IMPROVING DIRECTORY-LESS WLAN POSITIONING

BY DEVICE WHISPERING

Karl-Heinz Krempels and Martin Krebs

Department of Computer Science, Informatik 4, RWTH Aachen University, Ahornstrasse 55, 52074 Aachen, Germany

Keywords:

Indoor Positioning, WLAN, Geo Tagging, Whispering.

Abstract:

Existing positioning systems do not provide the required positioning accuracy for navigation systems in indoor

environments. Novel system approaches are based on ﬁngerprinting and triangulation techniques. Thus, they

suffer on low positioning accuracy due to multipath propagation and different sending power of the considered

access points. Other approaches are based on tagged WLAN (Wireless Local Area Network) access points or

GSM (Global System for Mobile Communication) base stations with their corresponding position stored in a

central tag directory. This would cause high communication costs for a mobile device that queries the directory

frequently. In this paper we present a ﬁltering technique for access points to determine the closest ones to the

mobile device. The geographical position of the mobile device is calculated from the geo tags broadcasted

by the access points in the mobile device’s vicinity. Systems based on this approach will provide the same

accuracy as directory-based positioning systems at a low cost.

1 INTRODUCTION

Since wireless networks are becomingmore present at

many places, the vision of ubiquitous and pervasive

computing can now become true. However, naviga-

tion and guiding systems need the current position of

a mobile device to provide the according information

to its users’ context. Current outdoor positioning ap-

proaches are based on GPS that does not works indoor

due to the limited reception of GPS signals inside of

buildings. This paper proposes a new directory-less

approach for WLAN indoor positioning which can be

used to realize indoor navigation and guidance sys-

tems at airports or railway stations, e.g. to guide the

passenger to his gate or to the next restaurant. This

approach does not need any additional server infras-

tructure or additional transmitter antennas, because it

uses already existing WLAN infrastructures.

This paper is organized as follows: In section

2 we discuss well known existing positioning ap-

proaches applicable in GSM and WLAN infrastruc-

tures. Furthermore, we discuss the positioning accu-

racy for mobile devices provided by approaches pre-

sended in the past and the arising expenses for mo-

bile users. Section 3 discusses the proposed approach

of SSID-based WLAN positioning by device whis-

pering, the involved components of the IEEE 802.11

WLAN standard, and the push and pull interaction

models for SSID WLAN positioning. Section 4 dis-

cusses a short application scenario and Section 5 com-

prises the conclusion.

2 POSITIONING APPROACHES

Positioning systems estimate the position of a mobile

device with the help of calculus concerning the atten-

uation of communication signals send and received

by the device. Well known approaches for position-

ing are: proximity detection and trilateration (or mul-

tilateration). Systems based on proximity detection

operate upon a set of antennas with well known po-

sitions and use the position from the strongest signal

(Yeung and Ng, 2007) to determine the position of

a mobile device. Systems based on trilateration are

using the strength values of three (or more) signals

send to or received by a mobile device to estimate its

position. Proximity detection requires a dense grid

of antennas for a high accuracy which would cause

very high installation costs in indoor environments.

Trilateration (Staras and Honickman, 1972) (Warren

et al., 1972) provides an acceptable accuracy at a

lower cost, but suffers on a high inﬂuence of envi-

ronmental changes, e.g. wall humidity in buildings

resulting from rain, moving (radio) indoor obstacles

like people, etc. An improved version of the trilat-

225

Krempels K. and Krebs M. (2008).

IMPROVING DIRECTORY-LESS WLAN POSITIONING BY DEVICE WHISPERING.

In Proceedings of the International Conference on Wireless Information Networks and Systems, pages 225-229

DOI: 10.5220/0002027002250229

 SciTePress

eration approach is the envinronment ﬁngerprinting

method (Jan and Lee, 2003), due to that for all the

rooms in a building a set of radio signal vectors (ﬁn-

gerprints) is collected and saved in a central directory.

A device interested in its position will send its actual

signal vector to the central directory and the directory

will send the closest position to the provided signal

vector back to the device.

