A Survey on Cellular Based Localization and Location Based Services

Abstract

As pervasive and ubiquitous computing concept become more and more popular, accurate positioning plays an important role in the scenario while location-aware applications often need precise geographic information.

Once it is known, we can then employ LBS (Location Based Services) afterward. This report aims at providing a survey on related research projects and industrial developments on cellular-based localization and location-based services.

Then we will discuss some possible issue on this matter and potential research directions.

1. Introduction

Following the evolution of mobile communication, services are enhanced as technologies improve significantly. Carrier seeks more services in order to gain more profits.

Consumers seek more services in order to facilitate their daily lives. On the other hand, in 1996, the Federal Communications Commission (FCC) directed the rules that wireless service companies have to provide location identification for wireless emergency call, and the accuracy of location estimation should be within 125 meters [1].

Recently, LBS (Location-Based Services) becomes an emerging domain of interests by industries, while this kind of service may customize each consumer and offer a better personalized service.

There are numerous ways to locate people, such as GPS, Wi-Fi [13], and Ultra-sound [14]..etc. In this report, we will focus on cellular based positioning methods and their applications thereafter.

2. Cellular-Based Localization

One of the most significant technologies behind Location-Based Service is positioning. Among all the methods of positioning, the Global Positioning System (GPS) is the most recognized one. However, GPS is good for outdoor environment. Its performance is bad in urban

technologies are network based positioning and typically rely on various means of triangulation of the signal from cellular sites serving a mobile phone. In addition, the serving cellular site can be used as a fix for location of the user.

Cellular-Based localization is considered a good solution because base stations (BTS) signals exist almost everywhere in civil world and their location is regarded as stable in a certain period of time since the planning and re-allocation cost is high. This stability characteristic is a key consideration while choosing localization references. In addition, cellular-based positioning requires no additional hardware in the mobile devices and practically every mobile device can measure the signal strengths. What we need to do on the mobile device is to install an extra positioning software, which may calculate the position by itself or communicate with network-based positioning servers.

2.1. Dual Channel Location Estimation

KKH Kan et al. [2], from Hong Kong, provides a method to compensate the weakness of GPS signals in urban area (eg.

the blind spots), by introducing additional GSM signals.

They also provide a mechanism to calibrate base stations from GPS data with a mathematical model, relied on Maximum Likelihood. They show an experiment result by obtaining an average error of 50.53 meters with Weighted GC method and 40.29 meters with Maximum Likelihood method. However, the accuracy is still not enough for a positioning application in an urban area. 40~50 meters in the urban area may mean blocks away.

A Survey on Cellular Based Localization and Location Based Services

連矩鋒 Chu-feng, Lien (p93922007) p93007@csie.ntu.edu.tw

Department of Computer Science and Information Engineering

2.2. Positioning GSM Telephones

In this paper [3], Christopher Drane et al. examine GSM solutions for positioning. They classify mobile services into 4 categories of application: Location-Sensitive Billing, Increased Subscriber Safety, Intelligent Transport Systems (ITS), and Enhanced Network Performance. This paper also offers their own architecture for GSM positioning, but that require the modification of GSM standard, while this is too difficult unless it is initiated from a group of influential government agencies or enterprises

2.3. GSM RSSI-based positioning using Extended Kalman Filter for training Artificial Neural Networks

The paper [4] provides an artificial training method to locate a user by his GSM mobile device, using RSSI (Radio Signal Strength Indication) information. They introduce Extended Kalman Filter (EKF) as a training technique for both Multi-Layer Perception and General Feed Forward ANN (Artificial Neural Network). From the experiment, they achieve the accuracy in the range of 50~200 meters, which is lightly better than previous works, but we still cannot see a significant progress with this method

2.4. Cell Identification Codes for Tracking Mobile Users

In the paper [5], the authors use CIC (Cell Identification Code) for tracking mobile users. By periodically broadcasting a short message which identifies the cell and its orientation relatively to other cells in the network, the mobile users can update their locations efficiently. They propose several CIC encoding schemes and evaluate the performances. In the end, they conclude that CIC-based strategies are shown to outperform the geographic-based method currently used in existing networks, and the timer-based method, over a wide range of parameters.

Moreover, this superiority increases as the number of users per cell increases. However, the method would need to modify the coding method in current communication system, while this is not possible since protocol rarely changes.

2.5. Accurate GSM Indoor Localization

Otsason1 et al. [6] use GSM radio fingerprints (RSSI) to locate people in an indoor environment. They obtain more than 6 strongest cell signals that a regular mobile phone may read from its GSM modem. The increasing information from other weak-signal cells plays an important role of the accuracy of this experiment. They claim the accuracy of 3~5 meters in an indoor environment. Besides the pure GSM method, they try to include additional Wi-Fi fingerprints as calibration, but the result is just slightly better than pure GSM method. We do not see the necessity and usefulness of the combination.

