Directions Magazine
Hello. Login | Register

Articles

All Articles | Post Comment

GIS and GSM Network Quality Monitoring: A Nigerian Case Study

Saturday, November 26th 2005
Comments
Classified Ads:

Summary:

August 2001 was a pivotal date in the history of Nigeria.That was when the first Global System for Mobile (GSM) communications call was made under a democratic government.This article, by GIS Specialist Ireti Ajala of MTN Nigeria Communication Ltd., describes how the company is using GIS to monitor network quality and improve service.

August 2001 was a pivotal date in the history of Nigeria.That was when the first Global System for Mobile (GSM) communications call was made under a democratic government.This event heralded the dawn of a new era "" the era of GSM technology, which has completely changed the face of doing business in Nigeria.

However, four years after the first GSM call was made, the GSM industry in Nigeria has changed a lot.Competition for subscribers is getting fierce.Operators have resorted to "price wars" to win subscribers. Subscribers, on the other hand, have more choices than ever regarding which GSM operator to use.To attract, maintain and move subscribers to high-value services such as voice, network operators must provide an unprecedented quality of service.Providing quality service will require monitoring and quality assurance with a view to optimizing the network.

The network statistics captured in the switch are monitored and processed at the Network Monitoring Center (NMC), using network monitoring tools that are based on the traditional row-column format. However intelligent these tools are; they only provide information on what is happening and very little on where.Field engineers who usually fix problems affecting the network have to depend on their "intuitive wet knowledge" to understand where the subscribers are experiencing low service quality.

This article will describe how MTN Nigeria Communication is using GIS capabilities with a GSM network monitoring tool to find geographic areas where subscribers are suffering low quality, and help fix the problem.The GIS-based tool described also increase the network efficiency by more than 70% and indeed helps make fast and informed technical decisions in resolving these network issues.There may be other, more refined methods to create the dynamic maps described in this article, but this approach is working for us at MTN at this point.

Background
When Nigeria gained her independence in 1960, there were only 18,724 functional telephone lines for an estimated population of 45 million, which was a "teledensity" ratio of 0.04 telephones per 100 people. During the thirty-odd years of military rule, there was very little by way of investment in telecommunications, and other sectors did not fare any better.

According to the International Telecommunication Union, by 1996 Nigeria's teledensity ratio was a mere 0.36.It rose slightly to 0.4 by 1999; according to the Nigeria Communication Commission (NCC). Nigeria's teledensity is a far cry from the African average of 1.67. Even the NCC admits that Nigeria has had a very limited telephone network for many years, and the waiting list is estimated at over 10 million people, who have applied to the incumbent monopoly, NITEL (established in 1985) for services.

However, with the liberalization of the telecommunication industry in 2001, the story changed dramatically.The teledensity ratio had tripled within just one year of GSM operation.By May 2005 Nigeria, with an estimated population of 128,771,988, had more than 9 million GSM subscribers, making the country one of the fastest growing GSM markets in the world.At the moment, there are four GSM operators in Nigeria: MTN (for whom I work), V-Mobile, GloMobile and MTEL.MTN enjoys the greatest patronage, with over 4 million subscribers.It was predicted that between 2003 and 2006, Nigeria's GSM market would be Africa's fastest-growing mobile market, and this prediction had been fulfilled. The competition is getting fiercer by the day as operators have to compete desperately for the same potential subscribers.

Four years after the start of the GSM era in Nigeria, the focus is gradually shifting from providing coverage to providing quality service.The euphoria of owning a phone set is gradually giving way to complaints of dropped calls and congestion.

The operators are fast realizing that they are in a highly competitive environment where subscribers can make or break them.Dissatisfaction by subscribers gives rise to a high rate of subscriber churn and low revenue for the operator.The performance of the network has a direct impact on the revenues.The NCC is bringing pressure to the operators to step up the quality of service offered Nigerians and had even gone a step further to award contracts to private companies to conduct comparative analyses of the quality of service offered by each of the operators.The NCC is further threatening to sanction any operator that fails to pay attention to quality.

It therefore behooves all the operators to ensure that the subscribers enjoy the best of service.The determining factors that produce customer satisfaction are answered by these three questions:
  • Can the call be made (or received)?
  • What will it sound like?
  • Will the call drop?
Statistics and Traffic Measurement Subsystem
This paper will concentrate only on network monitoring and problem detection in an Ericsson equipment based network.Monitoring in an Ericsson GSM Network is done at cell, Base Switching Center (BSC) and Mobile Switching Center (MSC) levels to have both a local and global view of the network.Different events are counted and collected by a subsystem called the Statistics and Traffic measurement Subsystem (STS).In the BSC, these events can be handovers, call setups, dropped calls, allocation of different channels, etc.There are also a number of status counters, reporting the status of equipment within the network, such as the current number of occupied channels.By continuously supervising the results from STS, the operator can obtain a very good overview of the radio network performance, which can help detect problems early.

