WO2010052695A1 - Method and apparatus for assessing communication quality - Google Patents

Abstract

A method and apparatus provide for analyzing communication of a user or user-group from call detail records, and assessing quality of service of an operator or network the user is associated with, or of another network or operator providing services to a user or a device with which the user communicates. Analysis is based on examining the call detail records against predetermined models and rules representing the user behavior, which may also include multi-user behaviors or specific cases. The method and apparatus thus analyze quality measure based upon the user's behavior, also in cases of successfully terminated communication sessions. Technical data can be used to complement the call detail records and enhance the analysis results.

Description

METHOD AND APPARATUS FOR ASSESSING COMMUNICATION

QUALITY

TECHNICAL FIELD

The present disclosure relates to assessing communication quality in general and to assessing the communication quality of a network based on user behavior, in particular.

BACKGROUND

Communication in general, and mobile communication in particular is a constantly increasing necessity. In recent years, electronic technology, and communication technology in particular, has changed our everyday lives. Portable electronic devices such as cell phones have become an inevitable part of the lives of large percents of the world population. Mobile communication and computing devices have become the means by which millions of people conduct their personal as well as professional communication with the world. It has become almost impossible for many people who use these devices as a means to improve productivity, to function without access to their communication devices. Therefore, the reliability and proper functioning of the devices, the communication networks and the equipment has become crucial.

A major problem related to mobile communication, as well as to other communication channels, is the quality of communication. Low level of communication quality is an important reason for which people switch operators. However, the communication quality experienced by a user depends not only on the quality provided by the operator he or she are associated with, but also on his or her mobile device, and on the communication quality provided by the operator of-the-person-or-device-he-is-communication-with— Further,— the-quality-of-any— operator may vary according to the geographic location, local and temporal load or other factors. Therefore, when a person is experiencing low level service, it may be due to problems caused by his or her equipment, his or her operator, problems for which the other operator is responsible, or problems with the device he or she are trying to communicate with. When a user calls his or her operator to complain about the service quality, the operator may not be able to help them, since the operator cannot tell problems with the user's equipment from short-term problems with the operator's own equipment or infrastructure, and further from problems with the other operator, wherein the problems of the other operator may also be constant, temporal, local, or relate to the device the user is communicating with. Some problems are easy to detect. For example, in mobile communications, if there are multiple interruptions in communications by people moving between two cells, the problem will be first searched for within the cell transfer mechanism, while if a problem repeats at a certain area, a problematic cell will be searched for. Known techniques for locating further problems within a network involve analysis of technical or engineering data available to, or gathered by the operator. Such data may include data exported by the equipment, by the infrastructure, or by testing equipment. Additional data may be gathered, among others, by testing equipment, such as dedicated network monitoring equipment probes, or performing drive tests.

However, such data provides no information regarding problems related to the user's equipment, or to other networks. Furthermore, the data provides little information to the quality of service experienced by the user. The user's operator has no access to information related to the other operator, so only problems with the operator's equipment or infrastructure can be detected.

Further, testing is limited to the conditions, including load and external conditions at the testing time, so some problems cannot be realized retroactively. Such systems also cannot offer suggestions for solving the problems, and can at most enable the identification of some of the problems. There is thus a need in the art for a method and apparatus for assessing the quality of a first network, as well as the quality of other networks with which the users of the first network communicate.

SUMMARY

A method and apparatus for evaluating the quality of a communication network from the call detail records, including drawing conclusions for calls that were completed successfully. One aspect of the disclosure relates to a method for assessing a quality parameter of a first network or a second network, the first network providing a communication service to a device, the device used by a user, the method comprising: receiving one or more call detail records of communication sessions associated with the device; and analyzing the call detail records against a communication model, and assessing the quality parameter. Within the method, the communication model optionally comprises one or more rules. Within the method, the rules are optionally provided by a human. Within the method, one rule optionally indicates that two communication sessions held during a predetermined time interval, such as a time interval shorter than five minutes between two users indicate a network failure. Within the method one or more of the communication sessions was optionally terminated successfully. Within the method, the quality parameter optionally relates to the quality of voice or data transmission within the first network. Within the method, the device is optionally communicating with a user or a device in the second network, and the quality parameter optionally relates to the quality of voice or data transmission within the second network. The method optionally comprises the step of generating a user or multi-user communication model based on training call detail records of the device, wherein the communication model comprises a user or multi-user communication model. The method optionally comprises the step of generating one or more recommendations for improving the quality parameter. The method optionally comprises a step of enhancing the assessment of the quality parameter using technical call data.

