Print Email Download Reference This Send to Kindle Reddit This
submit to reddit

Indoor Positioning System Using Wifi Signal Strength Computer Science Essay

Location determination is considered as vital element for producing valuable and inventive location base services. Such services not only increase the network performance but also efficiently improve the user experience. One such system that provides location awareness is Global Positioning System (GPS).The problem with GPS is that it only works outdoors not indoors, and even in outdoors it gives an accuracy of about 20 meters. Therefore, there is a need of a system which can work indoors to provide Location Based Services (LBS).

Out of many Indoor Positioning Systems currently deployed, one such system is WIFI based Indoor Positioning System. It uses the Received Signal Strength Information (RSSI) of WIFI signals from the neighbouring Access Points (APs), to locate a user device inside the building.

Finger Print approach is used to determine user device location. WIFI environment was chosen because of increase in development, ease of use and low cost in deployment of Wireless LAN technology.

1. INTRODUCTION

1.1 BACKGROUND:

Cellular network infrastructure provided the background for the evolution of positioning systems [3]. The enhancement in mobile computing and wireless networking have enabled users to move freely inside an indoor wireless network environment .Location of a user inside this environment will not only help in the management of Networks e.g resource management, load balancing, network planning etc , but will help in providing Location Based Services [2].

All these applications are only possible if the user device is accurately located.

Location Based Services:

Indoor Positioning System can be used for variety of services .Some of the Key Applications are:

Network management and security:

Wireless LAN based Indoor Positioning System can enhance management of networks e.g resource management, assets tracking and load balancing etc.

If a user device location is known then we can enforce location base authentication and authorization [3].such systems are already available in commercial market [5][3].

Context-Awareness:

It is a term derived from Ubiquitous Computing. Context-Awareness systems can change environment based on a certain information such as location .Wireless LAN base Indoor Positioning System can locate a user device.

Problem Description and Motivation:

Motivation for the project came from the useful applications as described above. GPS networks best works outdoors but suffer the limitation of indoor tracking because its signals are weak enough to penetrate the walls. The other problem is that it cannot identify a device location based on floor in a building. Therefore a need for Indoor Positioning System was needed.

There are many techniques and systems through which we can deploy Indoor Positioning System, will discuss later. But the optimal reason for choosing Wireless LAN was the low cost, Less Processing and easily availability of WIFI Access Points. No additional hardware is required for implementing Indoor Positioning System based on Wireless LAN.

The goal here is to implement Wireless LAN Indoor Positioning System in Glasgow Caledonian University (GCU) building, inside one of the floors and check the feasibility of WIFI signals as the mean for locating user device.

A new search algorithm will be developed, which will resolve some of the problem in WIFI Indoor Positioning System

Current status and Developments of Research or Technology:

An overview of the current wireless positioning systems with related work is discussed below.

Literature Survey and Analysis:

Global Positioning System (GPS):

Global Positioning System provides location awareness in outdoor environment. It is used for many military and commercial applications. It uses triangulation technique to track a receiver. So it requires having at least four satellites with clear line of sight to calculate the distances to GPRS receiver.

In case of obstructed path such as buildings, trees etc the GPRS receiver won’t be able to maintain clear line of sight with the satellites, which can make significant impact on its performance.

Location determination is very essential in mobile computing. The wide range of context-aware applications has inspired researchers to design and implement location aware systems specifically in the area where Global Positioning System don’t work, Such as indoors and path where there is no clear line of sight [6].

2.1.2 Indoor Positioning Technologies:

2.1.2.1 Infra Red ( Indoor positioning System):

Infrared based systems works on the principle that each user device emits a proprietary infrared beacons, which is directed by the receiver and pass it to the locating software. These infrared beacons are unique codes for specific devices. But the problem with it is that infrared signals requires line of sight and does not penetrate through opaque objects such as walls. Therefore a number of receivers are required to efficiently implement this system indoors.

The advantage is that it won’t interfere with other frequencies. The disadvantage here is the cost and complexity of implementation. As many Infrared reader is required, especially around the corners of the building and if the asset itself is blocking the line of sight to the Infrared reader.

Systems that based on Infrared are Xerox Parc Tab[4] and Olivetti Active Badge System [7].

