 25, 26), wherein a channel selected, in each case from the network (10) of user stations (12, 13, 14, 15) is checked before a possible data transmission as to whether a primary user (24, 25, 26) is currently occupying this channel, and, depending on this check, either a data packet is transmitted via the selected channel or else the next channel in the respective frequency hopping pattern (21; 47, 51; 49, 53) is selected for checking and possible data transfer.