WO2010022470A1

WO2010022470A1 - Systems and methods for server based zone detection

Info

Publication number: WO2010022470A1
Application number: PCT/AU2009/001123
Authority: WO
Inventors: Malcolm David Macnaughtan; Craig Andrew Scott; Mario Joseph Sammut
Original assignee: Seeker Wireless Pty Limited
Priority date: 2008-08-29
Filing date: 2009-08-28
Publication date: 2010-03-04

Abstract

Methods and systems of generating a profile representative of a region about a mobile radio terminal in a radio communications network wherein a location of the mobile radio terminal includes some uncertainty are disclosed. The methods and systems are characterized by the steps of receiving at a server, a registration request including at least one value of a plurality of radio signal parameters in the region about the mobile radio terminal; estimating a location of the mobile radio terminal and an estimated uncertainty associated with the location of the mobile radio terminal based on the at least one value of a plurality of radio signal parameters; performing propagation modeling to obtain a set of predicted received signal levels associated with nearby cells at a plurality of points within the region about the mobile radio terminal, wherein the distribution of the plurality of points corresponds to the uncertainty associated with the location of the mobile radio terminal; and processing the set of predicted received signal level to generate a profile representative of the region about the mobile radio terminal.

Description

SYSTEMS AND METHODS FOR SERVER BASED ZONE DETECTION

PRIORITY

The present application claims priority from US Provisional Patent Application No. 61/136,346 filed on 29 August 2008. The entire content of this document is hereby incorporated by reference.

CROSS-REFERENCE TO RELATED APPLICATIONS

The present application is related to the following co-pending patent applications:

PCT/AU2006/000479 entitled "Mobile Location"

PCT/AU2006/000348 entitled "Enhanced Mobile Location" PCT/AU2006/000348 entitled "Enhanced Mobile Location Method and System"

PCT/AU2006/000478 entitled "Enhanced Terrestrial Mobile Location"

PCT/AU2008/000344 entitled "Enhanced Zone Determination"

PCT/AU2006/001577 entitled "Detection in Mobile Service Maintenance"

PCT/AU2006/001576 entitled "Mobile Service Maintenance Management" PCT/AU2008/001374 entitled "Systems and Methods for Triggering Location Based Voice and/or Data Communications to or from Mobile Radio Terminals"

PCT/AU2008/001783 entitled "Methods and Systems for Zone Creation and Adaptation."

The entire content of each of these applications is hereby incorporated by reference. Furthermore, the entire contents of the following references is hereby incorporated by reference. W. C. Y. Lee, Mobile

Communications Engineering, McGraw-Hill, 1982, and P. L. H. A. S. Fischer, "Evaluation of positioning measurement systems", T1P1.5/97-110, December 1997, and IEEE VTS committee, "Coverage prediction for mobile radio systems operating in the 800/900 MHz frequency range", IEEE Transactions on VTC, Vol. 37, No. 1, February 1998, 3GPP TS 05.08, and C. R. Drane, Positioning Systems, a Unified Approach, Springer Verlag, 1992, S. R. Saunders & A. Aragon-Zavala, Antennas And Propagation For Wireless Communication Systems: 2^nd Ed, Wiley, 2007.

FIELD

The present disclosure relates to determining the locations of mobile radio terminals in a wireless telecommunications network and, in particular, to systems and methods of server based zone detection of mobile radio terminals.

BACKGROUND

In the majority of markets, charges for use of mobile telephony services are typically greater than the charges for using the equivalent fixed line services. As a consequence, if there is a fixed line available, many mobile subscribers will typically choose to use fixed line in preference to their mobile. This may be at home or at the office.

Mobile operators however are continually seeking ways to increase the usage of their networks. One such way is to encourage mobile users to continue to use their mobiles even when at home, in preference to fixed line services by lowering their charges to be comparable with the fixed line charges. Typically the mobile network operators seek to offer the lower tariffs only in restricted geographical locations such as the home or office, thereby preserving the so-called mobility premium at other times. This is sometimes referred to as zone based charging. The mobile subscriber is charged a normal mobile rate while away from home, but a lower rate whilst at home. A technical challenge that must be addressed in order to offer such services is for the mobile network to be able to distinguish between when the mobile caller is at home and away from home. While there are existing systems which are designed to provide this capability, the performance achieved by these systems is unsatisfactory from a commercial perspective, due either to inadequate reliability within the zone or else excessive zone size leading to excessive leakage of the mobility premium outside the zone. PCT Application No. PCT/AU2006/000478, by the present applicants disclosed systems and methods for home zone monitoring that addressed many of the performance issues, in part through the use of client software deployed at the subscriber's terminal. In some cases however it may not be possible to deploy such software at the subscriber's terminal due either to economic or technical reasons. Certain embodiments of the present disclosure can provide a method and/or system for determining whether a particular mobile is in or out of a particular zone, which overcomes the performance limitations of prior systems without relying on any software deployed at the subscriber's terminal. In the following sections, for simplicity, the description refers to the zone based rating or voice calls. It should be noted however that the zone based rating may equally be applied to other mobile telephony services such as video calling, SMS, USSD, Push To Talk Over Cellular (POC) & packet data.

In conventional systems, a common solution is for the network to rate a call using the identity of the cell from which the call is originated, since the identity of this cell to some extent represents the location of the caller. The identity of the originating cell is checked against a list which defines the zone. This list may be assembled in one of several ways.

The key performance metrics associated with such zone detection systems are the spatial precision and the reliability. The precision is the degree to which the system can discriminate calls made from within the target location (for instance a home) and the rest of the network. Commonly the precision is reported by way of a leakage area- in other words the size of the area where a caller is likely to be deemed to be within the zone and have the discount tariff applied. The term leakage reflects the fact that in this area a mobile operator is losing the premium charge which could otherwise be levied for mobility. This common approach for zone based billing has several problems. One is that the degree of unpredictability in mobile radio propagation makes it difficult to define a precise list of cells which cover a particular location. Including too few cells in the list results in poor reliability and corresponding customer dissatisfaction because they are charged too high a rate despite making calls from within their zone. The usual technique to address this is to add additional surrounding cells to the list. This causes the second major issue which is large leakage areas. In some networks / geographies, the home zone list may include a large number of cells leading to large home zones with correspondingly large revenue leakage.

Some existing systems utilize data produced by radio planning tools in selecting the cells which should define a zone. Such tools can generate for each cell in the network a definition of a polygon which represents the footprint of that cell. These are often referred to as best server polygons. For a given cell, the corresponding best server polygon defines the region within which that cell is likely to be chosen as the serving cell by a terminal.

The zone definition process then consists in selecting one or more cells based on their best server polygons to compose the list of cells defining the zone. One significant challenge in making this selection is that although the polygons may reflect with good accuracy the best server footprint for each cell, they do not directly indicate the smallest number and indeed the specific set of these polygons that should be combined in order to achieve the target zone reliability at the target zone location. As a result a common approach is to select either a fixed number of neighbouring cells are added, in some cases also taking account of the range from the zone location to the respective cells. Some of the performance issues with these existing systems arise from this cell selection approach and the fact that the selection is not directly based on the probabilities of the neighbouring cells serving at the specific zone location or on the target reliability for the zone.

A further problem for existing systems is that this coarse approach to selecting the cells in the zone definition enables only a very coarse ability to adjust the zone size versus reliability if such an adjustment should be required, either by adding or removing a certain number of cells.

Another problem is that as radio networks are modified and as the local environment changes either seasonally or due to construction activity, the pattern of mobile reception around the home may change. Existing systems provide very little in the way of tools to address this. In particular there is no mechanism by which the zone definition can be adapted over time in response to subscriber activity in order to optimize the zone definition relative to the mobile operator's performance requirements.