2.1 Positioning Accuracy of Mobile

Devices

Proximity detection in GSM networks could use the

Cell-IDs to estimate the position of a mobile device

(Kunczier and Anegg, 2004). The very low accuracy

of this systems in outdoor scenarios will be worser

in indoor scenarios, making them unsuitable for in-

door positioning or guidance applications. The accu-

racy for Cell-ID based approaches could be improved

by probability calculus (Borenovic et al., 2005), but

not in that matter to make them suitable for indoor

positioning. In (Krempels and Krebs, 2008) a new

approach for indoor positioning is proposed that is

based on geo-tagged WLAN access points and tri-

alteration. The proposed solution assumes that there

exists a WLAN infrastructure in that all access points

are tagged with their geographical position and mo-

bile devices will use this tags to estimate their posi-

tion. The accuracy of this system seems to be suit-

able for guidance and navigation applications inside

of large buildings.

Trilateration provides only a very low accuracy in

GSM networks, due to the large diameter of GSM net-

work cells (Kos et al., 2006). Thus, it is not very use-

ful for navigation or guidance systems in indoor envi-

ronments. One way to improve the positioning accu-

racy is to increase the density of antennas that would

increase the network operation costs. However, this is

not the objectiveof the network provider and operator.

The deployment of the trilateration approach in exist-

ing indoor network infrastructures like WLAN would

reduce the costs, but the position estimation becomes

difﬁcult due to the different signal strengths of the

WLAN access points transmitter and multipath signal

propagation (Wallbaum, 2004). The positioning accu-

racy could be improved with the help of probabilistic

calculus, but this is also highly inﬂuenced by envi-

ronmental changes. Benchmarks regarding the posi-

tioning accuracy of famous wireless location systems

proposed in the past are discussed by Wallbaum and

Diepolder in (Wallbaum and Diepolder, 2005).

2.2 Mobile Device Positioning Costs

Wireless positioning systems based on proximity de-

tection in GSM networks using Cell-ID’s (Kunczier

and Anegg, 2004) (Borenovic et al., 2005) require a

directory on the network site that maps a Cell-ID to

the coordinates of the corresponding location. This

means, that a mobile device must either establish a

connection to this directory for each positioning re-

quest or must remain connected all the time. Both

cases would produce continous positioning costs. A

directory is also necessary for wireless positioning

systems in GSM networks that are based on trilater-

ation or ﬁngerprinting causing also continuous costs

for the positioning of a mobile device.

Trilateration positioning systems for WLAN in-

frastructures proposed in the past are also directory

based (Wallbaum and Spaniol, 2006) (Wallbaum and

Diepolder, 2005). To determine its position, a mobile

device scans its vicinity for WLAN access points and

send this list to a directory server. The server pro-

vides the coordinates which are related to these ac-

cess points. If there is no positive match, the server

responds with an error and the client cannot deter-

mine its current position. The main drawback of di-

rectory based systems is that clients must establish a

connection to the directory server that would cause

high costs for the positioning information, which is

required very often for navigation and guidance ap-

plications.

The indoor positioning approach based on geo-

tagged antennas (WLAN access points) proposed in

(Krempels and Krebs, 2008) operates directory-less.

Therein, the central directory required by other ap-

proaches was removed in favour of reduced position-

ing costs for a mobile device. The approach is suit-

able for indoor and outdoor scenarios with an existing

WLAN infrastructure without GPS coverage.

3 DIRECTORY-LESS INDOOR

WLAN POSITIONING

In the directory-less indoor positioning approach

based on geo-tagged antennas (Krempels and Krebs,

2008) the Service Set Identiﬁers (SSID) of each

WLAN access point encodes the geographical coordi-

nates of the access point. A mobile device with an em-

bedded WLAN receiver will receive the broadcasted

SSIDs from a number of access points and could de-

code their geographical coordinates immediatly.

WINSYS 2008 - International Conference on Wireless Information Networks and Systems

226

3.1 Service Set Identiﬁer

In the following we describe the Service Set Identi-

ﬁers which is deﬁned by the 802.11-1999 (Commit-

tee, 1999) standard. For the discussed approach in

this paper only the SSID is useful:

• The SSID indicates the name of the WLAN cell

which is frequently broadcasted in beacons. The

length of the SSID information ﬁeld is between

0 and 32 octets. A zero length information ﬁeld

indicates the broadcast SSID.

• Extended Service Set Identiﬁer (ESSID): Multiple

APs have the same SSID and are connected to a

larger cell on layer 2 which is than called ESSID.

• The Basic Service Set Identiﬁer (BSSID) is a 48-

bit ﬁeld of the same format as an IEEE 802.11

MAC address. It uniquely identiﬁes a Basic Ser-

vice Set (BSS). Normally, the value is set to the

MAC address of the AP or a broadcast MAC ad-

dress in an infrastructure BSS.

