NZ540378A - Method and device for the optimisation of cellular wireless message networks - Google Patents

Abstract

Device for optimizing communication networks, in particular of cellular, wireless communication networks with mobile subscribers, comprises: a radio access network for wireless connection of the mobile subscribers; a communication network with fixedly networked transmission channels and a database with appropriate bi-directional interfaces to the communication network. The database contains, in particular, current position data, configurations of the network elements, measured data of various quantities from the communication network, counter data from the network elements, call statistics and planning data. Simulation means are provided, which generate a virtual model of the communication in a computer device, the configuration of which model is loaded from the database with the current state of the network configuration, and the current state of which model, in turn, can be transmitted to the database. An iteratively optimized network configuration is provided by accumulating appropriately weighted service specific subscriber traffic values of the rectangular area elements and by locally limited adjustment of the original network configuration.

Description

54037 1 Method of and Device for Optimising Cellular Wireless Communication Networks Specification Subject Matter of the Invention The invention relates to a method of optimising the throughput of messages in cellular wireless communication networks in accordance with the preamble of claim 1 and a device for executing the method in accordance with the preamble of the independent claim 14.

Prior Art Introducing cellular, wireless communication networks presented, from the beginning, the problem of finding, during construction and later extension and operation of the 20 networks, configurations which permit optimal utilisation of the available resources of hardware and frequency band width. The goal is, above all, maximisation of the throughput of messages with acceptable network quality and adequate network availability and network coverage, respectively. There, the concrete target quantity may be different. Thus e.g. the jam rate may be minimised, as, for example, explained in US 25 patent 5,826,218. WO 98/53621, again, is based on measured pilot reception power values. In order to achieve the goals of network optimisation, different parameters of the network configuration are adjusted, usually locally limited. This may be, for example, the adjustment of the aerial diagram or the pilot power, respectively, as in US patent 5,276,907, or the receiver sensitivity, as in WO 01/37446 Al.

In cellular phone networks, which, as a rule, only transmit information from stationary transmitting stations to subscribers, who may be mobile, comparable to conventional broadcasting, methods are used which optimise the network configuration mainly with K:\53\5387\5387.01\W05387.01-NV-050429-EnglUebers.doc , 1 respect to the network coverage, based on the radiowave propagation characteristics in the neighbourhood. In doing so, substantially radiowave propagation simulation and/or measurements of the average reception power as well as of the service quality at certain locations in the network are employed. The basic approach for such methods has been 5 described in Norbert Geng, Werner Wiesbeck, "Planungsmethoden fur die Mobilkommunikation", Springer-Verlag Berlin, 1998, ISBN 3-540-64778-3.

In communication networks such as the GSM-standard which use many frequency bands, the frequency bands are allotted to the individual cells of the network. In doing this, 10 appropriate distribution of the frequency bands on the cells of the network play a decisive roll in frequency minimising and thereby, throughput maximising. Such methods of frequency planning of the network are often summarised under the term "automatic frequency planning" (AFP).

With increasing importance of the various mobile data services and the development towards mobile communication networks of third generation, for example in accordance with the UMTS-standard, which is being developed within the framework of the international "3rd Generation Partnership Project" (3GPP), a network planning that meets the demands with regard to the service-specific subscriber traffic becomes more and 20 more important. The various available services have quite different characteristics with regard to their burst behaviour, the line- or packet-switched transmission mode, the data rate, the average transmission power and the symmetry/asymmetry between downlink and uplink. In addition, the partially quite different physical transmission characteristics of the various services have to be taken into account, as, as a rule, different transport 25 channels are used therefor. These specific demands on third generation mobile communication networks are not met by the prior art and conventional methods and devices for the purpose of communication network optimisation.

Problem of the Invention From the view of the network operator, it appears important that all provided services are available with substantially equivalent subjective quality and equal covering. Therefore, it appears necessary to employ a network optimisation program, which provides the K:\53\5387\5387.01\W05387.01-NV-050429-EnglUebers.doc 3 available services in the communication network all over the supply area with adequate quality.

It is the problem of the invention, to optimise the existing network configuration of a 5 communication network quickly, cost-effectively and efficiently, taking into account the heterogeneous demands of different services for telecommunication, such that subsequently an equally good, usually, however, improved network configuration with regard to throughput and network quality is available, while maintaining the existing network coverage. In doing this, any change of cell location is not a proper optimising 10 measure. It is the object of the invention to achieve this optimisation or at least to provide the public with a useful alternative.