A further limitation of existing systems arises from their operation according to coarse rules where certain numbers of cells are included in a zone definition. If a customer complains that the service is unreliable in their zone, there is limited facility to investigate the complaint and make specific adjustments to the zone definition to resolve the issue. In some cases, this results in a policy of increasing the zone size by some proportion by default, in turn encouraging subscribers to make further complaints to gain further increases.

SUMMARY OF THE INVENTION

According to exemplary embodiments, a method of generating a profile representative of a region about a mobile radio terminal in a radio communications network wherein a location of the mobile radio terminal includes some uncertainty. The method comprises the steps of receiving at a server, a registration request including at least one value of a plurality of radio signal parameters in the region about the mobile radio terminal; estimating a location of the mobile radio terminal and an estimated uncertainty associated with the location of the mobile radio terminal based on the at least one value of a plurality of radio signal parameters; performing propagation modeling to obtain a set of predicted received signal levels associated with nearby cells at a plurality of points within the region about the mobile radio terminal, wherein the distribution of the plurality of points corresponds to the uncertainty associated with the location of the mobile radio terminal; and processing the set of predicted received signal level to generate a profile representative of the region about the mobile radio terminal.

A method of controlling the performance of a system for determining whether or not a mobile radio terminal is within a predefined region in a radio communications system is also disclosed. The method comprises the steps of associating a zone magnification parameter with a predefined region in a radio communications system; and adjusting the zone magnification parameter to achieve a desired performance.

A method of handling customer complaints in a system for rating customer calls based on an estimated location of a customer in a radio communications network is also disclosed. The method comprises the steps of: receiving a customer complaint having a time, a date and a provisioned zone profile of a call; selecting a network database version describing a configuration of a radio communications network that was in effect at the time of the call; identifying a base station associated with the call from the network database; verifying that the base station, the time, the date, and the provisioned zone profile are consistent with the call being a mis-rated call; determining at least one change to be made to the provisioned zone profile based on a current configuration of the radio communications network.

A system comprising a processing system and a memory coupled to the processing system are described wherein the processing system is configured to carry out the above-described methods. Computer programming instructions adapted to cause a processing system to carry out the above-described methods may be embodied within any suitable computer readable medium. Certain embodiments of the present disclosure can provide systems and methods for zone detection based on the current serving cell, which overcome the limitations in existing systems by providing greater in- zone reliability while at the same time reducing zone size.

Certain embodiments of the present disclosure can provide systems and methods in which the operation of the system can be configured explicitly in terms of key performance measures such as in-zone reliability.

Certain embodiments of the present disclosure can provide a facility to adjust the zone size & reliability with greater precision than existing system, whereby the average zone size may be adjusted in steps substantially smaller than a unit of a cell serving area

Certain embodiments of the present disclosure can detect cases where errors in the radio network database are affecting the performance of a particular zone, and automatically update the zone definition to mitigate the effect of these errors ensuring that the subscribers enjoy an acceptable level of service.

Certain embodiments of the present disclosure make use of provisioned coordinates defining the desired center of the zone, both to verify that any initialisation measurements are collected in the desired location and also to achieve a greater reliability on average for a given zone size.

Certain embodiments of the present disclosure can employ usage statistics to aid the processing for detecting issues with the performance of a particular zone and adjust the zone definition in order to restore the target zone performance.

Certain embodiments of the present disclosure can provide tools for customer care to efficiently investigate customer complaints and determine whether one or more calls in question could indeed have been rated incorrectly as out of zone, and in this event to make any specific changes needed to prevent a recurrence of that issue.

BRIEF DESCRIPTION OF THE DRAWINGS These and other features, aspects, and advantages disclosed herein will become better understood with regard to the following description, appended claims, and accompanying drawings where:

Fig. 1 illustrates an exemplary zone validation process flow in accordance with certain embodiments;

Fig. 2 illustrates an exemplary flow chart of zone generation processing including validation in accordance with certain embodiments; Fig. 3 illustrates exemplary combinations of provisioned and estimated zone locations in accordance with certain embodiments;

Fig. 4 illustrates an exemplary comparison of cell serving probabilities for different zone location uncertainty;

Fig. 5 illustrates exemplary serving probabilities for an idealized network;

Fig. 6 illustrates a comparison of exemplary zone profiles for different zone magnifications in accordance with certain embodiments;

Fig. 7 shows a chart illustrating an exemplary distribution of the number of cells per zone as a function of the magnification in accordance with certain embodiments;

Fig. 8 shows a cumulative distribution of measured zone reliability for different magnification levels aggregated across different environment types in accordance with certain embodiments;

Fig. 9 shows a cumulative distribution of estimated zone size for different magnification levels aggregated across different environment types in accordance with certain embodiments;

Fig. 10 is an exemplary chart illustrating zone reliability versus median zone size in accordance with certain embodiments;

Fig. 11 is an exemplary chart illustrating zone reliability versus median and 95th centile zone size in accordance with certain embodiments;

Fig. 12 shows an exemplary comparison of measured zone reliability with & without using provisioned location in formation in accordance with certain embodiments;

Fig. 13 shows an exemplary comparison of estimated zone size with & without using provisioned location information in accordance with certain embodiments;

Fig. 14 is an exemplary chart illustrating measured zone reliability at varying magnifications with & without using validation in accordance with certain embodiments;

Fig. 15 illustrates an exemplary high level sequence diagram for customer complaint handling in accordance with certain embodiments; and Fig. 16 illustrates an exemplary workflow for customer complaint handling in accordance with certain embodiments.

DETAILED DESCRIPTION

Certain embodiments of the present disclosure will now be described in detail, examples of which are illustrated in the accompanying drawings. The examples and embodiments are provided by way of explanation only and are not to be taken as limiting to the scope of the inventions. Furthermore, features illustrated or described as part of one embodiment may be used with one or more other embodiments to provide a further new combination. It will be understood that the present inventions will cover these variations and embodiments as well as variations and modifications that would be understood by the person skilled in the art.

The following paragraphs detail a method for estimating the probability that a given cell or cells will be selected as the serving cell at a particular location. This method is employed in several different aspects of the processing carried out by certain embodiments of the present disclosure. These include consistency checking between provisioned coordinates and one or more reported cells during registration. Another example is in the selection of cells to include in the zone profile. The serving probabilities are generated by radio network propagation modeling. The model may use information on the configuration of the radio network including the location of cell sites, the height and orientation of cell antennas, the radiation pattern of the antennas as well as the channel frequencies and/or any other codes allocated to each cell. The model may also cover the loss in signal power as radio signals travel from transmitter to receiver.

Predicted received signal powers may be generated for any or all cells in the network. These power levels may also be used to derive interference level estimates so that the received quality of a signal from any particular cell can be predicted. Examples of the application of network model for predicting received signal levels and interference levels can be found in W. C. Y. Lee, Mobile Communications Engineering, McGraw-Hill, 1982, and P. L. H. A. S. Fischer, "Evaluation of positioning measurement systems", TlPl.5/97-110, December 1997, and IEEE VTS committee, "Coverage prediction for mobile radio systems operating in the 800/900 MHz frequency range", IEEE Transactions on VTC, Vol. 37, No. 1, February 1998. Other suitable predictive models may also be used. Using the model, predictions for all cells within a suitable range of the desired zone may be obtained. Preferably in addition to providing a signal power prediction, the modeling tool can also provide a prediction for the degree of variation that may typically be expected for signals in that area and or a confidence interval around the predictions.

The allowance for the variation in the power levels can be adjusted to account for other variations due to effects such as multipath & shadow fading. One suitable representation for the received signal level is a statistical log normal distribution. The parameters of this model are the mean and standard deviation.