3.2 Pull Model

In the pull model every wireless access point (AP)

broadcasts the same SSID like ’geo’. In the next step

the client associates with the AP and gets an IP ad-

dress with DHCP. In the last step the client will query

the positioning service provided by the access point to

retrieve the GPS coordinate of the AP. However, due

to the long interaction chain this model works but it is

not very suitable in a time and energy saving world.

3.3 Push Model

The push model is more efﬁcient with respect to

the communication overhead, but limited to querying

GPS coordinates only. Additional information can-

not be transmitted due to SSID space limitations of

32 characters. Every AP broadcasts a unique SSID

which is exactly the coded GPS coordinates of the

AP. The client needs only to scan for speciﬁc geo

SSIDs and selects the geo SSID with the highest sig-

nal strength. It is not necessary that the client asso-

ciates with the base station, because the client can re-

trieve all information from the already scanned SSID.

An example for the location of the city of Aachen

(longitude:6.1, latitude:50.7667) in the speciﬁed tag

format h

geo :longitude,latitude

i would be

geo:6.1, 50.7667

i and the corresponding broad-

casted SSID

geo:6.1,50.7667

3.4 SSID WLAN Positioning

The position of the mobile device could be estimated

with the help of interpolation calculus, by using only

the coordinates of the m strongest signals from n sig-

nals received by the device, or even of a combination

of both, e.g. by interpolating the coordinates encoded

in the m strongest signals. However, the result will be

an area or even a space, and the position of the mo-

bile would be close to its center of gravity (in the ideal

case).

Figure 1: WLAN SSID-Positioning.

Determining the position of the mobile device

only with the help of the strength of the received

signals is highly inﬂuenced by the different sending

or changing sending power of the considered access

points. Thus, we could not assume forthermore, that

the strongest signal is received from the closest access

point. In Figure 1 the signal received from AP

could

be stronger than the signal received from AP

3.5 SSID WLAN Whispering

In Fig. 1 the mobile device MD receives the signals

and SSID’s from the access points AP

, AP

, . . . , AP

To select the closest geographical vicinity of the mo-

bile device, we introduce the whispering approach.

Due to the fact that a mobile device is able to con-

trol its WLAN radio interface it can control also its

sending power. The characteristics of its receiving an-

tenna are not inﬂuenced thereby, so that the list of ac-

cess points received by the mobile device would not

IMPROVING DIRECTORY-LESS WLAN POSITIONING BY DEVICE WHISPERING

227

change. WLAN radio whispering consists in reduc-

ing the sending power of a mobile device to a mini-

mal value and queriing a subset of the visible access

points for management information (Fig. 2).

Figure 2: Radio Whispering to Detect the Close Vicinity.

Due to the reduced sending power of the mobile

device only the access points, that are geographically

very close to the mobile device will receive its query

and will answer to it. Thus, the effect of whispering

is a ﬁlter that is robust against signal multipath prop-

agation and power oscillations or automated adaption

of access points. An idealistic abstraction of the whis-

pering effect is shown in Fig. 3.

In the WLAN communication range of the mo-

bile device MD the access points AP

, AP

, . . . , AP

are visible (Fig. 1). AP

and AP

will receive the

information query send with very low power by the

mobile device (Fig. 2) due to their close vicinity to

it. Access point AP

answers to the query (Fig. 3)

and the mobile device can extract its position from

the SSID of AP

4 APPLICATION SCENARIO

Many indoor navigation and guidance applications

suffer on high positioning costs and on low positoin-

ing accuracy. The business cases of a subset of this

systems are based on low cost or free positioning

and do not require high accuracy positioning. Thus,

it seems that even with a low positioning accuracy

Figure 3: Answer of the Close Vicinity.

(less than twentyﬁve meters) navigation and position-

ing applications could be deployed and used. In Fig.

4 a guidance scenario is shown that could be im-

plemented with the help of the positioning approach

discussed in this paper. The scenario is based on a

planned trip consisting of a travel chain. Each el-

ement of the chain has an expected duration and a

travel mode (e.g. walking, ﬂying, travalling by bus,

etc). For the most travel modes the operation vehicle

(e.g. bus, train) and its route is known in advance.

Thus, the positioning accuracy could be improved if a

determined (rough) position is mapped to well known

trajectories of a planned route at the respective time,

e.g. corridors, stairs, etc. The scenario in Fig. 4 shows

a travel chain element with the travel mode walking.

A traveller is guided with the help of discrete position

points mapped to his planned route to the right gate,

e.g to take his plane.