Solution of the Problem This problem is solved with the method mentioned in the beginning by the features of the 15 characterising portion of claim 1. Furthermore, the problem is solved with a device of the type mentioned in the beginning by the characterising features of claim 14.

According to the invention, the problem is solved, in particular, in that by one-time or multiple, iterative variation of certain settings of the network configuration, in particular 20 of the aerial orientation or of the transmission power of a reference signal, for example of a pilot channel, an improved network configuration is determined in cells selected by the method, for which configuration the cell area-referenced subscriber traffic in the considered network area is balanced as far as possible. With ideally balanced traffic distribution between the cells, the traffic can be optimally conducted with the 25 predetermined cell locations. The method is supported by a graphic display of the cell areas, which are coloured with colours which correspond to the traffic of the respective cell. Thereby, problem areas with high traffic can easily be identified, and the network configuration can be specifically adjusted, at these locations, in accordance with the invention. Another possibility of graphic support of identification of problem regions 30 with regard to other quantities in a mobile communication network is suggested in WO 00/28756. 4 On principle, the method of the invention may be employed for any cellular, wireless communication networks. In particular, the method is appropriate for third generation networks, which, for example, operate in accordance with the UMTS-standard and which offer to the mobile subscribers a plurality of different services of telecommunication. It is 5 an advantage of the method that any number of provided services in the communication network having different demands, when optimising the network configuration, can be included.

A further advantage of the method is the high speed, as compared to other methods, with 10 which results can be achieved. The reason is that the method is based exclusively on measured data or planning data, respectively, of the communication network or on a combination thereof, these data, as a rule, being available beforehand, as they are required in the planning procedure. No simulation of the communication network behaviour within the iterative optimising process is required, as this is the case, for 15 example, in DE 196 19 208 Al. The method can also easily be automatised, which permits computerised realisation. The optional verification of network configurations matched by the method of the invention is independent of the optimising method proper, whereby the validation method can be freely selected and even exchanged. Besides of this, the validation because of this independence, gets much more significance and, on 20 the other hand, is possible without expensive field measurements. The accuracy of the method can be enhanced as desired by taking into consideration additional data from the, under certain circumstances, simulative validation of the matched network configuration as well as further measuring data from the real communication network.

Advantageous modifications and alternatives of the method as well as of the device are subject matter of the sub-claims.

Explanation The invention relates to a method of optimising the throughput of messages in cellular, wireless communication networks as well as to a device for carrying out the method. These are, in particular, cellular, wireless communication networks with mobile subscribers, which offer a service or a mixture of a plurality of services for K:\53\5387\5387.01\W05387.01-NV-050429-EnglUebers.doc telecommunication to one or a plurality of subscribers simultaneously, as this is the case, for example, in mobile telecommunication networks in accordance with the UMTS-standard. An exemplary configuration of such a communication network with four base stations B110, B120, B130 and B140 is illustrated in Fig.l.

It is a decisive feature of the invention that no elaborate simulations, investigations or measurements are required for the carrying out of the method. Rather is the method based on data, which, as a rule, are available and are required anyhow in the network planning and optimising process. This ensures a quick carrying out of the method based on the 10 requirements of network planning. The general course of the method is illustrated in the block diagram of Fig.2.

In accordance with step 210 of Fig.2, it is assumed that location-referenced data for the communication network or for the network section to be optimised, respectively, with 15 respect to the subscriber traffic and the radio network coverage are available. In this respect, it is necessary that these data are provided with sufficient resolution per area, to ensure that for the covered region of any cell of the communication network, in each case a substantial number of values of the said quantities are available. An appropriate, though not compulsory, arrangement of these data has the form of a rectangular matrix 20 consisting of many, relatively small rectangular area elements, each of which has associated therewith the respective value of the said quantities.

The subscriber traffic is the traffic per service provided in the communication network which traffic has been determined at different locations. The subscriber traffic data can 25 be obtained both by measurements, numerical or analytical methods or simulations and by any combination thereof.