2.1.2.2 Radio Frequency (RFID) Identification:

To implement RFID,RFID scanners are installed inside the building. These scanners actively detect active and passive tags that are attached to the objects (user device to be detected). These transceivers (tags) are identified based on its unique electronic code. The difference between the two is that active tag uses batteries and can be identified by a scanner from a maximum of 20 feet , while positive tag receive power from the RFID scanner and it has to be in close proximity of RFID scanner.

The disadvantage here is again the cost and complexity of implementation and it interference with other wireless technologies, such as WIFI. So if a client wants to implement wireless technology to be implemented in the facility, then the performance of RFID base location determination can be degraded.

One of the system that is based in RFID is 3D-ID from PinPoint [8]. Scanners are installed around the facility which transmit signals of 2.4 GHZ. Active or passive tags that are attached to devices emit signals 5.8 GHZ. Each active tag is identified by unique electronic code, RFID detects it and send it to centralized station. Where all the tag information is stored. Another system that RFID base system is Personal Shopping Assistant by AT&T [9].

2.1.2.3 Location based on Ultrasonic :

Location determination based on Ultrasonic is used by BAT System at AT&T laboratories Cambridge [10].

2.1.2.4 Sensor Fusion:

As indoor environment is very complex in nature. It is very difficult to implement a single system solution for location awareness applications. One such solution is to use Sensor Fusion to improve accuracy of user device location.

Sensor Fusion is define as “The process of combining several and independent observations to obtain improved accuracy and robustness” [11][12].

[11] Proposes an algorithm for fusion. It is explained in the figure below.

Sensor fusion, source [11]

2.1.3.1 WIFI Based Positioning Systems:

Most of the current and WIFi based positioning systems utilizes the Signal Strength as metric for location determination purposes. WIFI positioning systems uses the Strength of receive Signal from Access Points. The mobile device measure the Signal Strength receives from the Access Points.

RF (Radio Frequency) signals fade with distance, especially at indoor environment because of obstructions such as walls, furniture etc. Signal received by mobile device is an indication of that mobile device from the Access Point [2]. The accuracy of this technique depends on the stability of Signal Strength values and uniqueness of these values between different points [2].

This technique is usually implemented in two phases. OFFLINE PHASE and ONLINE PHASE. In OFFLINE PHASE, Signal Strength is measured from all the Access Points on specific positions and RADIO MAP of Signal Strength is created. In ONLINE PHASE, Signal Strength is calculated at the real time by mobile device and is compared with the results obtained in OFFLINE PHASE. The results are compared based on some algorithm and user current location is determined.

2.1.3.2 RELATED WORKS:

A lot of research has been done on how to calculate user position using WIFI Signal Strength. Some of the work done in this field is explained below.

2.1.3.3 RADAR AND WIFE PROJECT:

RADAR, is an In Building RF-based user location and tracking system [1]. It was one of the first projects which used Wireless LANS for location determination at Indoor Environment.

RADAR system consists of two phases. Online and Offline phase. In Offline phase, the Signal Strength Received Indicator (RSSI) and Signal to Noise Ratio (SNR) information from all visible Access Point’s is obtained on specific locations inside the building. The locations are also called calibration points. A Radio Map of the building is composed based on these calibration points. The process of collecting RSSI and SNR on all the calibration points is called fingerprinting[14][1].

As Signal Strength degrades because of opaque objects such as walls, furniture and even user blocking the line of sight between Access Point and WLAN device driver. The difference between a user blocking line of sight and non blocking is reported as 5db[14][1].

In the online phase a user scan the Area for Active Access Point’s and measure the RSSI and SNR. This real time Signal Strength (SS) and SNR Finger Print is then compared with the Radio Map Finger Print already obtained in Offline Phase, using positioning algorithm. In the last step the user device location is estimated, based on the known location from Radio Map entries [2][1]. RADAR uses K-Nearest Neighbour Signal Strength algorithm (K-NNSS) to compute user location. K represents the number of Nearest Neighbours that has the minimum distance from the user real-time location in Signal Space.

The K-Nearest Neighbour coordinates are averaged, which gives a better guess of User Location than if a single Neighbour was used. The figure below shows how averaging of three Nearest neighbour (N1, N2, N3) gives us location (Guess Point, G) that is more closer to the user real-time location, than any of the single neighbours [1].