One of the weaknesses in this paper is that the training stage may take a long time and cost manpower since the granularity of the measurement grid is between 1 and 1.5 meters and there is not a automatic method for training phase.

Regardless of the inconvenience of system training, this paper provides a breaking record to achieve the accuracy almost as good as Wi-Fi does. We can learn lots of thing from the paper.

.

2.6. Network based localization

In addition to above mentioned mobile station based positioning techniques, there are some more cellular positioning methods running with network based solution.

The following are parts of the methods:

location of the caller. The accuracy depends upon the cell area and the accuracy can be upto 150 metres for an urban area. Although the accuracy is not high and cannot be applied for emergency usage it is popular amongst the operators as it does not require any modifications in the handset or the network, hence it is comparatively cheap to deploy.

Time of Arrival (TOA)

Here the difference in the time of arrival of the signal from the mobile to more than one base station is used to calculate the location of the device. This method needs synchronization of cellular network using GPS or atomic clock at each base station. The cell sites are fitted with Location Measurement Units (LMUs). By measuring the signal from the mobile phone, the LMUs can triangulate the user’s position.

TDOA (Time Differential of Arrival)

TDOA is a method to calculate an object’s location by the difference of arrival time perceived by various cellular stations. It is used for multilateration, also known as hyperbolic positioning,

U-TDOA(Uplink Time Difference of Arrival)

The U-TDOA method calculates the location of a transmitting phone by using the difference in time of arrival of signals at different receivers known as Location Measurement Units (LMUs). The mobile phone transmits a signal that is received by different receivers at times that are proportional to the length of the transmission path between the mobile phone and each receiver.

AOA (Angle of Arrival).

AoA is a technique for determining the direction of propagation of a radio-frequency wave incident on an antenna array. The technique calculates the direction by measuring the Time Difference of Arrival (TDOA) at individual elements of the array -- from these delays the AoA can be calculated. Generally this TDOA measurement is made by measuring the difference in received phase at each element in the antenna array. This can be thought of as beamforming in reverse. In beamforming, the signal from each element is delayed by some weight to "steer" the gain of the antenna array. In AoA, the delay of arrival at each element is measured directly and converted to an AoA measurement

AGPS(Assisted GPS)

A-GPS is hot in recent mobile phone advances. It is likely to be the standard of positioning in 3G worlds. Mobile phone chip vendors, like QUALCOMM, have built in the A-GPS feature together with their CDMA solution. A-GPS is a technology that uses an assistance server to cut down the time needed to determine a location using GPS. It is useful in urban areas, when the user is located in "urban canyons", under heavy tree cover, or even indoors.

A-GPS method.

Enhanced Observed Time Difference (E-OTD)

E-OTD is a technology for identifying the location of a cellular caller. This is in competition to GPS-based location identification. Main difference between the two methods is that GPS uses a constellation of satellites maintained by the US department of defense. On the other hand, E-OTD uses a mathematical algorithm to identify the location of the caller based on the time signal takes to reach a set of base stations and then through a triangulation scheme, determine the approximate area where the caller might be

3. Location-Based Services

Location is a strategic asset of wireless carriers. Leveraging this information enables the user to experience value-added services and the mobile network operator to offer differentiation and incremental profitability. Many people are familiar with wireless Internet, but many don't realize the value and potential to make information services highly personalized. One of the best ways to personalize information services is to enable them to be location based.

An example would be someone using their Wireless Application Protocol (WAP) based phone to search for a restaurant. The LBS application would interact with other location technology components to determine the user's location and provide a list of restaurants within a certain proximity to the mobile user.

3.1.

The Prospects of LBS

According to a poll by Integrated Data Communications, Inc. (IDC), two-thirds of Americans want wireless location-based services, with safety and security issues (emergency call, roadside assistance, and driving directions) as the primary reasons for wanting location-based services.

Respondents to the survey on LBS are said to be even willing to pay a premium to have location capabilities in their handsets or receive advertising on their handsets to reduce or eliminate telemetric service charges

3.2. Location based billing

The ability to have preferential billing is provided by this type of application. Through location based billing, the user can establish personal zones such as a home zone or work zone. Through arrangements with the serving wireless carrier, the user could perhaps enjoy flat-rate calling while in the home area and special rates while in other defined zones. This type of application can be especially useful when use in conjunction with other mobile applications such as prepaid wireless.