Another aspect of the disclosure relates to an apparatus for assessing a quality parameter of a first network or a second network, the first network providing a communication service to a device, the device used by a user, the apparatus comprising: a call detail record receiving component for receiving one or more call detail records of communication sessions associated with the device; and an analysis component for analyzing the call detail records against one or more communication models and assessing the quality parameter. Within the apparatus, the quality parameter optionally relates to the quality of voice or data transmission within the first network. Within the apparatus, the device is optionally communicating with a user or a device in the second network, and wherein the quality parameter relates to the quality of voice or data transmission within the second network. The apparatus can further comprise a component for generating the communication model based on training call detail records of the device. The apparatus can further comprise a component for generating the communication model based on one or more rules. Within the apparatus, one of the rules optionally indicates that two communication sessions held during a predetermined time interval, such as a time interval shorter than five minutes between two users indicate a network failure. Within the apparatus, the communication model is optionally a user communication model or a multi-user communication model. The apparatus can further comprise a recommendations component for generating one or more recommendation for improving the quality parameter. Yet another aspect of the disclosure relates to a computer readable storage medium containing a set of instructions for a general purpose computer, the set of instructions comprising: receiving one or more call detail records of communication sessions associated with a communication device; and analyzing the call detail records against one or more communication models, and assessing a quality parameter of a network associated with the device. BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will be understood and appreciated more fully from the following detailed description taken in conjunction with the drawings in which corresponding or like numerals or characters indicate corresponding or like components. Unless indicated otherwise, the drawings provide exemplary embodiments or aspects of the disclosure and do not limit the scope of the disclosure. In the drawings:

Fig. 1 is a schematic illustration of a typical environment in which the disclosed method and apparatus are used; Fig 2 is a flowchart of the main steps in an embodiment of a method for quality assessment of communication, in accordance with the disclosure; and

Fig. 3 is a block diagram of the main components in an apparatus for communication quality assessment, in accordance with the disclosure.

DETAILED DESCRIPTION

The disclosed subject matter provides a method and apparatus for a communication operator to assess the communication quality provided to its users, as well as the communication quality provided by other operators associated with other users or devices with which the users communicate. The method and apparatus start with an optional training or preparation stage, in which rules are introduced or information is gathered related to communication networks providing services to end-users. The gathered information includes behavioral information, indicating the communication performed or received by users. The rules may include description of cases, use-cases, or situations in which the user may be receiving low quality service, possible with no other indication. For example, two communications between the same two users at a short time interval, even if at least the first communication ended successfully, may indicate low quality of service. The gathered information may relate to users' communication activity and routines, such as details about the user's communication in a predetermined time frame. Then one or more models indicating the routine behavior are constructed for each user or group of users.

In run-time, which may be in real-time, near real-time, or off-line, the user's activity is automatically analyzed against or in comparison to the model. The user's experience is analyzed, and if possible, improvement suggestions are made to improve one or more quality parameters. The analysis is thus based on the call detail records (CDR) rather than on technical or engineering data, although such data can also be used by the system.

The user's activity, as gathered during the training stage as well as during run-time, may include but is not limited to any subset of the following: the user's incoming and outgoing communications, optionally including the termination manner, i.e., completely successfully or dropped by either side, the numbers he or she are calling and receiving communications from, the operator associated with said numbers, the locations of the user, the date, weekday, date and time of the communications, and any other parameters. Further gathered information relates to data transferring, including destinations, number of sessions, number of packets, number of accesses to services, or the like.

Then, by analyzing the activity of one or more users, an operator can assess the communication experience of users, and may make suggestions or take steps in improving the experience.