Values for the standard deviation typically range between a 2 to 3 dB and 20 dB, depending on the degree of variation anticipated. Optionally the variation may be set differently for different cells based, for example, on their local environment. In a dense urban area larger values would typically be used, for instance between 9dB and 15dB, between 12dB and 2OdB, or between 14dB and 25dB. For less dispersive environments or environments with less clutter smaller values may be suitable, for instance between 3dB and 9dB, between 6dB and 12dB, or between 8dB and 15dB. The values may also be varied according to characteristics of the respective cells such as antenna height.

Alternative models include Rayleigh and Rician distributions. These may be suitable depending on the specific application. For example, a zone associated with an indoor or mixed area may be more suitably modeled with a Rayleigh distribution since there is a lesser likelihood of a direct line of sight to the cell antenna. Conversely a zone associated with chiefly outdoor areas may be more suitably characterized with a Rician model. The expected variation may also be suitably represented by other measures such as inter-quartile range.

Having obtained the signal power and variation predictions, the serving probability for each cell, or some subset of the cells, in the vicinity of the zone may be calculated by assuming that the serving cell selection process in a mobile terminal operates by simply choosing the strongest cell at any time. In an alternative embodiment, the serving cell selection process may be modeled to take into account mobile radio terminal and radio network variations, for example Preferred Roaming Lists in CDMA systems and the idle cell reselection algorithm parameter settings in GSM and UMTS. Thereby it is possible to calculate the probability for each of the cells that it will be selected as the serving cell whilst the terminal is in the zone. For example, this can be done in the following fashion:

Number the cells 1 through N. Define R₁ as the RxLev of the /' cell. Denote P(R₁ , ... , R_N ) as the joint probability density function of R₁ , ... , R_N . From C. R. Drane, Positioning Systems, a Unified Approach,

Springer Verlag, 1992, page 77, the probability of the first cell having the highest RxLev (and so being the serving cell), is given by

P_l

..., fi_jV) (1.1)

The probability that the i'^h cell is the serving cell can be derived from the equation ( 1.1 ) by a simple rearrangement of the terms.

Equation (1.1) is generally applicable to any joint probability distribution function ("p.d.f."). In a preferred embodiment, it is assumed that the RxLev's are independent from each other and are given by a standard Gaussian p.d.f,

where R₁ is the mean RxLev of the i'^h cell site and σ, is the standard deviation of the i'^h cell site. Assuming this form for the joint p.d.f, then equation (1.1) becomes P₁ = C G (^f) dR₁ Ji G (5^5) dR₂ ... Ji G (^f) dR_N (1.3)

After some manipulation, equation 1.3 can be reduced to

where OXx) is the cumulative distribution function of a standard Gaussian random variable and

δ_u = R, - R_l . For / > 1 we have that

where S_{1 k} = R_k - R, , with i ≠ k .

Equations 1.4 and 1.5 expresses the probability of being serving cell as a function of the i?₍ and σ, in a form that is readily and quickly evaluated using standard numerical techniques.

If additional information on the network configuration parameters which affect the serving cell selection is available then these can be incorporated in the probability calculations as well. Such information in GSM systems may include, without limitation, BCCH Allocation lists per cell, Cl and C2 thresholds as well as penalty times. The GSM idle mode cell selection process is described in 3GPP TS 05.08 which is hereby incorporated by reference. In a UMTS system, such information may include, without limitation, Qqualmeas, or Qrxlevmeas. The UE cell re-selection process is described in 3GPP TS 25.304 which is hereby incorporated by reference.

An extension to the use of modeling tools can be used where real measurement data is available for the region of interest. So-called drive-test data if it has been collected in the target zone can be incorporated in the profile generation process. This drive-test data may, for example, provide accurate measurements of shadowing or other deviations in the radio propagation environment. These measurements may be used to further refine the radio propagation model.

The serving probability calculations described above are based around a specified location. In some cases it may be necessary to perform these calculations around a location for which there is some uncertainty. Preferably the uncertainty is expressed as a 2x2 covariance matrix. The extended serving probability calculation is done by modifying the processing for serving probability calculation described previously as follows. The propagation modeling to obtain predicted rxLevs at the zone location is repeated at multiple points, distributed around the provisioned location in a fashion which reflects the uncertainty associated with the provisioned location. Since the provisioned location uncertainty is preferably modeled with a 2x2 covariance matrix, probability contours will be elliptical in shape. One suitable distribution of sampling points consists of points distributed on an ellipse corresponding to the covariance matrix. 8 points distributed evenly in angle is typically suitable although between 4 and 16 points could be used. The tradeoff is between achieving sufficient coverage with the sampling points and at the same time minimizing the relatively computationally intensive processing for the propagation modeling. Having predicted signal levels for the cells neighbouring the provisioned zone location at multiple sampling points, those levels are averaged before being applied to the processing described previously to obtain probabilities for each of the cells. Typically propagation models predict received levels in logarithmic units of dBm. For the averaging described here, it is first necessary to convert the levels to linear power units, for instance Watts or milliWatts before averaging. The averaged power levels are then converted back to log space before the probability calculation described above is carried out.

Figure 4 illustrates an exemplary calculation of cell serving probabilities. The 3 subplots show the results of the calculations at the same location, for different levels of uncertainty associated with the zone location. Plot (a) corresponds to a location uncertainty of 100m; plot (b) 250m; and plot (c) 500m. The axes of the 3 plots have been scaled identically to enable the relative changes between them to be observed visually. Note that as the uncertainty region increases, the serving probabilities for the nearest cells decrease while the probabilities for more distant cells increase. In each plot, the cells having predicted serving probabilities greater than an arbitrary threshold of 0.001 have been plotted. As expected, the number of cells having probabilities greater than this threshold increases as the uncertainty region grows.

An alternative approach for calculating the serving cell probabilities is to perform the radio propagation simulation over a grid of points and to weight the simulations at each point according to the probability of the zone being located at that point. The uncertainty of the zone location is represented by a position x and a covariance matrix N centred on x. A grid of points is distributed about the nominal location of the zone B. Typically the grid will be evenly spaced with the space between points in one direction denoted δa and in the orthogonal direction denoted δb. Each grid point thus representing an area of A = δα δb. The extent of the grid will encompass the ellipse defined by a multiple of the covariance matrix N. The multiple will typically be between about 2 and about 10. The number of grid points will typically be between about 4 and about 10000.

Each grid point a* = (x_hy,) representing an area A₁ is assigned a weighting according to the probability/?,:

At each grid point the received signal strength r_k from each BTS b_k is estimated using propagation modelling. The received signal levels in watts are weighted by the probability to provide

Pm = ^rkVi (1.7)

For each BTS b_k the weighted power across the grid is computed p — ' ^r< k ∑iPi (1.8)

The averaged power levels are then converted back to log space before the probability calculation described above is carried out.

Yet another alternative approach uses the same modeling as the previous approach but uses a different basis for selecting the cells that define the profile. At each grid point Xj the Rx levels for each cell are used to compute the serving cell probabilities P_ik for each cell C_k at that point. Each cell C_k is assigned a weighting w_k

The profile is defined by the set of cells for which w_k >= T where T is the decision threshold, which may be, for example, from approximately 0.001 to 0.05, from approximately 0.01 to 0.5, or from approximately 0.2 to 1.

In certain embodiments, the previous approach can be used to implement a zone over a larger area. The zone would be defined by a boundary such as a polygon or a circle. The grid points on the inside of the boundary would be assigned a probability p_; = 1 and those outside would be assigned a probability pi = 0. In effect the zone profile becomes defined by the set of cells for which the probability of that cell being a serving cell at one or more points within the boundary exceeds a specified threshold. The computational efficiency of the above embodiment could be improved by only evaluating the serving cell probabilities at a series of points on the boundary and automatically adding to the profile all cells that lie within the boundary.