The network coverage required as input is derived from the cell-referenced reception power of a reference signal transmitted with constant transmission power by each cell, 30 the reception power being detected at different locations. The reference signal may, for example, a pilot channel, as used in many mobile communication networks, among others for the radio identification of subscriber mobile stations. If no such reference signal exists in a communication network, a virtual reference signal may be assumed, K:\53\5387\5387.01\W05387.01-NV-050429-EnglUebers.doc which is transmitted by each cell of the communication network with the same transmission power. In the same way as with the subscriber traffic, the data of the reference signal reception power may be gained both by measurements, numerical or analytical methods, respectively, or simulations and by any combination thereof. Often, location-referenced path loss data of the electric field strength from each base station location exist, for example in matrix form, by means of which the location-referenced reference signal reception power can be computed, taking into account the transmission power of the reference signal, the aerials used and other radio radio propagation losses, as for example, a body damping.

At each location where reference signal reception power data are available, the cell having the best reception power is determined in a model of the communication network. In this way, association of area elements with cells is provided. All area elements which are associated with a certain cell are, at large, considered as cell area (in the following also merely "cell"). Because of the physical propagation conditions in the radio channel, which are valid for all services provided in the communication network, as well as due to the cell change algorithm for ensuring the subscriber mobility, which, as a rule, is based on the reception power of the reference signal as measured at the subscriber mobile station, the cell areas determined in the communication network model can be used as good approximation of the actual cells occurring in a communication network. In the exemplary section of a mobile communication network of Fig. 1, the cells being formed around the base stations in the manner just described are plotted in solid lines.

The aim of the method is the improvement of the throughput of messages of in the communication network by an optimised distribution of the provided subscriber traffic between the cells. It has been found that the subscriber traffic of all services can then be conducted with best quality, if a traffic distribution as balanced as possible prevails. Insofar, this method aims at establishing this balance of the traffic between the cells. Thus, in order to achieve this object, certain cell areas have to be shifted relative to each other to permit interchanging of the subscriber traffic, which, apart from this, is invariable and stationary, between the cells. Generally, the subscriber traffic per cell area is determined by accumulating the present traffic values within the respective cells in the model of the communication network. As, with a plurality of services, the traffic per K:\53\5387\5387.01\W05387.01-NV-050429-EnglUebers.doc 7 service at a certain location can be quite different, the traffic values of different services have to be summed up appropriately weighted. The weighting depends on the radio interference effectively generated by the respective service and the utilisation of the resources. Therefore, depending on the communication network, parameter as, for 5 example, the system band width, the channel band width, the ratio to be achieved of signal power and interference power, the spreading factor, the average or maximum data rate, the activity factor, the code limitation enter the traffic weighting. In Fig. 1, the traffic values per line summed up in this way are designated by V 110, V 120 ans V 140.

After the cell-related values of the subscriber traffic are available, at first in accordance with step 220 in Fig.2, the reached approximation on the desired substantial balance of the subscriber traffic between the cells has to be evaluated. To this end, as one possible variant of the invention, a graphic display of the investigated mobile communication network can be used as assistance, similar to Fig.l, the individual cell 15 areas being filled with color associated with the subscriber traffic carried out therein. If, after investigation in step 230 of Fig.2, an adequate balance of the subscriber traffic between the cells has already been reached, the method can be terminated, and the network configuration thus found can, in the simplest case as shown in Fig.2, be made available to the network planning as improved basis for planning in accordance with step 20 270. If this state has not yet been reached, this display serves, in step 240 of the method, to quickly locate problem zones, in which adjacent cells carry out quite different traffic. In another variant of the invention, the identification of the problem regions can also be effected in automated way by a numerical method, in which the differences of the subscriber traffic values between adjacent cells are investigated. This automated step is 25 then incorporated in the total optimising method consisting of the steps 220 to 250 of Fig.2 for minimising the differences of the subscriber traffic values between the cells.

