География и Геология/5.Картография и Геоинформатика

 

Golovan K.R., D.Sc. Pogorelov A.V.

Kuban State University

About the territorial structure of the traffic in the mobile network of second generation

 

Statement of the problem. A study of the territorial structure of cellular communication, despite the rapid development and growing importance in modern Russian economy is still having episodic nature. Available territorial generalizations have federal and regional scale [1, 3]; level of a large city – the main consumer of services, related to the particular characteristics of the properties of its territorial organization, hardly been studied.

Problem of determining the spatial distribution of traffic in cellular networks [7, 8] seems very urgent. This information can be used for a wide range of secondary tasks: selecting a location for the base station (BS) [4, 5] or office, advertising, design tariffs, etc.

Statistics by the number of voice traffic gathering in within a cell – a fundamental element in the cellular network [6]. With this organization of the data, it is possible to make only a superficial understanding of the spatial distribution of traffic for three reasons:

1. Low accuracy of the cell boundaries.

2. Cells overlap to each other.

3. Large cell size (length from hundreds of meters in the city to tens of kilometers beyond), hence the lack of spatial resolution for traffic estimates.

The main objective of this study – the development of methodology for estimating the spatial distribution of voice traffic in 2G network with high detail, i.e. displaying traffic structure within the cell.

The original data. As the initial data we used statistics of one of the mobile operators for September 18, 2013 (Wednesday) in Orenburg. City was selected due to the possibility to produce expert verification of the results obtained during the research by specialist with years of experience in the field of planning a cellular network in Orenburg.

In work we used the following statistics:

1. Total amount of traffic, measured in Erlang.

2. Compensation of propagation time from the phone to the BS Timing Advance (TA); based on this parameter it is possible to get a distance from the BS to the subscriber. [2]

Algorithm of calculation. Each value of TA corresponds to percentage of the total traffic. With known cell's total number of the traffic should be distributed according to the percentage threshold value of the parameter TA, whereby we obtain the amount of traffic for each range of TA (Table 1).

Table 1

The source data after the first stage of processing

 

Cell 1

Range of TA

1

2

3

4

5

6

7

Traffic, Erlang

90,7

3,9

1,4

1,7

0,3

0,3

0,3

Cell 2

Range of TA

1

2

3

4

5

6

8

Traffic, Erlang

41,9

27

1,8

1

0,3

1,7

9,6

 

Radiowave propagation speed equal to the speed of light, so knowing TA, we calculate the distance traveled by a signal from the BS to the subscriber by the formula 1:

                                         (1)

Thus, for each cell, we can calculate the zones 547 m wide and knowing the amount of traffic in each zone we obtain the scheme of subscriber activity in the first approximation (Fig. 1).

 

Fig. 1. The scheme of subscriber activity in a first approximation

 

This scheme may be informative in respect of a single cell, while there are thousands cells in a city so it should be better to present traffic in the form of separate points, and then analyze the density. For this purpose, we created script, using the Python programming language, allowing randomly distribute traffic to a point on the arch which is spaced from the BS by a distance in accordance with the TA for the azimuth direction of the antenna. As input data the script receives a text file with the following information:

1. Coordinates of antenna, to calculate the distances between "traffic point" and transmitter in accordance to the TA.

2. Azimuth of antenna, which determines the direction of the cell; for example, in Figure 1 the antenna’s azimuth is 180°.

3. TA and the amount of traffic that determine how much and at what distance points will be generated.

Coordinates of each point of traffic calculated by the formula 2:

,         (2)

where x, y –antenna’s coordinates, A - azimuth, rnd - random number in the range from -60 to 60.

Value of rnd corresponds to 120° antenna’s sector disclosure. In reality, the signal propagation is not restricted and may exceed 120°. However, such cases are unlikely and can be neglected, especially in cities which have high-density of cells.

At the conclusion of the script we obtain a text file with the coordinates of “points of traffic”, next step we render it in the program ArcMap (Esri, USA) using the command “Display XY data” (Fig. 2).