Cellular networks may include cells which by design only propagate over small areas such as the inside of a building or over a street intersection. Such cells are commonly referred to as pico and micro cells as compared to cells providing the majority of the coverage area which are referred to as macro cells. The discrete nature of the grid sampling process combined with the small area covered by pico and micro cells could result in such cells not being included in a profile. This may be due to the cells not being deemed likely to serve at any grid point due to no grid point being sufficiently close to the cell or it could be due to the averaging process due to the cell having a negligible received power at most of the grid points. Increasing the density of the grid points would improve this but at the cost of increased computation. In some embodiments an alternative approach may be employed in which cells deemed to be pico or micro cells are selected based upon spatial criteria with macro cells being selected using criteria derived from propagation modeling. In certain embodiments the Mahalanobis distance from the zone location to the pico and micro cells is computed and all such cells where the Mahalanobis distance is less than a specified threshold are included in the profile. Typical ranges for specified threshold are between 1 and 10. The criteria may optionally include an upper limit on the number of cells so included. If the upper limit is reached the set of cells with the smallest Mahalanobis distance are included. In other embodiments the spatial criteria may be to include all pico and micro cells within a specified distance of the zone location. This distance may be a fixed value or varied according to the density of the network in the vicinity of the zone.

The zones defined by certain embodiments of the present disclosure correspond to a physical zone location. The process of calculating a suitable zone profile characterizing the zone location is based on modeling the operation of the radio network in that location. Accordingly, in certain embodiments this processing relies on information describing the desired location as a point in a suitable reference frame as an input. Suitable reference frames may include a pair of Latitude & Longitude values in the WGS84 geoid or a pair of Easting & Northing values in a locally defined grid. In some cases there may also be some uncertainty associated with the point. This uncertainty could arise from inaccuracies in data sets held by Geographic Information Systems, which mean that translations from civic addresses to physical coordinates are inaccurate. Such uncertainty is preferably represented as a 2X2 covariance matrix.

This information on the desired location may be obtained directly from the subscriber when they subscribe to the service or alternatively may be captured by the system based on radio communication between the network and the subscriber's terminal. The following subsections describe exemplary options supported by certain embodiments for collecting and using this information. In some cases the MNO may know the address or geographic location where the subscriber wishes to utilize a zone based service. This could be the case because the service is offered on-account and the subscriber has to either visit a store or call customer care to enroll in the service at which time the address can be recorded and provisioned in the system. Alternatively the information could be captured via a self service internet portal. If this information on the target zone location is available, it is advantageous to use it in the process of defining the zone profile. The location could be captured as a civic address and geo- coded by a Geographic Information System (GIS) to obtain the corresponding location expressed either in geographic coordinates (latitude & longitude) or any convenient local grid (in units of eastings & northings). Alternatively a map based application could be used, enabling the subscriber to indicate the location, the mapping application recording the corresponding coordinates. In the following descriptions, these coordinates will be referred to as the provisioned zone location. The means for using this provisioned location will be described in a later paragraph.

In some cases, the provisioned zone location may be captured only with coarse accuracy. This could be the case for instance because the GIS exhibits poor accuracy or street numbers are not represented in its database requiring an approximation for instance by the midpoint of the street. To accommodate such cases the present system provides a facility for the accuracy of the provisioned zone location information to also be specified. Specifically, the provisioned coordinates (hereafter denoted V), are accompanied by a single accuracy value corresponding to the estimated 2DRMS error. The system translates this value into a corresponding diagonal covariance matrix (making the assumption that the error distribution is adequately modeled as a 2D Gaussian distribution with uncorrelated errors in the two dimensions). The covariance of the provisioned location is hereafter denoted as ∑_v. In any downstream processing in the system where the provisioned location is utilized, the algorithms take into account the uncertainty in that location, expressed by ∑_v.

In some cases, the system may capture information indicating the location of the zone based on communication between the user's terminal and the radio network while the user is situated in the zone. This can be achieved by providing a registration mechanism whereby the subscriber can call or send an SMS or USSD request. In an Intelligent Network (IN) based home zone service, such sessions can be subject to special service logic in which the originating cell ID is forwarded to the zone server for use in the zone generation. (Hereafter we refer to the originating cell as the reported celt). The system may use any suitable cell ID based location estimation method as known in the art to obtain corresponding location and associated uncertainty estimates. In the description hereafter we will refer to any zone location information captured in this fashion as the estimated zone location, and designate it as E. The associated uncertainty estimate, represented as a 2x2 covariance matrix, we will designate as ∑E. Compared to the case where the zone location is determined by geo-coding a civic address, there is typically greater uncertainty associated with the zone location in the present case. In order to achieve an acceptable level of zone reliability, the system explicitly accounts for the greater uncertainty in generating the zone.

In certain embodiments, the system may operate using only provisioned coordinates. In other cases, only estimated location information may be used. This latter case may be preferable if the location of the zone is not known to customer care when the service is initialized. One example is when the zone service is offered to prepay customers. The service initialization may be done wherever the prepaid subscriber wishes without requiring them to contact customer care.

In yet other embodiments, both provisioned as well as estimated location may be captured and used in the zone profile generation. In such cases, it may be advantageous to first carry out a consistency check between the two sources of information before proceeding with the zone profile generation. This could detect instances where the address for the zone has been incorrectly captured or an error has occurred in the GIS and incorrect coordinates have been calculated. Without such a consistency check, a zone definition could be created which may not achieve the required reliability in the zone. This is likely to be reflected in incorrect charges in the next bill to the subscriber. Incorrect charges in turn will damage the MNO's reputation leading to operationally expensive support calls and potentially even subscriber churn to competing service providers.

In some cases this consistency check also serves another purpose which is to ensure that the service is being operated at a location which confirms to local regulatory requirements for fixed mobile substitution. As an example, in some markets it is a requirement when offering a mobile home zone service, to restrict the zone to within some geographic region. If the provisioned location is confirmed to be within the applicable region, the consistency check described here can ensure that the resulting zone is consistent with the provisioned location. This consistency check will be referred to herein as zone validation in the description that follows.

The validation processing operates by estimating the probability (using the method described above) that the reported cell would be the serving cell at the provisioned location. This probability is then compared with a threshold probability to decide whether the reported cell is consistent with the provisioned location. In some cases the validation threshold probability may be a single value. Suitable values may range between 0.001 and 0.25. The larger the probability threshold, the greater the degree of consistency required between the provisioned location and the serving footprint of the reported cell. In certain embodiments, the validation threshold probability may be defined as a system parameter which can be adjusted by the operator to achieve a suitable degree of strictness, and which can be adjusted while the system is in operation. In yet other embodiments, the validation threshold probability may be defined on a per-user or even a per-zone basis. Figure 1 illustrates an exemplary workflow for the validation processing. In Figure 1, the system first determines that the validation coordinates have been provisioned. If so, then the system next calculates the probability that the reported cell would be the serving cell at the provisioned location. If the calculated probability exceeds the probability threshold, then the system validates the registration and combines the estimated and validation coordinates. Otherwise, the system rejects the registration.

Typically, the provisioned location will have an associated uncertainty region. In the case that the provisioned location is defined with high accuracy (signified for instance by the provisioned accuracy being set to <= 50m), the serving probability calculation is carried out by predicting the received levels from surrounding cells at the provisioned location. The predicted signal levels are then used together with an assumed standard deviation in that region due to shadow fading and fast fading to compute the respective serving probabilities for each of the nearby cells. The selection of a suitable value for the fading standard deviation can be made as described above. For larger uncertainty values, the method described above for serving probability calculation while taking into account uncertainty associated with the location can be utilized.

