Customer Tracking Systems based on Identifiers of Mobile Phones

D. M. Mikhaylov

, A. V. Zuykov

, S. M. Kharkov

, S. V. Ponomarev

, S. V. Dvoryankin

and A. M. Tolstaya

National Research Nuclear University MEPhI (Moscow Engineering Physics Institute),

Kashirskoe highway 31, 115409, Moscow, Russia

Concern Radio-Electronic Technology, Ltd, Moscow, Russia

Keywords: Non-Functioning Communication Cell, Counting Users by Mobile Phone Ids, Collection of Visit Statistics.

Abstract: Gathering statistics about visitors finds more and more applications in various fields of business and

commerce. This paper describes the system of impersonal counting of unique visitors by their mobile

identifiers. Counting is carried out using a non-functioning communication cell (the system does not provide

communication services to users of mobile networks). The system masks itself as the base station of a

mobile operator. Mobile devices automatically connect to the system even in case of a strong signal from

the towers of mobile operators. Once connected, the user identification data is received. The proposed

solution allows to compare data about the number of visits to a particular site in various periods of time and

to identify the re-occurrence of the visitors. The system is inexpensive and shows 99% accuracy in the

identification of users (compared to the real data about the visitors).

1 INTRODUCTION

It is well known in the industry of online shopping,

that collecting statistical data with further data

mining procedures can help in marketing of

products. In online shopping business it is quite easy

for a programmer with certain skills to store

information about pages that users of the shop

visited, how long did they stay on these pages, what

they actually purchased, etc. All these data are used

in analytics, the main task of which is to transform

the page, site navigation and functionality to

increase the sales.

At the same time, much more products are still

sold "offline" – in shops, malls, hypermarkets and

other places where data collecting is limited. In

digital form, the shop usually receives only

information about what was sold, how much goods

was sold on a specific day at a specific time, etc.

Therefore, the collection of statistics about

visitors finds more and more applications in various

fields of business. Many companies seek to obtain

reliable statistics on the number of visitors, duration

of eye contact with the product or advertising signs

and other indicators. These statistics can be used, for

example, to increase the efficiency of an advertising

campaign or to encourage the consumption of a

certain type of goods or services (Positioning 2015,

Mikhaylov 2014a).

Many scientific papers focus on the issue of

tracking visitors (Cai 2013, Mikhaylov 2014b,

Samoylov 2015, Wang 2014, Wang 2013).

However, a significant number of systems is aimed

at counting the number of people staying in the

room. (Types of counting systems 2015) There are

also powerful systems for analyzing the faces of

visitors, but the processing is carried out on a remote

server, or requires the use of a separate powerful

server (Audience Engagement 2015, Erns 2009).

For example, there is a well-known system

RetailNext (2015), which uses sensor technology

(cameras and radio frequency identification) for

counting visitors and traffic. DVR CountBOX

(2015) is designed to count visitors; it includes a

calculation STB-1 unit and a video camera. Counter

MC-Ethernet (2015) combines traffic to two

different channels – considers entrance and exit of

the visitors separately, which allows watching the

dynamics of entry and exiting traffic of the site.

Another system – Statistics Dome (2015) – is a

professional tool for counting visitors through

surveillance algorithms of Orwell 2k computer

vision. Video counter Statistics Dome effects a bi-

directional people counting with the help of built-in

video analytics.

Mikhaylov, D., Zuykov, A., Kharkov, S., Ponomarev, S., Dvoryankin, S. and Tolstaya, A..

Customer Tracking Systems based on Identiﬁers of Mobile Phones.

In Proceedings of the 7th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2015) - Volume 1: KDIR, pages 451-455

ISBN: 978-989-758-158-8

451

The disadvantage of such systems is the

omission of such problems as visiting of malls by

family customers, as well as a constant entry and

exit of buyers from the room several times. They are

also expensive and often inaccurate.

Therefore, it is proposed to use the system for

counting the identifiers of mobile phones to improve

the efficiency of the resulting statistics on the

number of unique visitors, for example, in the mall.

The objective of the system is to eliminate the above

flaws with the new system giving improved

performance.

2 COUNTING VISITORS BY

IDENTIFIERS OF MOBILE

PHONES

The proposed solution is based on non-functioning

communication cell. A non-functioning communi-

cation cell (NFCC) is an independent cell of the

cellular network, which generates a signal of a

mobile operator's base station, but it does not

provide access to communication services. NFCC

are successfully used for many civilian and military

applications.

At the same time, the real operator is an

officially registered service provider for the selected

user.

Customer tracking systems by identifiers of

mobile phones include:

 non-functioning connection cell for receiving

data; wherein NFCC is associated with the

central server;

 central server that analyzes received traffic,

wherein the server is connected to the central

computer of the operator;

 computer of the operator, where there are reports

of new events in the system arrive.

Figure 1 shows the structure of the system under

development. The number of NFCC can vary from 1

to N.

Inside the central server, to which the NFCC are

connected, an analysis of the data is carried out, but

the interim results of processing are not saved in the

memory of the equipment and the external memory.