For example, analyzing the activity of multiple users of a cellular communication operator can provide the following input: two or more communication sessions, any of which ended successfully or by dropping, between the same two users at a predetermined short time difference, such as under thirty seconds, one minute, five minutes, or the like, may suggest low service quality on at least the first communication, so that the users tried to establish a second communication hoping the service improves. Multiple disconnections for users in a particular area may suggest failure of a particular cell. A similar conclusion may be arrived at from the engineering data, but analyzing the CDR can provide additional details, such as failure time, load at time of failure or other details which are not always available as part of the engineering or technical data.

In another example, multiple disconnections between users and devices associated with another operator, may suggest failure of a cell or another equipment of the other operator.

The analysis of communication by a particular user can lead to further conclusions, for example: two or more consecutive calls between the user and a particular number, held at a time difference smaller than a predetermined value, for example 30 seconds, can suggest a low quality of service the user is experiencing, thus the first session was ended and one of the users started a anew session, even if the first one was not dropped.. A similar conclusion can be drawn if a user is having a communication which is significantly shorter than his usual communications in general, or with a certain number, in particular. The occurrence of such types of problems to multiple users in the same local, may indicate a regional network problem. Alternatively, If no signs for network failure or another problem exist, whether for the network of the user or the network associated with the other user or device, then the situation may suggest failure of the user's device or the other person's device. If the same scenario repeats for the user in multiple occasions in multiple geographic locations, it may be the case that the user's device is faulty, in which case a suggestion can be made to replace or fix the device. If the same scenario repeats in the same geographic area, then the conclusion may be that the user is calling from a low-coverage area, in which case measures may be taken to improve coverage.

Throughout the disclosure, the terms user and device may be used interchangeably, since the subject matter relates to a user who uses a device to communicate with another user who is also using a device, for example when holding a conversation, or a user sending a message or another content to a device with no intervention by the other user. In both cases, the analysis relates to the communication exchanged between the devices or between a device and infrastructure equipment.

Referring now to Fig. 1, showing a schematic illustration of a typical environment, in which the disclosed method and apparatus is used. The environment comprises a user 100 speaking or otherwise communicating, for example sending or receiving a short message (SMS), or surfing the Internet via a mobile device 104. Mobile device 104 may be a portable phone, a personal digital assistant with communication capabilities, a landline phone or any other communication device. Device 104 is in communication with a server 112, which may be operated by an operator. The communication is optionally carried over cellular equipment, including for example cell or antenna 108. Alternatively, the communication is carried over any other channel, such as a landline. Server 1 12 is generally in communication with server 1 16 associated with another operator. Alternatively, if the communication is within the same network, then server 116 can be omitted. Server 1 16 (or server 1 12) sends and receives information to and from cell or antenna 120, which further transfers the information to and from device 124, optionally used by user 122. The communication exchanged between user 100 or device 104 and user 122 or device 124 may include cellular conversations, text messages, images, video streams, web pages, WAP data, or any other data associated with at ;least one end user or end-device. It will further be appreciated that another optional environment includes only one user or one device, such as user 100 or device 104, which communicate with a server rather than with another end user or device, for example when downloading WAP pages.

It will be appreciated that Fig. 1 is exemplary only, and multiple other configurations may exist in which the disclosure may be used, including environments employing mobile or landline communication, one or more servers, one, two or more operators, or the like. The essential parts are one or more end devices and one or more servers communicating or transferring communication with the device. Referring now to Figs. 2A and 2B, showing flowcharts of the main steps in a method according to the disclosure. Referring now to Fig. 2A, showing the main steps in constructing user communication models. The method, generally referenced 200, starts on step 204 in which training call detail records (CDRs) are received. The call detail records are preferably collected over a period of time which enables tracking a user's communication patterns, typically between a few hours and a few weeks. Each CDR preferably comprises the number or numbers of the device the user is using, the number of the other device, if any, the operator associated with the number, the areas at which the user is present when making the communication, the time, date, day of week and duration of call, or any other details. The data may also relate to services other than conversations, for example data related to data transfer, such as text, pictures, multi-media or the like. The data may include re-transmission of packets, entries to servers and their frequency in the time line. The CDRs are optionally collected for multiple users. Optionally, an operator may choose to collect data related to users consuming a lot of communication services, to VIP users, or to any other group of users.

On step 208 a user communication model is generated for a particular user from the data related to the user as collected on step 204. The model represents the normal or expected communication patterns for each service the user is consuming in each network, over a predetermined period of time. The model can comprise the average and standard deviation of the duration of a user's conversations, in general or with particular often-called numbers, frequency of the same, retransmissions of data packets, or the like.