An alternative form of consistency check could be implemented which involves computing an estimated location from the registration cell ID and then calculating the range to the provisioned location and comparing this with a threshold range. A disadvantage with this type of consistency check is that the accuracy of the estimated location will vary according to the cell density. Without some normalization for this environment dependent factor, it is difficult to set an effective range threshold. Preferably the criteria used for the consistency check compensates for the effects of the network topology such as cell density.

One such metric that may used in certain embodiments would be probability criteria. The consistency check would take the form of estimating the probability of observing the given registration cell ID if located at the provisioned location. The threshold would then be specified as a probability. Such a threshold could be in the range of, for example, approximately 0.05% to 10% or approximately 5% to 30%. For example, if the threshold were set to 1%, the system would nominally reject registrations correctly made at the provisioned location 1% of the time.

In certain embodiments, establishing the probability criteria could be done by computing the Mahalanobis distance between the estimated location and the provisioned location. The Mahalanobis distance compensates for the cell density because the distance is normalized by the covariance of the location estimate. Preferably, the covariance matrix used would be the sum of the covariance matrices of the location estimate and provisioned location. The consistency check can then be performed against a

Mahalanobis distance threshold. The threshold could be in the range of, for example, approximately 1 to 5 or approximately 2 to 10. If Gaussian errors are assumed then alternatively the Mahalanobis distance could mapped to a chi-squared probability and the threshold set as a probability.

An optional aspect of the validation processing is to reject the zone generation request if the reported cell is not known since it is not possible in this case to measure the consistency with the provisioned location. This could happen if the reported cell has been newly added to the network and the database has not been updated yet. Preferably the operation of the system is configurable via a user controlled parameter. If the parameter is set not to reject the request in such circumstances, the zone profile is generated purely based on the provisioned location. The reported cell is added as a temporary cell as described in PCT Application No. PCT/AU2006/000478.

Figure 2 shows an exemplary process flow when the system is configured to reject zone generation requests when the reported cell is unknown. In the exemplary process, the system first attempts to identify the registration cell. If the cell is not identified, then the registration is rejected. However, if the cell is identified, then the system estimates the origin of the provisioned zone. The system next determines whether the validation coordinates have been provisioned. If so, then the system performs validation processing, for example as described in relation to Figure 1. If the registration is validated then the system combines the estimated and validation coordinates. The system then runs the network modeling routines (including uncertainty estimates) and selects candidate cells to generate the zone.

The zone definition is created through a process involving propagation modeling at the zone location. Because the zone location may not be known perfectly, the propagation modeling takes into account the uncertainty in the zone location as a means to ensure that acceptable reliability is achieved. The actual coordinates used in the propagation process depend on the available information. In the discussion that follows we will denote the location coordinates used to generate the zone definition as Z and the associated covariance as ∑z-

In some cases, the only information on the location of the zone may be the provisioned coordinates and associated uncertainty. In this case we set: Z = V

∑z = ∑v (1.10)

In other cases, no provisioned coordinates may be available, however an estimated location may be available as described above. In such cases we set: Z = E

∑Z = ∑E (1.11) In yet other cases both provisioned location and estimated location may be available. A combined location estimate and associated covariance is obtained by taking the mean of the provisioned and estimated locations after weighting by the inverse of their respective covariances as follows:

_{2 =} VL? + E∑_E-*

(Σ- + Σ- )^"1

∑_z = (∑? + ∑→y^ι _{(U 2})

Figure 3 shows exemplary combinations of different provisioned & estimated locations. The magenta cross & circle 300 represent the provisioned location and associated uncertainty while the green cross & circle 302 represent the estimated location and associated uncertainty. Finally the blue cross & circle 304 show the combination of the two locations and uncertainties as described above. In some cases, the input of both provisioned zone location as well as a Cell ID/LAC to support an estimated location enables the system to work around issues with the radio network database. If a subscriber initiates a call or other event as described above to register the service and the corresponding cell is missing from the radio network database, the system can fall back on the provisioned location only, using that to generate the zone and then also adding a temporary profile entry containing the unknown LAC & Cell ID. This temporary entry would then be resolved when an updated network database was applied, containing an entry corresponding to the unknown cell. The benefit in this case is that the subscriber is not inconvenienced by the lack of the serving cell information in the database, and their zone registration and service initiation completes as normal.

Figure 10 & Figure 11 illustrate exemplary performance statistics measured in a real GSM network across a large number of locations. The zone reliability is the percentage of time for which the serving cell selected by a mobile located within a zone matches one of the cells that model that zone. The ideal zone reliability is 100%. The zone size is the area, expressed as an equivalent radius, over which a mobile has a better than 50% probability of serving off a cell that is a member of the zone profile. Figure 10 illustrates that the system can be tuned to achieve a nominated performance. By reducing the threshold that determines which cells are selected for a zone, the reliability can be improved with the trade-off being that the size of the zone, the leakage area, increases. The zone size on this figure represents the median size across all locations. At the highest threshold setting used the median zone size was approximately 2.75km. At this setting approximately 84% of locations has a zone reliability of 90% or better, approximately 74% of locations has a zone reliability of 95% or better and approximately 63% of locations has a reliability of 99.5% or better. The curves on the graph show that as the threshold is reduced the percentage of locations that meet a given reliability target increases but the tradeoff is the zone size increases. At the lowest threshold setting used approximately 88% of subscribers has a reliability of 99.5% but the corresponding median zone size is approximately 4.3km. Figure 11 presents the same median zone size curves as figure 10 but also includes for comparison the 95% zone size statistic. For example when the threshold was set such that the resulting median zone size was approximately 2.8km, 95% of zone sizes were less than approximately 1 lkm.

Figure 12 & Figure 13 show exemplary performance results measured in a real GSM network across a large number of locations in different environment types. Each plot presents a comparison between two system configurations. In the first case, the system utilized both provisioned zone location information as well as estimated location, based on the reported cell at zone registration while in the second, only estimated location was used. In the former case, the source of the provisioned location was somewhat unreliable. As a result, the provisioned location accuracy was set to 500m for all tests. The comparison between the zone reliability curves shows that notwithstanding the inaccuracy of the provisioned locations, the resulting zones yield greater reliability than if the provisioned locations are not used. The comparison between the zone size curves shows that the availability of provisioned zone location information also results in smaller zone sizes. The reason can be understood from which shows that the uncertainty region resulting from the combination of provisioned as well as estimated zone location is typically smaller than the uncertainty associated with the estimated location alone. Accordingly, the zone generation makes a lesser allowance for location uncertainty in selecting the profile cells.

The process of generating a zone profile involves selecting all the cells needed to achieve a specified target reliability in the zone. Typically the objective is to do this while at the same time minimizing the leakage associated with the zone. The leakage refers to the revenue lost through charging a lower rate when a subscriber was actually outside the zone.

The method approach for selecting the cells in the profile is to start with the cell having the highest estimated serving probability and then iterate across the remaining cells in descending order of estimated serving probability, including additional cells until the sum of the probabilities for the included cells is equal to or greater than the target probability. The serving probabilities for the cells in the vicinity of the zone are calculated as described above, taking into account the uncertainty associated with the location as appropriate. To illustrate the process with a simplified example, Table 1 lists some simulated probabilities for the cells shown in. The cell ids listed in the table correspond to the numbers labeled on the plot. (For clarity only the 20 strongest cells have been shown). In this case, assuming the target in- zone reliability was 95%, only the first 4 cells (yielding a total probability of 0.954) would be included.