5 CONCLUSIONS

In this paper, we presented the whispering technique

to improve the positioning accuracy in directory-less

indoor WLAN positioning. The advantage of this ap-

proach is that there is no need to establish an online

Internet connection, and that it applicable indoor and

outdoor. The positioning accuracy is determined by

the radio range of the access points which could be

seen by a mobile device, and the WLAN radio of

the device itself. The proposed solution enables mo-

WINSYS 2008 - International Conference on Wireless Information Networks and Systems

228

Bus

Walk

Train

Walk

Plane

Gate 1

Gate 2

Gate 29

Gate 30

Figure 4: Application Scenario.

bile devices to be used in large indoor environments

with an existing WLAN infrastructure for guidance

and navigation, e.g. for ealderly people in airports,

train stations, etc. In a next step the solution will be

integrated in a hybrid positioning system, based on

GPS, GSM, and WLAN. The short term objective is

to provide a directory less, best effort positioning sys-

tem.

REFERENCES

Borenovic, M., Simic, M., Neskovic, A., and Petrovic, M.

(2005). Enhanced Cell-ID + TA GSM Positioning

Technique. Computer as a Tool, 2005. EUROCON

2005.The International Conference on, 2:1176–1179.

Committee, L. M. S. (1999). ANSI/IEEE Std 802.11, 1999

Edition, Part 11: Wireless LAN Medium Access Con-

trol (MAC) and Physical Layer (PHY) Speciﬁcations.

IEEE Standard.

Jan, R.-H. and Lee, Y. R. (6-9 Oct. 2003). An Indoor Ge-

olocation System for Wireless LANs. Parallel Pro-

cessing Workshops, 2003. Proceedings. 2003 Interna-

tional Conference on, pages 29–34.

Kos, T., Grgic, M., and Sisul, G. (June 2006). Mobile

User Positioning in GSM/UMTS Cellular Networks.

Multimedia Signal Processing and Communications,

48th International Symposium ELMAR-2006 focused

on, pages 185–188.

Krempels, K.-H. and Krebs, M. (2008). Directory-less

WLAN Indoor Positioning. In Proceedings of the

IEEE International Symposium on Consumer Elec-

tronics 2008, Vilamoura Portugal.

Kunczier, H. and Anegg, H. (5-8 Jan. 2004). Enhanced Cell

ID Based Terminal Location for Urban Area Loca-

tion Based Applications. Consumer Communications

and Networking Conference, 2004. CCNC 2004. First

IEEE, pages 595–599.

Mayrhofer, A. and Spanring, C. (2007). A Uniform

Resource Identiﬁer for Geographic Locations (’geo’

URI), draft-mayrhofer-geo-uri-01. IETF Internet

Draft. work in progress.

Staras, H. and Honickman, S. (Feb 1972). The Accuracy

of Vehicle Location by Trilateration in a Dense Urban

Environment. Vehicular Technology, IEEE Transac-

tions on, 21(1):38–43.

Wallbaum, M. (5-8 Sept. 2004). Tracking of Moving Wire-

less LAN Terminals. Personal, Indoor and Mobile Ra-

dio Communications, 2004. PIMRC 2004. 15th IEEE

International Symposium on, 2:1455–1459 Vol.2.

Wallbaum, M. and Diepolder, S. (19-19 July 2005). Bench-

marking Wireless LAN Location Systems Wireless

LAN Location Systems. Mobile Commerce and Ser-

vices, 2005. WMCS ’05. The Second IEEE Interna-

tional Workshop on, pages 42–51.

Wallbaum, M. and Spaniol, O. (Oct. 2006). Indoor Posi-

tioning Using Wireless Local Area Networks. Modern

Computing, 2006. JVA ’06. IEEE John Vincent Atana-

soff 2006 International Symposium on, pages 17–26.

Warren, W., Whitten, J., Anderson, R., and Merigo,

M. (Aug 1972). Vehicle Location System Experi-

ment. Vehicular Technology, IEEE Transactions on,

21(3):92–101.

Yeung, W. M. and Ng, J. K. (21-24 Aug. 2007). Wireless

LAN Positioning based on Received Signal Strength

from Mobile Device and Access Points. Embedded

and Real-Time Computing Systems and Applications,

2007. RTCSA 2007. 13th IEEE International Confer-

ence on, pages 131–137.

IMPROVING DIRECTORY-LESS WLAN POSITIONING BY DEVICE WHISPERING

229