The model can take any form, such as a statistical model indicating a probability for certain data, such as communication duration, a location, numbers communicated with, or other parameters. Alternatively, the user communication model can be a case-based model, such as indicating that a very short call to or from a number with whom the user usually holds longs conversations indicates a disconnection. Alternatively, combinations of such models, or other types of models can be used. The model is optionally generated using artificial intelligence techniques such as clustering, data mining, cause analysis, or others. Alternatively the model can be based on use-cases and generated or refined, by a human, based on experience, expertise and knowledge about problems and indications thereof.

On optional step 212 a multiple-user model is generated, comprising information relevant for groups of users. The multiple user model can also be a statistical or case-based model, and can be constructed using the same techniques as the user model generated on step 208. Optionally, a user model may be enriched with multi-user information, so that a single model is applicable for a particular user, wherein the model also comprises information related to a group of users.

On optional step 216, models comprising cases or rules are generated, either by a human or by a dedicated system. Optionally, the case-based models are integrated with the user models generated on step 208 or with multiple-user models generated on step 212. The additional cases can relate to normal and expected behaviors, as well as to abnormal behaviors of all users, and optionally their implications or improvement recommendation to be made. For example, a rule can be added which indicates that in case a very short call, or two calls during a short time interval, such as 30 seconds, were held between the user and a particular number, regardless of the caller, the length or the termination reason, then the following cases can be distinguished: 1. If one of these two cases occurred for at least a predetermined number of users, in multiple areas, who called numbers associated with multiple other operators, then a failure in the network occurred. 2. If any of these cases occurred for multiple users in a particular area, then a problem in a cell, antenna, or another local equipment has probably occurred. 3. If any of these cases occurred for multiple users who called only numbers associated with a particular operator, then a failure with the other operator's equipment occurred. 4. If the situation occurred multiple times for a particular user when calling multiple numbers, a problem may have occurred with the user's device. 5. If the situation occurred multiple times for a particular user when calling a particular number, a problem may have occurred with the other device. It will be appreciated that the scenarios detailed above are exemplary only, and that additional scenarios and conclusions can be identified and integrated into a model. It will be further appreciated that any one or two of steps 208, 212 or 216 can be omitted.

It will be appreciated an initial model can be constructed upon data collected during a short period of time, such as few hours. The model can be used as detailed in association with Fig. 2B below, while additional data is collected, so that the model is continuously or periodically enriched or updated, to provide a more indicative representation of the user's behavior.

Referring now to Fig. 2B, showing the main steps in using the models constructed on Fig., 2A in order to detect, analyze and suggest improvements for a user's communication. The method, generally referenced 220, starts on step 224 in which run-time CDRs are received for one or more users. The method can be performed in real-time or near-real-time, i.e. analyze users' communication as they occur or shortly after, or offline, i.e., every predetermined period of time, such as one or a few hours, a day, a week, or the like. In yet another alternative, the method can be activated upon demand, i.e. when a user complains or when there is another reason to investigate the communication made by a user or a group of users. Optionally, an operator may choose to activate the method for selected users, such as users consuming a lot of communication services relatively to other users, VIP users, or any other group of users.

The CDRs preferably comprise the number or numbers the user is calling from, the numbers most often called, the operator associated with these numbers, the areas at which the user is present when making calls, the time, date, day of week and duration of calls, or any other details. The CDRs may also contain data related to other services, as detailed in association with step 204 above.

On step 228, the collected data is analyzed against one or more user models, as constructed in step 208 of Fig. 2 A above. The analysis depends on the type of model used. For example, if the model is a statistical model, the probability of one or more CDRs is assessed, and if the probability is smaller than a predetermined value, an abnormal behavior is indicated. If the model is case- based, then it is assessed, for example using pattern recognition techniques, whether the behavior complies with any of the cases. The analysis produces an assessment to one or more equality parameters related to the communication. The quality parameters may relate to voice, video or data transmission, and to the network associated with the user or to another network.

On optional step 232, CDRs collected for multiple users are analyzed against the multiple user model generated on step 212, in order to identify abnormal behaviors associated with a group of users. If a user model also comprises multi-user information, or if there is no multi-user information, then step 232 can be omitted.