A local displacement of cell areas following after step 250 in Fig.2 can be achieved by the variation of certain network configuration parameters at individual network elements 30 in the model of the communication network, in particular by appropriate matching of the aerial inclination angle or of the aerial height of the respective applied aerial diagram or, for example, matching the transmission power of the reference signal at selected cells. Because, as described above, all these parameters enter the determination of the reference K:\53\5387\5387.01\W05387.01-NV-050429-EnglUebers.doc 8 signal reception power, which is of decisive importance for the form of the cell areas, cell surfaces can be changed also in this way. If, for example, in the network configuration of Fig.l a locally limited high subscriber traffic V 110 was detected in the cell of base station B 110, an appropriate adjustment of the network parameters should be made, 5 according to the invention, which causes reduction of the cell area and, thereby, causing transfer of part of the traffic from this cell into adjacent cells. After the network configuration has been adjusted, the displacement of the cell areas shown by dashed lines has been reached. Thereby, the stationary partial traffic values V 112, V113 and VI14 have been subtracted from the traffic V 110 of the cell at base station B 110. These 10 values were distributed to the cells of the base stations B 120, B130 and B 140, as shown in the Figure, these cells now having the traffic values V 120 + V 112, VI30+ V 113 and V 140 + V 114, respectively. Then the network configuration thus achieved is better, if thanks to the changes made the subscriber traffic values of the individual cells are more balanced than before.

Preferably, the variation of the cell areas to achieve substantial balance of the subscriber traffic between the cells is made iteratively, so that the original network configuration can be approximated step-by-step to an optimised network configuration. In one iteration step, only one or few parameters of the network configuration are changed. Thereafter, 20 the approximation to the desired balance of the subscriber traffic between the cells is checked anew in accordance with the steps 220/230 of Fig.2, the new cell areas as defined by the varied network configuration forming the basis therefor. From the result of the test, the necessary variation of the network configuration in the next iteration step is derived. This procedure may also be automated by means of numerical methods. The 25 stepwise approach can be terminated, when a sufficient balance of the subscriber traffic between the cells has been reached. In particular, it is a criterion that the traffic in the cells with the originally highest subscribers traffic has been clearly reduced as compared to the average subscriber traffic in the communication network model. Such a reduction of the traffic reduces the local interference level, whereby also the network quality may 30 be improved.

If a network configuration has been reached which exhibits a considerably more balanced distribution of the subscriber traffic between cells than the original configuration, the K:\53\5387\5387.01\W05387.01-NV-050429-EnglUebers.doc 9 result of the network matching of step 260 of Fig.2, according to an advantageous variant of the invention, may be validated. The validation may, for example, be effected by means of a network simulation. In doing this, preferably it should be aimed at taking into account, with this simulation, the time response of the network, in order to be able to 5 assess the network quality, though this is not absolutely necessary. As a measure of network quality, for example, the number and locations of blocked switching attempts or broken connections, respectively, the developed error rate and delay of connections, or the throughput per connection or per cell may be used. Of course, also after the network matching, an adequate network coverage has to be ensured, i.e. the various services have 10 to be available at the predetermined locations.

Results of the communication network may, subsequently, as a further advantageous variant of the invention, be used to employ them in a repeated optimisation of the communication network, in order to increase the accuracy of the method. Such quantities 15 may, in particular, be the subscriber traffic actually occurring in the network simulation, but also quantities representing the radio interference. The latter include, among others, the so-called "noise rise", which characterises the proportion of the radio interference relating locally to one cell and caused by other subscribers as referenced to the total interference level of the cell in the uplink. This quantity provides, in particular for 20 CDMA-networks, interesting indications with regard to the local network utilisation. Also information with regard to the local occurrence of call rejections, interruption of connection or failed cell change attempts can be used as boundary conditions for subsequent optimisation of the communication network. Furthermore, also quantities such as, for example, the ratio of transmission power of the reference signal to the total 25 interference power per cell can be used for this refinement of the method of the invention.

In similar way, in a further variant of the invention, such network simulations discussed above can be applied already beforehand to the original network configuration, in order 30 to permit comparison with the optimised network configuration with regard to network quality. Thereby, also the mentioned additional quantities from the network simulation may enter the method of the invention. Besides simulatively determined traffic values as direct input quantity, in particular further boundary conditions for the method may be K:\53\5387\5387.01\W05387.01-NV-050429-EnglUebers.doc fixed. This applies, for example, to the various reasons and the local limitations of interruption and rejection of connections, the analysis of which provides interesting information for a fine control of the method of the invention. As a variant hereto, also measured values of other quantities from the real communication network, in particular 5 of the quantities mentioned before, can enter the initial validation of the original network configuration.