Table 1 : cell serving probabilities for idealised network

Typically when the present system is utilized in delivering a home zone service, the primary system performance requirement is in-zone reliability. In contrast to existing systems where the ability to adjust the zone performance is very coarse, based only on increasing or decreasing the number of cells in the zone definition, the certain embodiments of the present disclosure permit finer zone performance adjustments. With existing systems the design and tuning process typically consists of defining the number of nearby cells that will be included in the list. Some testing may be done to check whether this achieves the required reliability. The number of cells to be included may be increased or decreased depending on the results of the testing. Further cycles may be repeated until a number is chosen whereby including that number of cells in a zone list achieves the required performance. A limitation of such systems and the methods for tuning them is that there is very limited degree of precision available to modify the zone performance regardless of the particular nature of each location. Adjustment has to be carried out by modifying the number of cells to be added.

In contrast, embodiments of the present disclosure can enable the zone generation process to be adjusted with very fine precision. It should be emphasized that this does not mean an individual profile can be changed any more finely that be adding or removing cells, however the average performance across all zones can be. This means that a 1% adjustment can be made and while for many zones this may not result in any change, some zones in which the strongest candidate that was previously excluded from the zone by a small margin, will now be included.

This capability means that an operator can carry out some tests in their network and adjust the performance of the system with fine control to achieve the best tradeoff between zone reliability and size to suit their needs. When the system is being used for different classes of service (for instance home zone as well as office zone), different default sizes can be assigned corresponding to each class of zone. Furthermore each individual zone has an associated zone magnification parameter M. Advantageously, by adjusting the zone magnification parameter, the zone generation can be adjusted such that the average zone size measured across a large sample increases in fractions of a cell. For example, the zone magnification parameter can be adjusted such that the change in the average zone size, when measured across a large sample of zones may be smaller than a single cell serving area, for example between approximately 0.01 and 0.2 of a cell serving area, between approximately 0.05 and 0.25 of a cell serving area, or between approximately 0.1 and 0.75 of a cell serving area.

In certain embodiments, the dimensionless zone magnification parameter M, can be applied in the process of selecting cells for inclusion in the zone profile. As described above, the set of cells included in the profile can be formed by, for example, scanning from strongest to weakest predicted cell, including the smallest set of cells whose serving probabilities sum up to the target zone reliability. In this example, the residual or complementary probability would be 1 - target zone reliability. This process as described corresponds to a magnification of 100. For larger values of magnification, the sum probability threshold is adjusted by scaling the complementary probability by M/100. To illustrate, if the target zone reliability was set as 95% (corresponding to a complementary probability of 0.05), for a magnification of 200, the complementary probability threshold would be decreased by 0.05/(200/100) = 0.025. The corresponding sum probability threshold is 0.975, resulting in additional cells being included in the profile, thereby extending the physical extent of the zone. The initial size of the zone is based on the default zone size however via the system provisioning interface, such as the one described in PCT Application No. PCT/AU2006/001479, the magnification parameter may be adjusted, for example, between about 50% to about 300%, resulting in a change to only that specific zone.

Figure 6 illustrates the effect of the magnification parameter for a zone in a real GSM network. Three versions of the profile for a particular zone are illustrated, with magnifications of 100%, 150% and 200%. The extents of the three plots have been aligned so that a direct visual comparison can be made. As can be seen from the plots, an increase in the magnification results in additional cells being added to the profile.

Figure 14 illustrates exemplary performance statistics measured in a real GSM network across a large number of locations to compare the effect of the use of validation coordinates. The validation coordinates, on average, enables the system to more accurately determine which cells are likely to be selected by the mobile as the serving cell within a given zone. The effect is that for a given level of magnification, which is directly proportional to zone size, the reliability achieved when using validation coordinates is higher than when not used. Figure 14 illustrates that this improvement is consistent across a wide range of zone magnification settings. For example with the zone magnification value of 100, without validation coordinates 60% of locations achieve an reliability of 99.5% or better whereas that statistic improves to approximately 63% when validation coordinate are used. Similarly for a magnification of 200 the percentage of locations with a reliability of 95% or better improves from approximately 88% to approximately 92%.

Certain embodiments of the present disclosure apply methods to estimate the serving probabilities for the neighbouring cells and use this to select the smallest number which is required to achieve the target reliability. The result is on average less leakage for a given reliability.

Figure 7 illustrates some aggregate statistics from testing done using the present algorithms in a live GSM network. Each curve shows the cumulative distribution of the number of cells included in the profile. The zones cover a range of environments ranging from dense urban to rural. The variation in the numbers of cells per zone reflects the variation in the cell density in the different environments as well as the individual context of each zone, i.e. the proximity between each zone location and the surrounding cells and their respective configuration. The different curves correspond to different magnification settings as described above.

The system provides a default parameter (referred to as the default zone magnification) to set the operating point. The nominal value of this parameter is 100 (corresponding to a conceptual setting of 100%). In certain embodiments this parameter can be adjusted in steps of 1 percent between about 50% and about 300%. about 0.01 and 0.2 of a cell serving area., or between 0.05 and 0.25 of a cell serving area or between 0.1 and 0.75 of a cell serving area. A value smaller than 100 corresponds to a reduction of the zone size; while a value larger than 100 results in larger zones. A change of 1 in the value of this parameter corresponds to a 1% change in the zone magnification, although the real effect on the zone size on the ground may not correspond exactly to this, being dependent on local geography, cell configuration etc.

Following initial testing and tuning, the internal system default magnification can be adjusted so that the operators preferred operating point corresponds to a setting of 100 in the externally accessible magnification parameter. Thereafter, the operator can adjust the operating point for the system as required. For instance if a higher than expected rate of reliability issues was observed, the default magnification could be raised to 105.

In currently pending applications PCT/AU2006/001577 entitled "Detection in Mobile Service Maintenance" and PCT/AU2006/001576 entitled "Mobile Service Maintenance Management," methods are disclosed for updating zone profiles following one or more changes to the configuration of the radio network. Those same methods may be employed in certain embodiments of the present disclosure to maintain the zone profiles in a current state. For changes that pertain to cell identifiers such as Cell ID & LAC, a corresponding change is made to the profiles. For changes which affect the likely signal reception levels within the zone, the processing described above to generate the zone can be repeated using the updated radio network model to refresh the zone profile.

Some existing network based home zone deployments have experienced non-trivial rates of customer complaints that calls actually made from within the zone are not being charged at the discount rate. Such subscriber complaints may come in several forms ranging from the very vague "My zone doesn't work" to the very specific "The call I made from my kitchen at 3pm ... was charged at the higher rate".

An approach employed in existing systems to handle such complaints is to increase the zone size by adding a certain number of additional cells to the zone definition. This has several disadvantages however. Simply adding one or more of the nearest cells to the zone definition may not actually resolve the particular issue affecting that subscriber. The specific cell in question may be a remote cell, which serves within the home due to particular local morphology. Therefore after adding additional cells, the customer care representative will still be unable to confidently assure the caller that the issue has been solved and will not recur. It is preferable for the system to identify the specific cause for the issue and address it, enabling a decisive resolution for the customer. The benefits include happier customers, particularly when no repeats of the problem occur and lower operational costs due to fewer repeat complaints.

Certain embodiments provide facilities that enable customer care to focus the complaint handling process around specific calls which the subscriber identifies as having been incorrectly deemed as out of zone. By focusing the discussion around specific calls in this way, customer care can ensure that where needed, the zone profile is amended in a way that ensures the problem will not recur. At the same time spurious attempts to obtain a larger zone will be discouraged by the requirement to provide specific instances and the accompanying scrutiny of those instances.