On optional step 236, the collected CDRs are analyzed against the case- based models generated on step 216. If there are no case-based models, or the cases information is integrated into a user model or a multi-user model, and step 236 can be omitted.

The results of analysis steps 228, 232, or 236 include quality assessment of parameters associated with audio, video or data transfers. The assessment relates to the quality of conversations or data within the network, wherein the assessment may optionally use also available technical data. However, the assessment may also relate to the quality of services such as conversations or data transfer provided by other operators and networks, for which no additional data is available, and the only source of information is the user's behavior and pre- generated models.

On step 238, the analysis results are optionally enhanced with technical data, such as equipment status or other details, if available.

On step 240, one or more improvement recommendations, directed to the user or to the operator are optionally made. A recommendation to the user may be to upgrade his or her device. A recommendations addressed to the operator may be to improve a cell, antenna or a mechanism for switching, or otherwise upgrading or fixing equipment. .

On step 244 the analysis results or the recommendations are output, and sent to the user, to the operator, or otherwise used. For example, the operator can assess the quality of the network relatively to other operators, based on the relative number of disconnects caused by each network, a geographical or temporal analysis of the same, quality of other services, or the like, without access to other operator's data.

Referring now to Fig. 3, showing a block diagram of the main components in an apparatus for assessing the quality of service provided by an operator.

The components of the apparatus detailed below are preferably interrelated collections of computer instructions, such as software modules, executables, libraries, routines or other units implemented in any programming language and under any programming environment. The apparatus is preferably executed by a computing device associated with the operator. The computing device can be any computing platform provisioned with a memory device, a CPU or microprocessor device, and one or more I/O ports. The computing device can thus be a general purpose processor, or alternatively a firmware ported for a specific processor such as digital signal processor (DSP) or microcontrollers, or hardware or configurable hardware such as field programmable gate array (FPGA) or application specific integrated circuit (ASIC).

The apparatus comprises CDR collection component 300 which collects CDR data related to end-users of a network-based service. The data relates to conversations or data transfer within the network, or with users or devices of networks associated with other operators. CDR collection component 300 is useful both during training and in run-time.

Component 304 receives the CDRs from CDR collection component 300, or from another source, such as an external source, a storage device or the like. CDR receiving component 304 is also useful both during training and in run-time. The data received by CDR receiving component 304 during training of the system is transferred to model generation components 308, which comprise a single user model generation component 312 for generating a model relating to the behavior of a particular user, multi-user model generation component 316 for generating a model relating to the behavior of a group of users, and case model generation component 320 for generating a case-based model which may relate to one user, multiple users, or all users. Case model generation component 320 can be designed to deduce rules from training CDRs or to receive rules provided by a human. It will thus be appreciated that any one or two of components 312 and 316 are optional, and that as long as any of components 312, 316 and 320 exist, the system is functional.

It will be appreciated that model generation components 308 can be implemented as one or more components, and that the division to components 312, 316 and 320 represents only an exemplary implementation. It will be further appreciated that the components may generate a single model comprising all relevant information per a particular user, the model including data related to the particular user as well as data common to multiple users or cases.

The data received by CDR receiving component 304 during run-time is transferred to analysis and recommendations components 324. Components 324 comprise component 328 for analyzing data against a model. Component 328 is functional in analyzing data related to a particular user or to multiple users, against a single-user model or against a multiple user model. Components 324 further comprise analysis against cases component 332, for analyzing user or user-group behavior against a case-based model. It will be appreciated that components 328 and 332 can be implemented as a single component analyzing user or user-group behavior against all available behavioral data and models.

Once analysis is performed, analysis results are available, which include assessment of quality of service related to conversations or data transfer, within a network and with other networks. The analysis results are transferred to recommendations component 336, which issues recommendations for a particular user, recommendations for the operator related to a particular area, particular devices, infrastructure, or the like. Recommendations can also be generated which relate to other networks. Output component 340 is responsible for outputting the analysis results and recommendations. The results and recommendations can be sent to a particular user, to a person within the operator, stored, published or otherwise used. The results and recommendations can be sent be sent or made available to a person or system using any known technique, such as e-mail, text message, voice message, report, electronic billboard, or the like.