In order to take advantage of the improved network configuration determined by the method of the invention, the method is incorporated in the device of the invention shown 10 in Fig.3. For a more detailed explanation, there is discrimination between the radio access network 310 and the message switching network. 330, which can communicate with each other through a bi-directional interface. The radio access network is that part of the communication network which establishes the direct radio connection to the mobile subscribers, i.e. it comprises substantially the base stations of the communication 15 network and, if required, further components, inasmuch as those are necessary for the direct communication with the subscribers. The message switching network establishes the networking of the base stations in the communication network together with switching stations and gateways to other networks or service providers. Through the interface 320, both the transmitted subscribers' messages and manifold signalling and 20 control information can be transmitted in both directions. A database 350 is provided for the message network, in which position data, configuration data of the individual network elements, measuring data of various quantities from the message network, counter data from the network elements, call statistics, planning data etc are combined. In this connection, the database symbol in Fig.3 at first represents merely a logic 25 aggregation of all these data. Depending on the implementation with a network provider, this data collection may also be realised by a plurality of databases coupled as required. Data from the message network can, at any time, be stored in the database through the bidirectional interface 340, also new network configuration can be transmitted therethrough to the message network. In addition, optionally, quantities and parameters from the 30 message network measured by means of a mobile measuring device, not necessarily motorised as shown in the Figure, can be stored in the database K:\53\5387\5387.01\W05387.01-NV-050429-EnglUebers.doc 11 For carrying out the method of the invention in the device, there is a computer device 370, which contains a numerical implementation of the communication network in the form of a communication network model 380.This model reflects the initial network configuration, which has been loaded into the database through the bi-directional 5 interface. Now the method 390 of the invention, as shown in Fig.2, works in the computer device on the basis of the communication network in order to establish an improved network configuration. In doing so, the changes of the network configuration are made in the numerically implemented communication network model 380, whereby subsequently the improved configuration is set directly there. If required, optimisation of 10 the message network may be controlled and monitored by a communication network optimiser. This exertion of influence may, however, be substantially limited by the ^ automation described above.

Now, the thus determined improved network configuration can be transferred through the 15 interface 360 at first to the database and from there into the message network. Measuring values from the message network 310,330 itself or through a measuring device 399 permit completing evaluation of the achieved improvement of the throughput and the network quality. Such measured quantities may, in turn, also be submitted to the computer device, in order to contribute there to the validation of the improved network 20 configuration.

^ As, in a message network, in particular with a plurality of provided services, the local subscriber traffic per service can change quite heavily also short-time, i.e. several times during a day, it is appropriate to adapt the network configuration promptly to the changed 25 conditions with regard to traffic. This can be achieved with one variant of the device of the invention, wherein the high speed of the underlying method is advantageous. In this variant, the current state of the network configuration and of the traffic, which are stored in the database, are transferred to the computer device 370 and are there, at once, processed with the method of the invention. Subsequently, the promptly determined 30 network configuration better matched to the instantaneous network conditions is re-transferred to the database, from where it is transferred directly into the message network and becomes active there. This permits comparatively quick reaction to a varying subscriber traffic in the message network.

K:\53\5387\5387.01\W05387.01-NV-050429-EnglUebers.doc Though the present invention together with its advantages have been explained in detail, it should be made clear, that various further modifications and variations of the invention within the meaning of the above descriptions can be realised, without departing from the spirit and scope of the invention as covered by the claims.

The example of an embodiment described below serves to illustrate the method of the invention of optimising message networks by means of a selected section of a UMTS/FDD communication network, the general features of which are, for example, described in H. Holma and A. Toskala (ed.), "WCDMA for UMTS", 2nd Ed. John Wiley & Sons, Chichester UK, ISBN 0-470-84467-1. The starting point represents a region covering about 10 km2, in which a total of 15 transmitter stations with 3 cells each are arranged and configured in accordance with the planning methods to-day conventional and corresponding to the state of the art. For each transmitter station, the path loss of receiver locations arranged in a matrix (path loss matrix) is computed with a spacial resolution of 10m in accordance with the present state of the art by means of radiation-optical propagation models, making use of a detailed 3D-surrounding area model. Furthermore, information with regard to the location-related traffic density in the form of traffic matrices with a spatial resolution of 25 m are forms the basis of a speech service (Speech) and a WWW-service (WWW).