During a call set-up, several counters are affected.The allocation of a Stand-alone Dedicated Control CHannel (SDCCH) can succeed or fail based on congestion or the SDCCH could later drop due to low signal strength.Each event results in different counters to be stepped.The reason for a handover decision can be normal or it could be caused by conditions like bad quality.All these events are recorded by the STS and can be used for further analysis.

The central part in STS is a measuring database, where all measurements are collected from different blocks in the central processor.The user defines if the data should be transferred to an external system in a binary file format, or if reports should be generated as alphanumeric printouts on a terminal.

STS is implemented in a support processor, which is physically located in the Input / Output Group.The frequency of the collection is determined by the basic recording period parameter, which can be set to 5 or 15 minutes.The database consists of several object types.The object types correspond to different types of equipment, logical units or functions in the BSC.Every object type contains several objects (compare with records) that have a number of counters (compare with record fields).

Standard GSM Monitoring Tools
Some of these monitoring tools have in-built intelligence using scripts to highlight certain rows in different colors to draw attention to cells that had violated the network stipulated KPIs.

Fig 1: Sample GSM Network Monitoring Tool.(Click for larger image.)


A number of statistical plots like bar charts, histograms, pie charts, etc.have been incorporated to give better insights into understanding these parameters.With these charts, it is possible to do a comparative analysis for example, by plotting call drop rate against time (See Fig 2).

Fig 2: Sample statistical plotting of Drop call rate against time.(Click for larger image.)


With such powerful tools, engineers have a better understanding of the network status because they are able to detect abnormalities or spot irregularities in the reporting of the measured counters.As impressive as this feat is, engineers still depend heavily on their "wet knowledge" of the geographic area of the network to answer a very important question "" where? They are usually at a loss when it comes to associating network parameters with geographic places.They typically assign responsibility for a section of the network to a particular engineer.However, a network suffers greatly when the old engineers leave and fresh ones take over.Time and effort are expended on training these new engineers to bring them up to speed, not to mention the possibility of the operator being held "ransom" by the old engineers because of the fear of losing their knowledge.

Modern network monitoring needs a tool that addresses this issue.It should be easy to use and present a summary of the network status without trading off details.The most important question that a monitoring/optimization engineer wants answered is: what is happening where? GIS is the best technology for making faster, informed technical decisions, especially if such decisions are spatial, as in this case.

Applying GIS to GSM Network Monitoring
Using a map in network monitoring can provide a dramatic improvement over traditional optimization methods, allowing the engineers to see a precise, up-to-date picture of the entire network, and quickly identify the trouble spots.

From a practical point of view, we've found out that the use of maps and GIS in a standard network monitoring tool will reduce the monitoring engineer's load by more than 50%, and increase productivity by more than 70%.The requirements for this application are:
  1. Predicted coverage arrays
  2. Counter-based network statistics
  3. Map layers.
The predicted coverage array forms the bedrock for producing dynamic maps used for monitoring the network's status.It is a collection of geo-referenced polygons in space that represent the radial distance of the signal strength away from each cell based on the signal's interaction with environmental factors like the terrain, water, forest, residential areas, etc.It is produced using a specific algorithm in a standard radio frequency planning tool.The planning tool combines a number of coverage predictions into a raster, which contains the best coverage values for each location and other information regarding the serving cell.(See Figure 3.)

Fig 3: Coverage array with unique Cell ID.(Click for larger image.)


Network parameters, or counters, are monitored or captured based on network-defined temporal frequency.These are then stored in a standard database like Oracle or SQL server.Typical hourly counters or statistics account for the network status of every single operational cell in the network (see Table 1).

Table 1: Sample statistics for a few cells for just one hour.(Click for larger image.)



Vector layers like city street maps and point location layers (Figure 4) can then be used to further track network status down to micro level.

Fig 4: Vector layers (Lagos City Map).(Click for larger image.)


Spatial Network Monitoring Tool
Predicted coverage polygons can be imported into a standard GIS stand-alone application.However in this case we've found out that it's better to incorporate the GIS functionalities into the monitoring tool.The advantage is that it automatically creates a seamless connection to the counter based statistics in the Oracle database required to produce the dynamic maps, all within one integrated system, Instead, we would have to import the coverage polygons from the RF planning tool into another tool entirely - in this case a standard GIS application like MapInfo Professional - and then creating an Open Database Connectivity (ODBC) to access the counter database.We have discovered that bundling the monitoring tool with GIS basic functionalities into one integrated system will result in learning to use one tool instead of two, especially if the network-monitoring tool does not have GIS capabilities.

Having exported the coverage arrays into the GIS interface of the monitoring tool, the next task is to tie each serving cell/polygon to its corresponding network statistics resident in the database.Each polygon representing the coverage has a Cell ID (see Figure 3) that uniquely identifies that cell/polygon.This Cell ID could be used as the primary key to merge the static attribute table of the coverage array with the dynamic hourly counter-based network statistics using a simple Structed Query Language (SQL) script.This creates a GIS of a sort in which the geo-referenced arrays served as the spatial data while the dynamic statistics served as the attributes.Once this join had been accomplished, a thematic map is created that gives insight into the geographic spread of a particular KPI, e.g.network traffic or drop call rate for one particular hour.(See Figures 5 and 6.)