3.3. Emergency services

Hopefully we do not have to rely on dialing 9-1-1 from a mobile phone, but if we do, it is a location based emergency service application that pinpoints your location and relays it the appropriate authorities. The FCC has mandated that by October of 2001, all wireless carriers in the United States must provide a certain degree of accuracy in pinpointing the location of mobile users who dial 9-1-1. However, most GSM systems do not have enough capability to provide the information until 3G network is introduced (network-based positioning).

3.4. Tracking

Tracking is a large category that contains everything from the difficult fleet applications, human tracking, to enabling mobile commerce. Fleet applications typically entail tracking vehicles for purposes of the owning company knowing the whereabouts of the vehicle and/or operator.

Tracking is also an enable of mobile commerce services. A mobile user could be tracking and provided information that he has predetermined he desires, such as notification of a sale on men's suits at a store close to the user's current proximity.

3.5. Case Study [12] -- A Model of Location Based Services for Crime Control (Thailand)

The requirements for the modeling of location based services for crime control are derived from questionnaires from a survey of 100 Thai citizens and 100 Thai police between May to July 2003. According to the survey results, more than two-thirds of the respondents regard safety and security issues (emergency calls i.e. robbery, murder, theft, etc) as the primary reasons for wanting the system. They also want the system to provide information on crime for example crime warning messages, news reporting on crimes, report lost property and crime prevention information. In addition, contact information and digital maps showing the route to the nearest police station, other organizations that support victims, routes to get from one place to another and

information regarding road works and accidents are also required.

Figure 1 (a) Use Case Diagram for Crime Reporting (b) Use Case Diagram for Crime Information Retrieving

Location based services model for crime control

Both police and citizens connect to the system by PDA, mobile phone, PC, etc. The system will check for the authentication and authorization to identify the source of a message. Based on the security policy of the system, the authentication process can deny or allow the user to access the system. The authorization process is used to control access to resources once identity has been verified. Then the data will be obtained by data analysis and converted to longitude and latitude and sent to the database. The database consists of crime data and GIS data which will be integrated before sending to the server. The HTTP server/Map server/WAP server will do the computation.

The Web browser/WAP browser is used for viewing the results.

Location Based Services Model for Crime Control

System Development

They developed the crime control system by integrating Minnesota MapServer, Internet server and crime database into a Web based client/server environment. The MapServer is a system for developing web-based GIS applications. The core of the system is a Common Gateway Interface (CGI) application that allows developers to

presentation templates.

The PostgreSQL database server has been used to manage the attribute data and multi-media content.

PostgreSQL is an Object-Relational DBMS, supporting almost all SQL constructs, including sub-selects, transactions, and user-defined types and functions. The database consists of tables that hold all the attribute data and spatial data. The PostGIS adds support for geographic objects to the PostgreSQL object-relational database. In effect, PostGIS partially enables the PostgreSQL server, allowing it to be used as a backend spatial database for geographic information systems (GIS). Real-time update of authorized users is also made possible. An authentication and authorization mechanism to check which users can be allowed to enter new data or update the existing database is also provided.

Components and Information Flow

Conclusion

The example illustrates the requirements of a crime control system, derived from users’ demands. It illustrated a model of location based services for crime control and the features of the system that we are employing to guarantee user privacy and security. Most of the research presented in this example is still a work in progress. In the near future, the project hopes to access the performance of the model on a very large group of user who are behaving in a very realistic manner.

While the information and messages are generated by a mobile device and both of them are authenticated, they do not take into consideration that if the criminal purportedly put the mobile in one place and go to another place actually.

Since this is an application for criminal control, security and safety of other citizens should be strictly guaranteed. A good idea is to combine bio information to the mobile device and thus generates a unique certificate for authentication.

3.6. Other Applications

The extent of coverage of LBS has virtually covered all the walks of life from selecting the restaurants to emergency services to aid in navigation, it is all there. Some of the fields where network operators are moving in are:

． The ability to pinpoint location is going to revolutionize customer billing in the wireless industry as

billing models. For an example, a family could receive better rates for residential location.

． Fleet management using the remote tracking of taxis and trucks will be common place. One could know the proximity of their trucks or packages and in years to come even the precise location of their goods.

． Roadside assistance and driving directions will become more prevalent with the availability of improved location service. It could later turn out to be turn-by-turn driving directions a complicated location service.

． If you are driving to the theatre and are low on gas, your handset will know there is a gas station three blocks ahead on the right. It will display directions to the theatre and even indicate that you can upgrade your tickets to orchestra seating if you request it in the next few minutes.