 

 

Fig. 2. Visualize the results of the script

 

For visualization of the density in a raster format the obtained points are handled by the module «Spatial Analyst Tools», namely of the tool «Kernel Density», which calculates the density of point features around each cell of the output raster. A key parameter in the construction of the image is the radius of the search; changing this parameter sets the detalization of the simulated field. In this study, for temporary permission day, in our opinion, it is appropriate to select 5-8 main centers of traffic. Optimal result obtained with the required detail when searching points in a radius of 500 m.

At the last stage of processing, we executed raster-vector conversion and exporting images to Google Earth (Google, USA) for analyze and highlight areas with high traffic density, in other words, an integrated daily activity of subscribers.

Analysis of the data. In analyzing the resulting raster image highlighted 6 centers of traffic (Fig. 3).

 

 

Fig. 3. Density distribution with numbered centers of increased activity of voice traffic

 

The largest spot (№ 1), located in the center of Orenburg and restricted by streets Rybakovskaya , Proletarskaya , Krasnoarmeiskaya , Kazakovskaya. This area is a number of key administrative buildings, including government buildings of Orenburg region, the Ministry of Finance, Department of the Federal Penitentiary Service, ATC Orenburg municipal administration. Center number 2 is located in the northern part of the city, restricted by streets Volgogradskaya, Prostornaya, Dzhangildina, and Brestskaya. This center , probably formed due to the presence of a large number of educational institutions: kindergartens (№ № 102, 151, 161 , 123, 5 , 138, 107 , 118, 175 , 114, 144 , 148, 141 ), schools (№ № 67 , 8, 19 , 57 , 4, 69, 5 , 71).

Center of increased traffic density (№ 3) located inside the streets Samoletnaya, Yaltiskaya, Raskovoyi, Orskaya and matches the boundaries of a large residential neighborhood. Center number 4 limited Zhukov Street, Turkestanskaya, Leningradskaya, Chkalov Street. Here is a residential area and a number of educational institutions. Center number 5 is localized near the stadium "Olympic" between streets Shevchenko, Proletarskaya, New Street and Tereshkova Street. Center number 6, located between the streets Gagarin Street, Mednogorskaya, Buguruslanskaya corresponds to residential area.

Thus, calculations are performed and constructed maps of the distribution of voice traffic in a large city, reflecting the spatial inhomogeneity of the subscribers’ activity in a cellular network. Level of simulation - "intra-cell" with a temporal resolution day. We assume that the proposed method of assessing the territorial structure and display of integrated indicators of voice traffic within a cellular network of a large city is quite effective, and the results of the localization of traffic in Orenburg, according to expert opinion, explained by nature of the distribution of elements in urban infrastructure (residential and administrative buildings, etc.).

Literature:

 

1.        Головань К.Р., Погорелов А.В. География сотовой связи в Южном федеральном округе (на примере МТС) // Известия Кубанского государственного университета. Естественные науки. 2014. Вып. 3.

2.        Дворкина Н. Б., Намиот Д. Е. Использование opencellid api в м. Вып. обильных сервисах // Прикладная информатика. 2010.  №5.  С.92-101.

3.        Леснова Ю. В. География развития сотовой связи России: диссертация … кандидата географических наук: 25.00.24 / Ю. В. Леснова; [Место защиты: Моск. гос. ун-т им М. В. Ломоносова]. М., 2004. 209 с.

4.        Мухаджинов Р. Р. О постановке задачи выбора рационального размещения базовых станций сотовой связи // Вестник АГТУ. 2008.  №1.  С.127-129.

5.        Мухаджинов Р. Р. Применение генетического алгоритма к решению задачи «размещение станций систем мобильной связи» // Вестник АГТУ. Серия: Управление, вычислительная техника и информатика. 2009.  №1.  С.165-167.

6.        Ратынский М. В. Основы сотовой связи. / Под ред.  Д. Б. Зимина. –2-е изд., перераб. и доп. М.: Радио и связь, 2000. 248 с.

7.        Руфова А. В. Частотно-территориальное планирование сетей подвижной связи: учеб. пособие. / Под ред. В. Ю. Бабкова // СПбГУТ. СПб, 2002. 64 с.

8.        Tutschku K., Tran-Gia P. Spatial traffic estimation and characterization for mobile communication network design. University of Wurzburg. 1998.