_
Fig 5: Geographic distribution of congestion rate in Lagos on February 13, 2004 at 10am.(Click for larger image.)


Each thematic map representing the status of one hour for the same geographic area forms a layer.The process can be repeated for all the hours under review.This flow of information in real-time can be achieved using Web-based reporting tools like Business Objects or Fast Access Tool (FACT).These reporting tools are configured to display dynamic maps in a presentation format similar to PowerPoint presentations.The presentation shows the thematic maps of the same geographic area at different times of the same day.A single glance at the dynamic maps will reveal "red spots" (areas requiring urgent attention) as can be seen in Figures 5 and 6.

Producing this kind of map shows where there have been persistent problems through out the day.Personnel in the region with access to this information quickly know what geographic area is suffering from low quality service and take urgent steps to correct the problem.

Different types of dynamic maps are produced to show different aspect of the network.They are produced based on:
  • Traffic carried in Erlangs
  • Percent traffic channel congestion
  • Percent call drop rate
  • Mean hold time
  • Percent traffic channel availability
  • Percent network resource utilization
  • Cell down
  • Processor load balancing
  • Network dimensions
_
Fig 6: Geographic distribution of congestion rate in Lagos on February 13, 2004 at 8pm.


These dynamic maps can be produced for each of the GSM bands (in places like Nigeria, two frequency bands "" 1800Mhz and 900Mhz "" are used). This type of analysis reveals what band is contributing the most to the congestion in a particular place.Dynamic maps are also produced to show the BSC load balance.

We have achieved significant benefits using a spatial network monitoring tool.
  • Enhance response time to network issues.It simplifies the identification and resolution of network and service performance issues, resulting in maximized network performance, lower capital equipment costs and enhanced service
  • Accelerate deployment and support of new technologies and services.It allows key knowledge users to capture and embed technical expertise and best-practice processes for rapid roll-out and support of new services
  • Enhance understanding of customer and network behavior.Spatial network monitoring tools could help evaluate the impact of voice service on radio network resources to better determine the requirements for future network expansion and the addition of new services.
  • Lower network costs.It reduces operational expenses by minimizing the time and expertise required to perform common network analysis tasks; reduce capital expenditures by taking advantage of support for all commonly used hardware systems, legacy and future.
  • Create an insightful method of better dimensioning the network.This is especially true when introducing new BTS, BSC, and MSC in a more realistic way
Conclusion
Higher quality in a GSM service operation is achievable but only through fast and accurate network optimization.Using GIS in standard network monitoring tools are known to reduce the stress of the quality-monitoring engineer and increase productivity by more than 70%. The task of GSM network optimization is highly complex and specialized, but it is also a task with enormous potential rewards, as each incremental improvement in system performance can translate to huge cost savings and increased revenues for the operator.

Caveat

The views contained in this article are not those of MTN Nigeria Communications Limited or those of its Management or Board of Directors.

References
  1. Planet EV GSM Overview Brochure (RF Planning & Optimization Software)
  2. GSM NETWORK OPTIMIZATION- Motorola Lifecycle Services (http://www.motorola.com/networkoperators)
  3. Wu Jing, Yang Lu, Song Jun De Case Based Knowledge Management and Case Mining in Optimization of GSM Network
  4. Roni Abiri (2001) Optimizing service quality in GSM/GPRS networks
  5. Optima V3.4 User Training Note.Aircom International 2002
  6. Milan S.Petkovic, Slobodanka Djordjevic-Kajan, Dragan H.Stojanovic, Leonid V.Stoimenov.The Role of GIS in Telecommunication Network Maintenance
  7. John Scourias (1997) Overview of the GSM Cellular System Extended Abstract
  8. Ericsson's User Description, Radio Network Statistics 2000
  9. Service Verification Solution: ACTIX Product details http://www.actix.com/products/svs.htm
  10. Bala A.Muhammad (2001): Dawn Of The GSM- Hope and Despair in the Nigerian Telecoms Market.Presented at Annual Conference - South African Communication Association, Pretoria
  11. P.A.Burrough (1993) Principles of Geographical Information Systems for Land Resources Assessment
  12. Christopher Jones (1997) Element of Geographical Information Systems and Computer Cartography
  13. MapInfo Professional (User's Guide) Version 6.5
  14. Ireti Ajala (2002) Emergency Management in a Gas Pipeline Network.
  15. Understanding GIS-The Arc/Info method, Lesson 1: Why GIS? ESRI 1992
  16. US Government Fact book http://www.cia.gov/cia/publications/factbook)
  17. Ernest Ndukwe (NCC's Executive Vice Chairman)(2002) -One year of GSM Revolution, what future for the telecommunication "" A presentation.
  18. Guy Engon Zibi (2002)-Capitalizing on Africa's fastest growing Market


Bookmark and Share


Stay Connected

Twitter RSS Facebook LinkedIn Delicious Apple Devices Android Blackberry






Recent Comments


DirectionsMag.com

About Us | Advertise | Contact Us | Web Terms & Conditions | Privacy Policy
© 2014 Directions Media. All Rights Reserved