When answering a customer complaint, the customer care representative (hereafter CC) first requests the customer to identify the date and approximate time of the call(s). CC then uses the zone status audit facility to retrieve a list of the calls made from the user's terminal around that time. The core zone server provides an XML interface for a CC GUI application to retrieve these details. CC then attempts to identify one or more calls that the customer believes were treated incorrectly using the list of calls. For each call identified, CC invokes the second stage audit, causing the server to analyse the probability of a call being made on that cell while the subscriber was situated within the zone. The probability calculation involves the same processing as described above for checking the consistency between the provisioned location and reported cell at zone registration. Depending on the result of the analysis, CC can accept the complaint and proceed to modify the zone to address the issue or else reject the complaint on the grounds that the serving probability is too low. The decision threshold is not specified here as it will depend on customer complaint handling policies of the operator, the performance, both technical as well as commercial of the service however typical thresholds might be between, for example, approximately 0.001 and 0.05. If the decision is taken to accept the complaint and resolve the zone definition, CC issues another request to the server to update the profile to accommodate the cell at issue. This sequence is illustrated in Figure 15.

Figure 15 illustrates an exemplary messaging sequence when a customer contacts customer care in accordance with certain embodiments. Initially the customer contacts the customer care center with a complaint. CC then queries the customer for the zone and date of the alleged problem. The customer can then respond with a zone and date and the customer care center forwards this information to the location server (LS) to get subscriber details. The LS responds to the CC with the relevant subscriber details and the Provisioned Zone (PZ). Next, the CC queries the subscriber for a time associated with the alleged problem, to which the customer responds. CC issues a request to the LS for the call details associated with that PZ on the nominated date, receiving a list of records numbered with an Audit Sequence Number (ASN). Otherwise, the CC chooses a specific call at the time nominated by the customer to audit. The CC then submits an audit request to the LS, causing the LS to identify the particular cell from which the call originated and analyse that cell against the zone profile & network database history. If the cell is in the current profile, then the problem has been fixed since the selected call. If the customer's mobile radio terminal is rated HOME for the PZ then there are potentially zone notification issues that can be reported. Otherwise, the LS will report that the cell is not in the profile and will provide an estimated serving probability for that cell at the zone location. If the CC representative is satisfied with the probability returned then they can add the selected serving cell to the zone profile by submitting a command to the LS. The LS will verify the ASN, MSISDN, and PZ ID and generates a synchronous response that adds the cell that was most recently audited for the specified ASN. The LS will acknowledge the add selected serving cell command. The CC can also submit a request to remove previously added serving cells to the LS. In response, the LS will verify that the cell was added using the current interface and if so, perform the request and acknowledge the CC.

An exemplary workflow for this process is illustrated in Figure 16. In the exemplary workflow, initially the CC receives a customer complaint. The complaint should specify the date of an allegedly mis-rated call and the PZ in question. The CC then retrieves the subscriber's call data for the specified date and PZ from the LS. The customer is then queried to specify the time of the alleged mis-rated call and identify the call in a list returned from the LS. If the call was rated HOME, then there is a possible problem with the HOME notification for the user that should be reported. If not, then the CC determines whether the reported cell is in any current profile. If yes, then the CC notifies the subscriber that the problem has been fixed since the call was made. If not, then the CC determines whether a Serving Cell (SC) check resulted in a "NotProvided" message. If yes, then there is possibly an external problem with, for example, the network. If not, then the CC determines whether an SC check resulted in an "UnknownCell" message. If so then there is a possible network database (NWDB) problem. If not, then the CC audits the ASN index associated with the specified cell. If the probability of the cell serving a mobile radio terminal in the PZ exceeds a predetermined threshold, then the cell is added to the PZ. In some cases, the problem may have already been addressed (for instance call was made on a new cell before the NWDB was updated - since then the NWDB update has added that cell to the profile). In such cases the system will indicate this to the CC representative enabling them to advise the customer that the issue has already been resolved.

The use of this facility offers significant benefits to the operator of a zone service. Compared to existing solutions, the detailed, but automated analysis facilities will support an efficient workflow enabling rapid resolution of genuine issues. In addition, the changes made during this process are precisely tailored to address the specific issue. Existing systems require an expert radio network engineer to investigate the issue.

A complication arises when trying to address customer complaints relating to the zone status determination for past calls when the network configuration is subject to change due to ongoing optimization or capacity enhancements. The mechanism described above involves examination of past call records to identify the cell through which the call was established and then checking whether that cell should have been included in the zone definition. The identity of the cell in the call records is typically represented by Cell ID & LAC. If in the elapsed time between the actual call and the customer complaint, a change has been made which affects this cell, it may be difficult to identify the cell in order to perform the analysis described above. One example of such a configuration change is a re-parent where the BTS is associated with a different BSC, resulting in a change of the LAC associated with that cell. Other examples include a change of CID or even decommissioning of that cell. The following paragraphs describe methods for overcoming this difficulty. As disclosed in, for example, PCT Application Nos. PCT/AU2006/001577 entitled "Detection in Mobile Service Maintenance" and PCT/AU2006/001576 entitled "Mobile Service Maintenance Management," the system maintains a database describing the configuration of the radio network. This database may be updated from time to time to reflect the changes in the radio network. The system maintains a history of the changes, including the content of the database as well as the date & time at which each version was applied.

When a customer complaint is to be investigated, the first step is to use the time of the call in question to select the network database version describing the configuration that was in effect at that time. The system then uses the CID & LAC associated with the call to identify the base station. The system also retrieves the version of the zone profile which was in effect at that time.

Having identified the cell and the profile, the consistency check is carried out using the contemporaneous data on the cell and the zone to complete the initial determination as to the validity of the complaint, using the methods described above.

If the complaint is deemed valid, the next step in the process is to determine what if any changes are required to be made to the current version of the profile, given the current configuration of the radio network.

If the cell has since been decommissioned, no change will be made.

In some cases the cell may already be in the profile. This could happen through the automatic reliability enhancements mechanisms described in PCT Application No. PCT/AU/001783 entitled "Methods and Systems for Zone Creation and Adaptation." If the cell is already in the profile, the system advises the customer care agent of this enabling them to confirm to the customer that the issue has already been addressed.

If the cell is still missing from the profile, the system adds a new entry to the profile, corresponding to the cell. This entry would reflect the current details for that cell including any change of Cell ID for instance since the original call.

As noted earlier, while the description presented here is described in terms of voice calls, the same facilities and processing are applicable to other mobile telephony services including video calling, SMS, USSD, POC & packet data.

While the preceding paragraphs have described the use of certain embodiments of the present disclosure primarily in terms of a home zone service, the same system can also be applied to other zone detection scenarios. Location based mobile advertising is one such scenario. To illustrate, assume that a wireless access provider wishes to offer promotions to subscribers at particular locations (zones). These promotions could be related to the provider's own products & services or could be on behalf of other enterprises, as a delivery channel for a third party mobile advertising service.

Certain embodiments of the present disclosure can be used to opportunistically detect when subscribers are situated within one or more such zones, based on events such as call or SMS origination, location updating, and any other such events in which the subscriber's terminal interacts with the access network, revealing its proximity to that particular access point. The advantage in such cases is that the extent of the zones can be designed effectively according to the requirements of the advertising campaign, by adjusting the target reliability in the zone profile generation process.

A further option is to utilize the system as described here in conjunction with a mobile client based zone detection system for mobile advertising as described in PCT Application No. PCT/AU08/001374 entitled "Methods and Systems for Triggering Location Based Voice and/or Data Communications to or from Mobile Radio Terminals."