In the embodiment, the initial network configuration is analysed by means of a dynamic UMTS-system simulator with regard to capacity and quality of the telecommunication. Subsequently the communication network is optimised by means of the method of the invention and the changed network configuration is again subjected to a dynamic analysis with regard to effected changes in capacity and quality.

For each cell, the resulting equivalent traffic is computed on the basis of the underlying traffic models and the available location-related traffic density information. The local supply area of each cell as resulting from the wave propagation conditions and the network configuration can be maked in colours, the colour code corresponding to the equivalent traffic computed before. In this connection, not the absolute value is decisive but the relative deviation of the values of adjacent cells.

K:\53\5387\5387.01\W05387.01-NV-050429-EnglUebers.doc For cells adjacent in space a quite different traffic to be conducted is to be expected. After the analysis of the network configuration by means of a dynamic system simulator with regard to capacity and quality, according to the invention, a communication network is determined by iterative variation of the aerial orientation and the transmitter power of the reference signals, alone based on the given location-related traffic and radio propagation data, this configuration providing, as compared to the initial configuration, a considerably better balanced traffic to be expected.

It can be shown that network configurations determined by the optimising method of the invention provide a better balanced distribution of the traffic to the cells. Thereby, it should be possible to better utilise the available global communication network capacity.

It is apparent, that by means of the optimising method of the invention the original blocking and connection interruption probability can be reduced in the example. Nearly all service demands can be admitted and served with the optimised network configuration. In the example, an increase of 11 percent of the connection simultaneously served by the communication network under the general conditions described can be demonstrated.

K:\53\5387\5387.01\W05387.01-NV-050429-EnglUebers.doc 14

Claims (21)

Claims

1. Method of optimising communication networks, in particular cellular, wireless communication networks with mobile subscribers, wherein a service or a mixture of a plurality of services for telecommunication is provided simultaneously to one 10 subscriber, and wherein values both of the subscriber traffic per provided service and the cell-referenced received power of a reference signal transmitted with constant transmission power by each radio cell, both of which being obtained either by measurements, analytical methods or simulations or any combinations thereof, and referenced to location are provided, with sufficient resolution, for very small, 15 substantially rectangular area elements, in particular in the form of matrices of these quantities covering a certain region, wherein 20 an iteratively optimised network configuration is provided, in a model of the communication network for use in the real communication network, by accumulation of appropriately weighted, service-specific subscriber traffic values of the rectangular area elements and by locally limited adjustment of the original network configuration. 25 30

2. Method of optimising communication networks according to claim 1, wherein path losses during signal transmission, in particular those which are caused by the distance-related damping of the transmitter power, are taken into account in the model of the communication network.

3. Method according to claim 2, wherein the path losses are taken into account, when weighting the service-specific subscriber traffic values. INTELLECTUALrPR^^Tg^E.

4. OF AJ.

5. Z C:\Documents and Settings\th\Local Settings\Temporary Internet Files\OLKl\W05387.01-NV-05il429-EnglUebers.doc I 0 OCT 2005 15 Method according to any one of the claims 1 to 3, wherein the weighting of the individual service-specific subscriber traffic values is calculated, in each case, depending on the radio interference effectively generated by the associated service, and the utilisation ratio of the resources per subscriber, parameters as, for example, the system band width, the channel width, the ratio of signal power to interference power, the activity factor, the code limitation etc. of the respective service being taken into account depending on the radio standard vised in the communication network.

6. Method according to any one of the claims 1 to 4, wherein a calculatory implementation of a mathematical optimising method is used for iteratively changing certain settings of the network configuration to achieve maximal balance, between the individual cells, of the distribution of the subscriber traffic per cell.

7. Method according to any one of the claims 1 to 6, wherein the optimised network configuration is validated by subsequent network simulation of an adequate radio network model.

8. Method according to any one of the claims 1 to 6, wherein the network simulation applied to the optimised network configuration in the radio network model operates in accordance with a dynamic network simulation method, which is, in particular, distinguished in that it adequately models the time-related behaviour of the communication network in order to determine meaningful characteristics for the evaluation of the network qualities.