The apparatus further comprises or has access to storage unit or device 344 for storing the models. The storage unit can further store the training CDRs, the run-time CDRs, or the analysis results and recommendations, as well as any other related information. Storage device 344 can be a mass storage device, for example an optical storage device such as a CD, a DVD, or a laser disk; a magnetic storage device such as a tape or a hard disk; a semiconductor storage device such as Flash device, memory stick, or the like. The disclosed method and apparatus provide for analyzing communication of a particular user or user-group from call detail records, and assessing quality of service of an operator or network the user is associated with, or other networks or operators providing services to users or devices with whom the particular user communicates. The analysis is based on analyzing the call detail records of the user or user-group against predetermined models of the user behaviors, which may also include general behaviors or specific cases. The method and apparatus either do not use, or use as an enhancement technical data related to the operator, thus basing the analysis on the user's behavior as reflected by the CDRs. It will be appreciated that the disclosure relates to any communication operator, including cellular, mobile, landline, internet service provider, or others.

It will be appreciated that the disclosed method and apparatus can be implemented in multiple ways and use multiple characteristics of the user data as well as rules or cases, without deviating from the teachings of the disclosure. While the disclosure has been described with reference to exemplary embodiments, it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted for elements thereof without departing from the scope of the disclosure. In addition, many modifications may be made to adapt a particular situation, material, step of component to the teachings without departing from the essential scope thereof. Therefore, it is intended that the disclosed subject matter not be limited to the particular embodiment disclosed as the best mode contemplated for carrying out this invention, but only by the claims that follow.

Claims

1. A method for assessing a quality parameter of a first network or a second network, the first network providing a communication service to an at least one device, the device used by a user, the method comprising: receiving at least one call detail record of communication sessions associated with the device; and analyzing the at least one call detail record against a communication model, and assessing the quality parameter.

2. The method of claim 1 wherein the communication model comprises an at least one rule.

3. The method of claim 2 wherein the at least one rule is provided by a human.

4. The method of claim 2 wherein the at least one rule indicates that two communication sessions held during a predetermined time interval between two users indicate a network failure.

5. The method of claim 4 wherein at least one of the at least two communication sessions was terminated successfully.

6. The method of claim 1 wherein the quality parameter relates to the quality of voice or data transmission within the first network.

7. The method of claim 1 wherein the device is communicating with a user or a device in the second network, and wherein the quality parameter relates to the quality of voice or data transmission within the second network.

8. The method of claim 1 further comprising the step of generating a user or multi-user communication model based on training call detail records of the device, wherein the communication model comprises a user or multi-user communication model.

9. The method of claim 1 further comprising the step of generating an at least one recommendation for improving the quality parameter.

10. The method of claim 1 further comprising a step of enhancing the assessment of the quality parameter using technical call data.

11. An apparatus for assessing a quality parameter of a first network or a second network, the first network providing a communication service to a device, the device used by a user, the apparatus comprising: a call detail record receiving component for receiving at least one call detail record of a communication session associated with the device; and an analysis component for analyzing the at least one call detail record against an at least one communication model and assessing the quality parameter.

12. The apparatus of claim 11 wherein the quality parameter relates to the quality of voice or data transmission within the first network.

13. The apparatus of claim 1 1 wherein the device is communicating with a user or a device in the second network, and wherein the quality parameter relates to the quality of voice or data transmission within the second network.

14. The apparatus of claim 11 further comprising a component for generating the communication model based on training call detail records of the device.

15. The apparatus of claim 1 1 further comprising a component for generating the communication model based on an at least one rule.

16. The apparatus of claim 15 wherein the at least one rule indicates that two communication sessions held during a predetermined time interval between two users indicate a network failure.

17. The apparatus of claim 1 1 wherein the communication model is a user communication model or a multi-user communication model.

18. The apparatus of claim 1 1 further comprising a recommendations component for generating an at least one recommendation for improving the quality parameter.

19. A computer readable storage medium containing a set of instructions for a general purpose computer, the set of instructions comprising: receiving at least one call detail record of communication sessions associated with a communication device; and analyzing the at least one call detail record against an at least one communication model, and assessing a quality parameter of a network associated with the device.