Compared to prior systems, the systems and methods disclosed herein may offer one or more of several advantages:

In certain embodiments, zone performance goals (i.e. achieving target reliability with the smallest zone size) across the network are optimized because the system adjusts the number of cells per zone, taking into account the local network configuration around each zone.

In certain embodiments, a further advantage is that in the case where it is a requirement to increase the reliability of a zone, the system can select cells to add in order of the increase in the overall zone reliability that each is likely to contribute, thereby increasing the reliability with the smallest increase in zone size on average.

In certain embodiments, another advantage is the ability to make very fine adjustments to the system operating point, enabling an operator to respond to aggregate statistics on customer care calls and optimize the total cost of the solution achieving the desired balance between customer satisfaction and revenue leakage.

In certain embodiments, an advantage is that it makes allowance for the greater or lesser knowledge about the zone location in building the zone definition, thereby achieving the required reliability without having to unilaterally increase all the zone sizes to protect against the cases where the performance would be worse.

In certain embodiments, another advantage is achieved through the flexibility & robustness of the system to support different operating scenarios regardless of whether or not the operator is able to provision the location of the zone in advance. The robustness also means that even if the provisioned information is not highly accurate, the system can make use of it in proportion to the accuracy. In certain embodiments the facility to validate a subscriber registration against a provisioned location for the center of the zone is advantageous as it provides the operator the ability to limit the service to certain locations. Additionally it protects the user against poor zone reliability in the case that the cell measured during registration has incorrect parameters in the radio network database which without the validation step would result in an erroneous zone location estimate and a correspondingly invalid zone profile.

In certain embodiments, another advantage is the ability for customer care representatives using the sophisticated zone audit facility to definitively resolve genuine customer issues while discouraging spurious complaints.

In certain embodiments, a further advantage is the ability to work around cases where the reported cell is not present in the network database, if provisioned zone location is available.

In certain embodiments, another advantage is the ability to perform consistency checks between the provisioned location and the reported cell to detect errors in the provisioned location. This can protect the user from poor zone reliability and the inconvenience of having to contact customer care to complain. In turn this translates into lower operational costs for the operator and lower likely rates of churn from the service.

The exemplary approaches described may be carried out using any suitable combinations of software, firmware and hardware and are not limited to any particular combinations of such. Computer program instructions for implementing the exemplary approaches described herein may be embodied on a tangible computer-readable medium, such as a magnetic disk or other magnetic memory, an optical disk (e.g., DVD) or other optical memory, RAM, ROM, or any other suitable memory such as Flash memory, memory cards, etc.

Additionally, the disclosure has been described with reference to particular embodiments. However, it will be readily apparent to those skilled in the art that it is possible to embody the disclosure in specific forms other than those of the embodiments described above. The embodiments are merely illustrative and should not be considered restrictive. The scope of the disclosure is given by the appended claims, rather than the preceding description, and all variations and equivalents which fall within the range of the claims are intended to be embraced therein.

It will be understood that the term "comprise" and any of its derivatives (eg. comprises, comprising) as used in this specification is to be taken to be inclusive of features to which it refers, and is not meant to exclude the presence of any additional features unless otherwise stated or implied.

The reference to any prior art in this specification is not, and should not be taken as, an acknowledgement of any form of suggestion that such prior art forms part of the common general knowledge.

Claims

CLAIMS:

1. A method for generating a profile representative of a region about a mobile radio terminal in a radio communications network wherein a location of the mobile radio terminal includes some uncertainty, the method comprising:

receiving at a server, a registration request including at least one value of a plurality of radio signal parameters in the region about the mobile radio terminal;

estimating a location of the mobile radio terminal and an estimated uncertainty associated with the location of the mobile radio terminal based on the at least one value of a plurality of radio signal parameters;

performing propagation modeling to obtain a set of predicted received signal levels associated with nearby cells at a plurality of points within the region about the mobile radio terminal, wherein the distribution of the plurality of points corresponds to the uncertainty associated with the location of the mobile radio terminal; and

processing the set of predicted received signal level to generate a profile representative of the region about the mobile radio terminal.

2. A method as claimed in claim 1 wherein the at least one value of the plurality of radio signal parameters is obtained by the mobile radio terminal.

3. A method as claimed in claim 1 wherein the at least one value of the plurality of radio signal parameters is obtained by a radio communications network device in the region about the mobile radio terminal.

4. A method as claimed in claim 1 wherein the at least one value of the plurality of radio signal parameters is obtained by intercepting a communication between the mobile radio terminal and a radio communications network.

5. A method as claimed in any one of claims 1 to 4 wherein at least one of the plurality of radio signal parameters is a cell ID.

6. A method as claimed in any one of claims 1 to 5 wherein the step of processing the at least one value of the plurality of radio signal parameters is performed at a network processor of the radio communications network.

7. A method as claimed in any one of claims 1 to 5 wherein the step of processing the at least one value of the plurality of radio signal parameters is performed at a processor external to the radio Communications network.

8. A method as claimed in any one of claims 1 to 7 wherein the generated profile is transmitted to the mobile radio terminal.

9. A method as claimed in any one of claims 1 to 8 further comprising the step of validating the consistency of the at least one value of the plurality of radio signal parameters.

10. A method as claimed in claim 9 wherein the step of validating comprises calculating an approximate location of the mobile radio terminal and comparing the approximate location with a provisioned location.

11. A method as claimed in claim 10 comprising the further step of, if the approximate location of the mobile radio terminal and the location associated with the region are within a given distance of each other, accepting the at least one measurement, otherwise rejecting the at least one measurement.

12. A method as claimed in claim 9 wherein the step of validating the consistency comprises computing the probability the at least one measurement could have arisen at the provisioned location.

13. A method as claimed in claim 12 wherein if the probability of the at least one measurement arising at the provisioned location is above a given probability, accepting the at least one measurement, otherwise rejecting the at least one measurement.

14. A method as claimed in claim 1 wherein generating the profile of the region about the mobile radio terminal comprises selecting all the cells needed to achieve a specified target reliability in the region, wherein selecting the cells in the profile comprises:

selecting a first cell having a highest estimated serving probability,

iterating across remaining cells in descending order of estimated serving probability, and

selecting additional cells until a sum of the probabilities for the selected cells is equal to or greater than the specified target reliability.

15. A method of controlling the performance of a system for determining whether or not a mobile radio terminal is within a predefined region in a radio communications system, the method comprising: associating a zone magnification parameter with a predefined region in a radio communications system; and

adjusting the zone magnification parameter to achieve a desired performance.

16. A method as claimed in claim 15 wherein the zone magnification parameter may be adjusted from about 50% to about 300%.

17. A method as claimed in claim 15 wherein the zone magnification parameter may be adjusted such that the change in the average zone size, when measured across a large sample of zones may be smaller than a single cell serving area.

18. A method as claimed in claim 17 wherein the zone magnification parameter may be adjusted between about 0.01 and 0.2 of a cell serving area.

19. A method of handling customer complaints in a system for rating customer calls based on an estimated location of a customer in a radio communications network comprising the steps of:

receiving a customer complaint having a time, a date and a provisioned zone profile associated with that customer's service;

selecting a network database version describing a configuration of a radio communications network that was in effect at the time of the call;

identifying a base station associated with the call from the network database;

verifying that the base station, the time, the date, and the provisioned zone profile are consistent with the call being a mis-rated call;

determining at least one change to be made to the provisioned zone profile based on a current configuration of the radio communications network.

20. A method as claimed in claim 19 wherein the at least one change comprises adding a new entry to the provisioned zone profile corresponding to the base station.

21. A system for performing any of the methods described in claims 1 to 20.

22. A processor-readable medium storing instructions thereon for causing a processor to perform any of the methods described in claims 1 to 20.