The system is a non-functioning communication

cell, which is determined by the network as a base

station of a cellular operator, whereby the mobile

devices within the range of the complex will

automatically switch to work with it. Once

connected, there is a receipt identifying the user data

that are needed for the further calculation.

Figure 1: This caption has more than one line so it has to

be set to justify.

Further, the mobile device resumes functioning with

the network of its real operator. The central server,

when analyzing, can also identify the model of the

mobile device and the operator of the mobile device.

NFCC effects the switch of the mobile device to

collect identifying information even in case of a

good signal from the base station of the real operator

due to the high base station selection criteria (cell

selection and cell reselection criteria).

The described device does not allow any kind of

data transfer that is why the mobile devices

connected to the complex switch back to the

network of the real operator. This scheme of work

does not affect the quality of communication

services of the real service provider, so the complex

does not prevent the data transmission.

The system operates as follows: NFCC conducts

partial authentication, and after receiving the user

data for identifying the mobile device, switches back

to the base station of the operator. The system

allows keeping a record of the duration of a visit to

the selected area with up to six minutes accuracy. In

the case of inactivity, every six minutes a mobile

device of the subscriber updates its connection

status. Thus, it is possible to build the statistics

length of stay in the territory of the subscriber.

Automatic connection of mobile devices to the

NFCC within one standard connection occurs even

in the case of a strong signal from the towers of

mobile operators.

Figure 2 shows a schematic block diagram,

which can be realized by the proposed system of

counting visitors by identifiers of mobile phones.

Figure 2 shows:

1 – mobile device of the subscriber whose

identifiers are to be determined;

KDIR 2015 - 7th International Conference on Knowledge Discovery and Information Retrieval

452

2 – base station of an operator to which the

subscriber's mobile device is connected;

3 ... N – NFCC working by standards 2G\3G\4G,

which switches the mobile device of the user (due to

better reception). It is necessary to create conditions

for switching the mobile device to the new base

station, thereby the mobile device performs a new

authentication in the network, providing the ability

to process the required IDs;

N + 1 – central server that handles the received

identifiers;

N + 2 – computer operator who receives the

anonymous analysis results.

Mobile device

Non-functioning

communication

cell

Non-functioning

communication

cell

Base station of

the operator

Central server

Computer of the

operator

...

N+1

N+2

Figure 2: Schematic block diagram of system

implementation.

A customer tracking systems by identifiers of

mobile phones works as follows:

 Step 1 – In normal mode, the mobile device of

the subscriber is connected to the access point.

 Step 2 – At a pre-set time the mobile device of

the subscriber changes the connection from the

base station of the operator to one of the N

virtual base stations, thereby performing

authentication of the new network, which

provides the possibility of receiving the desired

identifiers. This is achieved by using a special

radioreceiver of signals during a new

authentication of the mobile device in the

network when switching from one base

station to

another, due to forced creation of conditions for

the best quality signal.

 Step 3 – Central server that is connected to N

non-functioning communication cells, processes

the information obtained.

 Step 4 – Unallocated analysis results from a

central server are transferred to the computer of

the operator.

 Step 5 – Once the correct identities of mobile

devices of subscribers are gathered, they switch

the connection from one of NFCC to the base

station of the real operator.

Thus, the system of counting the identifiers of

mobile phones allows determining the number of

unique visitors, without affecting the quality of

cellular connection. The system collects anonymous

data and does not receive any personally identifiable

information of the visitors.

In terms of information security the proposed

system can:

 redirect packet traffic in the data leak prevention

(DLP) system;

 record all calls with copies of conversations sent

to DLP system;

 control of SMS messages;

 control of employees presence on the workplace;

 selective blocking of subscribers;

 mass dispatch of SMS notifications.

It is possible to develop the client for the mobile

operating systems Android, iOS, Windows Phone,

with integrated DLP system that will allow to block

camera, microphone and sending images.

There are several spheres of the system

application, for example:

 marketing research, integration with loyalty

programs;

 communication systems;

 protection systems from data leakage (DLP

systems);

 intelligent connection blocking with no threat to

health;

 security systems, physical security of facilities.

The latter has the following advantages:

 large coverage area, it is advantageous to use for

the control of remote objects and sparsely

populated areas (pipelines, electrohighways,

border, etc.);

 possibility of drawing up white lists, operation in

mode “friend/foe”;

 ability to identify the offender (in case of special

permissions).

3 EXPERIMENTAL DATA

To test the effectiveness of the developed system for

counting mobile phone identifiers a special territory

Customer Tracking Systems based on Identiﬁers of Mobile Phones

453

was allocated. It was a bandwidth monitoring area of

200 m2 and 20 employees. Control of access to the

territory was carried out by using access cards,

allowing us to conduct a comparative analysis of the

system.

As a part of the experiment, which was

conducted during five weeks, antennas and amplifier

were installed around the perimeter of the zone so as

to completely cover-control it. For ease of testing,

the system was set up for one mobile operator (the

employees who worked in the territory, used the

services of the same operator). The results of the

experiment are given in Table 1.