Averaging of Three Nearest Neighbours

This technique gives us an estimated error in distance of about 2 to 3 meters. The difference between RADAR and WIFE (Wireless Indoor Positioning Based on Finger Print Evaluation) is that Signal Strength and Signal to Noise Ratio values are collected at four orientations (NORTH, SOUTH, EAST, WEST)for a single (x,y) location in signal space using RADAR, while WIFE collects Signal Strength values at eight orientations for a single (x,y) location in signal space.

The problem with both approaches is that it does not address a scenario where some person is standing beside the user device while the user is collecting online data for his location determination. The user device may receive very weak Signal Strength from that particular Access Point, whose Line of Sight is blocked by a person standing beside user device.

2.2 CHALLLENGES AND EXPECTED CONTRIBUTION:

The main objective of this project is to implement the Research work practically using Java in an indoor environment. The technical challenges would be

Orientation issues: User itself blocking Line of Sight to mobile device.

Environmental changes: Indoor environment changes constantly due to people’s movement, different placement of the furniture etc causes propagation losses, because of which signal strength do not remain stable [2].

Aliasing: When two points that are physically apart from each other but they have the same Signal Strength values is called Aliasing. This is mainly due to the complex and unpredictable nature of indoor environment [2].

The main contribution of this project would be to propose a solution for the problem statement, where a person block the line of Sight to user device while the user is scanning his location for location determination.

Project Aim:

The main aim of the project is to implement indoor positioning system inside GCU building. Investigate and evaluate the feasibility of WIFI signals as a mean for location determination. Propose a solution to the problem, if a person blocks the line of Sight to user device while the user is scanning his location estimation.

Project objectives:

Objectives that are setup in achieving this aim;

To study the literature review for understanding the WIFI positioning system in location determination based on WIFI signal strength and Java as simulation software.

To study the indoor propagation environment and the effect of environment on radio wave propagation.

To create a test bed for the implementation in one of the floor in GCU building with at least four rooms.

To design and implement the software, which will be used in location determination of user device.

To check the feasibility of WIFI signals as a mean of locating user device by determining the level of acceptable accuracy and validate it with known results.

Testing and analysing the system after the propose solution to the above problem is applied.

Expected Outcomes/Deliverables:

The following outcomes are expected at the end of achieving the above objectives.

Understanding the creditability of WIFI signal strength as a mean of location determination.

Understanding Java (J2ME) software tool.

Practical implementation of the system and its limitations.

Critical analysis on the accuracy of the system.

Comprehensive and structured documentation of the work done.

The Method of the Project:

This project will be implemented inside the GCU building in one of the floors. To evaluate and physically implement this project a combination of the research approaches already explained and new search algorithm to the problem of someone standing beside a user device , when collecting Signal Strength information for location determination, will be explored. Analysis of the new algorithm will studied and compared.

The method to be employed by the project:

For practical implementation of this project a Test bed will be created and for analysis programme will be written in Java. To attain Aims and Objectives specified in section 3, the following steps are initiated. When a user with wireless LAN device driver enters an area occupied by Access Points, it receives beacon messages, which is broadcasted by all the Access Points in that area. These beacons messages help the client to select the best Access Point in that area. This process is also called Passive scanning. Best Access Point means which have the better Signal to Noise Ratio (SNR) compare to other Access Points. These also help in handoff. In Active Scanning phase the WLAN driver sends out probe request to all Access Point’s in that area. The Access Point’s respond with probe response frame. Mac-address of all the Access Point’s are advertised in the beacon and probe response frame. The signal strength from each Access Point can be calculated through the service provided by WLAN- MAC Layer [15]. In this way Signal Strength (SS) from all the Access Point’s is calculated with respect to the current position. It is the key parameter on which location determination calculation is based upon [14].

4.1.1 OFFLINE PHASE (DATA COLLECTION):

In Offline Phase the User Device measure Signal Strength information from the Access Points. This information is also referred to as “FingerPrints”[1]. These Finger Prints are collected at various locations inside the indoor environment. Depending on the direction when collecting the Signal Strength, a user may have orientation in such a way that there is no direct Line of Sight with the Access Point e.g User body is blocking the Signals. So for that reason Signal Strength is collected at four orientations (NORTH, SOUTH, EAST, WEST) for same (x,y) location. Thus the client device store tuples in the form (x, y, d, SSi,) [1]. (x,y)is the location and ‘d’ represents the orientation at that location e.g (NORTH, SOUTH, EAST, WEST), SSi is the signal strength , where i= {1,2,3,4}, represents four Access Points[1].These Signal Strength values for each location (x,y) are collected and stored in a database. A Radio Map of these Signal Strength values is created in Signal Space.