． On the leisure and sports sector, golfers would tag their ball with a satellite transmitter before and after each shot. The location data would be shot to a satellite and tracked. When users go back to their home they would be able to track their performance on a Web site, complete with maps of fairways marked with the flight of their ball.

4. Research Issues

While other techniques (ie. UWB) claim a more accurate positioning result within 2 meters. [9], cellular based positioning is still far from what people may need to acquire an exact location, especially in outdoor environments. One possible solution is we can try to combine Wi-Fi signals in compensating cellular methods. As the number of public APs grow exponentially around the city, it is feasible to realize cellular localization with the idea.

On the other hand, while the location of a person might be know during the enquiry stage, privacy issue will surely be raised. Barksuus et al. [8] provide a solution for LBS and show some experiment results. It sounds good to continue the idea with enhanced features. Further implementation can comprise the level of privacy desired by the users and adaptive privacy control.

Another interesting direction can be to develop a virtual on-line game, where virtual people (on the net) and real people (on the street) can play together, like Can you see me now? [11]. People will find a different type of game experiences if we can break the spatial barrier.

Child and elderly people tracking may be another meaningful direction of research, while this kind of application can immediately fulfill the safety request. It is very suitable to adopt cellular based healthcare because cellular network is ubiquitous and we do not need an extra device to be put on those who we care.

5. Conclusion

Based on the accuracy, positioning technologies can be

divided into 3 categories: basic, enhanced, and advanced.

Basic methods are those who build their applications on using cell identification. Cell ID can be used alone, or together with timing advance and network measurement report. E-OTD is often referred to as an enhanced positioning method. Assisted GPS (A-GPS) is an advanced positioning method

After we can correctly position a person or merchandise, there are plenty of applications we can develop to facilitate the efficiency or customize the services.

What I am going to do with my future research is to develop a cellular based positioning system where users can have the ability to locate their position with a regular mobile phone. The project will need a training phase for collecting GIS information, but we will enlarge the granularity of training density to save costs in this stage. After the training phase, we will use the database for testing and approximating runtime user locations. Considering the failure of positioning and changes of base stations, we will also develop an auto-correction mechanism for the purpose.

From that idea, we hope we can develop an auto-learning geographic database with least effort needed in initial setup.

References

[1] Federal Communications Commission, ” Revision to the commission’s rules to ensure compatibility with enhanced 911 emergency calling system”, July 1996, CC Docket No.94-102

[2] .KKH Kan, SKC Chan, JKY Ng., “A Dual-Channel Location Estimation System for providing Location Services based on the GPS and GSM Networks.”, Advanced Information Networking and Applications, 2003

[3] Christopher Drane, Malcolm Macnaughtan, and Craig Scott.,

“Positioning GSM Telephones.”, IEEE Communications Magazine April 1998

[4] Koteswara Rao Anne, K.Kyamakya, F.Erbas, C.Takenga, J.C.Chedjou, “GSM RSSI-based positioning using Extended Kalman Filter for training Artificial Neural Networks”, IEEE Vehicular Technology Conference, 2004.

[5] ZOHAR NAOR, HANOCH LEVY and URI ZWICK, “Cell Identification Codes for Tracking Mobile Users”, Wireless Networks 8, 73–84, 2002 2002 Kluwer Academic Publishers [6] Veljo Otsason1, Alex Varshavsky2, Anthony LaMarca3, and Eyal de Lara2, “Accurate GSM Indoor Localization”, ACM Ubicomp 2005

[7] GSM Association, “Location Based Services”, veriosn:3.1.0, January, 2003 (PRD SE.23)

[8] L Barkhuus, A Dey. “Location-based services for mobile telephony: a study of users' privacy concerns”, - Proceedings of INTERACT, 2003

[9] Ubisense http://www.ubisense.net/

[10] B Rao, L Minakakis. “Evolution of mobile location-based services” - Communications of the ACM, 2003

[11] BLAST Theory “Can you see me now?”

http://www.blasttheory.co.uk/bt/work_cysmn.html

[12] R Boondao, V Esichaikul, NK Tripathi, “A Model of Location Based Services for Crime Control” - Map Asia 2003

[13] YC Cheng, Y Chawathe, A LaMarca, J Krumm, “Accuracy characterization for metropolitan-scale Wi-Fi localization“ - Proceedings of the 3rd international conference on Mobile Systems, Applications, and Services, 2005

[14] NB Priyantha, A Chakraborty, H Balakrishnan, “The Cricket location-support system” - Proceedings of the 6th annual international conference on Mobile Computing and Networking (ACM MOBICOM ’00), 2000

[15] JJ Caffery, GL Stuber, “Overview of radiolocation in CDMA cellular systems” - Communications Magazine, IEEE, 1998