Table 1: Results of the experiment of counting visitors.

Timeline NFCC inspection

results

(entrance to the

facility)

Access cards

inspection results

(entrance to the

facility)

1 week 586 578

2 weeks 565 563

3 weeks 586 586

4 weeks 574 569

5 weeks 592 589

During the entire time of the experiment, the

system shows stable operation. The results obtained

with the help of NFCC were compared with the data,

fixed by means of electronic access cards to the

territory. The results do not differ by more than 1%.

Inaccuracy is due to the fact that some employees

had a tablet with access to the Internet or a second

cell phone while being at the controlled territory.

4 CONCLUSIONS

Thus, this paper describes a system of impersonal

counting of identifiers of mobile phones, which

would greatly increase the efficiency of counting

unique visitors in shopping malls, exhibitions and

other similar events. The system requires a small

financial cost for its installation and operation, it

does not affects the quality of services provided to

users by the operator of communication and has no

adverse impact on the information security of mobile

devices.

Ongoing testing of the prototype system

demonstrated its high efficiency, where the accuracy

of the results obtained by the device is 99%

compared to the actual data. In the future, this figure

will be increased to 100%.

The limitations of the study are the following:

 legal difficulties as there is no regulation of use

of such devices, albeit low-power, at frequencies

of mobile operators;

 the system only supports the 2G (GSM), other

cellular standards are not supported. It is

necessary to develop a system or jammers for

them. But the latter encounter legal difficulties.

Today, the system is being tested and further

developed. It is planned to conduct scientific

research to address the problem of the calculation of

several unique devices from one owner (for

example, the visitor has two mobile phones and a

tablet).

The scalability of the system described will also

be examined.

REFERENCES

Positioning (2015) Marketing tools and sales. Official site

of Development Program of the Small Business in

Saint-Petersburg. Committee for economic

development, industrial policy and trade of

Administration of St. Petersburg. URL:

http://www.spb-mb.ru/index.php?page=286.

Dmitry Mikhaylov, Timur Khabibullin, Andrey

Starikovskiy, Alexander Smirnov, Mikhail Froimson,

Maxim Aristov and Vladimir Konev (2014a) Method

and System for Gender-oriented Targeted Advertising.

ICORES 2014. Proceedings of the 3rd International

Conference on Operations Research and Enterprise

Systems. ESEO, Angers, Loire Valley, France. 6 - 8

March, 2014. Pages 153-158.

Zhaowei Cai; Longyin Wen; Dong Cao; Zhen Lei; Dong

Yi; Li, S.Z. (2013) Person-specific face tracking with

online recognition. 10th IEEE International

Conference and Workshops on Automatic Face and

Gesture Recognition (FG), 2013. Pages: 1 – 6.

Dmitry Mikhaylov, Anton Samoylov, Peter Minin and

Alexey Egorov. (2014b) Face Detection and Tracking

from Image and Statistics Gathering. Tenth

International Conference on Signal-Image Technology

& Internet-Based Systems, 2014. С 37-42.

Samoylov A.S., Mikhaylov D.M., Minin P.E., Egorov

A.D. (2015) Face Tracking for a System of Collecting

Statistics on Visitors and Quality Assessment of its

Functioning. Journal of Theoretical and Applied

Information Technology. Volume 71, Issue 3, 31

January 2015, Pages 353-366.

Guojun Wang; Alam Bhuiyan, M.Z.; Jiannong Cao; Jie

Wu. (2014) Detecting Movements of a Target Using

Face Tracking in Wireless Sensor Networks. IEEE

Transactions on Parallel and Distributed Systems,

(Volume: 25, Issue: 4). April 2014. Pages: 939 – 949.

Chao Wang; Yunhong Wang; Zhaoxiang Zhang; Yiding

Wang. (2013) Face Tracking and Recognition via

Incremental Local Sparse Representation. Seventh

KDIR 2015 - 7th International Conference on Knowledge Discovery and Information Retrieval

454

International Conference on Image and Graphics

(ICIG), 2013. Pages: 493 – 498.

Types of counting systems. (2015) Unique electronic users

accounting system. Counters of visitors. URL:

http://www.uchet.biz/vidy-sistem-podscheta-

posetitelei.php.

Audience Engagement. (2015) Intel AIM Suite. URL:

https://aimsuite.intel.com/aim-suite-work/audience-

engagement.

Ernst A., Ruf T. Kublbeck C. (2009) A modular

framework to detect and analyze faces for audience

measurement systems. Proc. GI Jahrestagung, 2009.

Pages: 3941-3953.

RetailNext (2015) Official website. URL:

http://retailnext.net.

CountBOX (2015) Official website. URL: http://count-

box.com.

The system of visitors counting MC-Ethernet (2015) "The

first BIT". URL: http://www.kkm.ru/shop/anti-

steal/sistemy-podscheta-posetiteley/sistema-

podscheta-posetiteley-megacount/44055.

Statistics Dome (2015) Official website. URL:

http://подсчет-посетителей.рф.

Customer Tracking Systems based on Identiﬁers of Mobile Phones

455