9. Method according to anyone of the preceding claims, wherein the network simulation applied to the optimised network configuration operates in accordance with a statisft'c network simulation method, which, in particular, is distinguished in that it models a large number of instantaneous states of the communication network ("Monte-Carlo-simulation"), as a rule without considering the preceding or following states, in order to determine statistic characteristics for the evaluation of network qualities. 16 Method according to any one of the claims 1 to 8, wherein the respective distribution of the subscriber traffic on the individual cell surfaces as well as, if desired, further parameters from input data of the method and/or network simulation results are graphically displayed in a map of the area covered by the communication network, at best before and after each change of the network configuration, at least, however, at the beginning and at the end of the method, in order to determine the optimal network configuration, the said traffic distribution to the cells being illustrated in such a way, that, in the graphic illustration, each cell surface is coloured with a colour unambiguously associated with the provided subscriber traffic.

10. Method according to claim 9, wherein the graphic display of the subscriber traffic per cell and, if desired, further parameters from input data of the method and/or network simulation results is used to identify those cells which clearly deviate from the average subscriber traffic provided per cell, to achieve, with these cells, a uniform distribution of the subscriber traffic between the cells, in iterative consecutive steps of the method by variation of certain parameters of the network configuration, if necessary taking boundary conditions into account, which are defined by the further displayed parameters.

11. Method according to any one of the claims 1 to 10, wherein , in each iteration step, a matrix is determined from an analysis of the geographic neighbourhood relations of the cells, the columns and rows of this matrix being designated by the cells of the investigated network section and in the fields of this matrix the differences of the traffic values of the respective cells and, if appropriate, further boundary conditions are inscribed, in order to determine, therefrom, in an automated process, at least one cell for which subsequently certain network configuration parameters are varied to improve the network configuration of the radio communication network.

12. Method according to any one of the claims 1 to 11, wherein the initial network configuration is evaluated by network simulation of an adequate intellectual property office i of n.z 10 OCT 2005 17 10 25 radio network model, and certain quantities , in particular the cell-related throughput as additional inputs as well as other results, in particular the reasons and locations of locally occurring denials of connection desires and connection interruptions, are entered as additional boundary conditions into the optimisation of the communication network.

13. Method according to any one of the preceding claims, wherein the initial network configuration is evaluated by entering certain quantities from the real communication network, in particular of the cell-related throughput, and these measured values are entered as additional inputs into the optimisation of the traffic distribution between the cells of the communication network.

14. Device for optimising communication networks, in particular of cellular, wireless communication networks with mobile subscribers, which comprises a radio access network for wireless connection of the mobile subscribers and a communication network with fixedly networked transmission channels, further comprising a database with appropriate bi-directional interfaces to the communication network, the database containing, in particular, current position data, configurations of the network elements, measuring data of various quantities from the communication network , counter data from the network elements, call statistics, planning data, wherein simulation means are provided, which generate a virtual model of the communication in a computer device, the configuration of which model is loaded from the database with the current state of the network configuration, and the current state of which model, in turn, can be transmitted to the database, the configuration of the model being matched by a method implemented in the computer device in accordance with any one of the preceding claims.

15. Device according to claim 14, wherein the fixedly networked transmission channels for message transmission are subdivided to be freely connected up intellectual property office 1 0 OCT 2005 RECElVPn 18

16. Device according to any one of the claims 14 or 15, wherein a matched network configuration transferred from the computer unit to the database is transferred to the message network. 5

17. Device according to anyone of the claims 14 to 16, wherein the database, if required, may sub-divided into a plurality of databases coupled with each other, depending of the type of the stored data.

18. Device according to any one of the claims 14 to 17, wherein 10 measured values recorded with a mobile measuring device in the communication network can be stored in a database.

19. Device according to any one of the claims 14 to 18, wherein the instantaneous network configuration and current traffic data from the database are 15 transferred to the computer device, in order to determine, by application of the optimisation method of claim 1, a network configuration better matching the current network situation, and to return this improved network configuration directly subsequently to the database and, from there, to the communication network, where it becomes effective in real time. 20

20. A method of optimising communication networks substantially as herein described with reference to accompanying figure 2.

21. A device for optimising communication networks substantially as herein described with reference to accompanying figure 3. intellectual PROPERTV office OF N.Z 10 OCT 2005 RECEIVED