4.1.2 ONLINE PHASE:

In the online Phase, User Device computes its Signal Strength information (SS1,SS2,SS3,SS4) at real-time from four Access Points , it is then compared with sets of Signal Strength(SSI’,SS2’,SS3’,SS4’) previously obtained at Offline Phase using Euclidian distance formula , sqrt((SS1-SS’1)2+(SS2-SS’2)2+(SS3-SS’3)2+(SS3- SS’3)2) [1]. This gives us Nearest Neighbour location of the current user device with respect to the Signal Space.

A technique is proposed in this Project to solve the issue discussed below. In this project we will be using four Access points. Consider a scenario in which user device is in the Online Phase and measuring the Signal Strength values from the Access Points and someone is standing beside him. The user device will receive less Signal Strength value from the Access Point whose path to the mobile host is blocked. Let i=i:{1....r} physical points and for each physical point (x,y) Signal Strength from the Access Points are calculated and stored in the set (SS1,SS2,SS3,SS4).

From the set of four Signal Strength, the two lowest Signal Strengths are eliminated and only the other two Signal Strengths are used to calculate user location, by comparing with the Signal Strength set (SS1',SS2',SS3',SS4') obtained in the Offline Phase using Euclidian Distance formula .

Say suppose SS1 and SS3 are the Lowest Signal Strength so the values that will be compared are:

sqrt((SS2-SS’2)2+(SS4-SS’4)2).

K-Nearest neighbour algorithm is applied, which represents the K number of nearest Neighbours from user real-time location in Signal Space. An average of the K-Nearest neighbour Signal strengths is taken and that point is User estimated Location.

This process is illustrated in the figure below.

Positioning Process Overview, source [2]

Project Method Illustration, Validation and Evaluation:

A real world scenario will be implemented by using Access Point’s. The whole project will be implemented using Java (J2ME). This projected will be physically implemented inside GCU Building. The main focus will be a single floor with around four rooms due to time restraints. This setup will require Four Access Point’s located in a fix positions.

As discussed earlier for receiving the signal strength (RSSI) from the Access Point’s, software programme will be written using J2ME. The software will consist of

User Interface

Offline And Online data Collection

Radio Map (Signal Strength Database)

Positioning Algorithm

User interface will help the users to operate the software. First of all, Offline Data will be collected on various locations of floor and rooms and will be store in the Database. In Online Phase, when a user collects Signal Strength (SS) from the Access Point’s in real time, will be store in the Database. User device will be located by applying Positioning Algorithm on the Offline and Online data collected.

The performance of our system can be estimated using the “Error Distance”. “It is the Euclidean distance between the actual (physical) location of the mobile device and the estimated location” [1]. The results obtained can be validated with the previous work done in this area [1] [2], and will be shown in terms of graphs and statistical data.

5. Project Plan:

5.1 Project Schedule

The following schedule will be followed throughout the course of project. Contains all the phases of project planning, development, implementation, validation, analysis and reporting in 12 week period.

NO.

TASK NAME

DURATION (Days)

Start date

Finish date

1

Research phase

4 weeks

Week 1

Week 4

1.1

Meeting with supervisor

1 day

Week 1

Week 1

1.2

Research and comprehensive study of Indoor positioning System using WFI Signal Strength.

13 days

Week 1

Week 2

1.3

Study and understanding of JAVA (J2ME)

13 days

Week 3

Week 4

1.4

2nd meeting with project supervisor, discussion about the test bed setup for the implementation.

1 day

Week 4

Week 4

2

IMPLEMENTATION PHASE

6 weeks

Week5

Week 10

2.1

Designing the software for Indoor Positioning system

10 days

Week 5

Week 6

2.2

Implementing the design in software programme (JAVA, J2ME)

4 days

Week 6

Week 6

2.3

Setting up test bed for the implementation of the system, placing Access Point’s in fixed locations. Map.

2 days

Week 7

Week 7

2.4

Offline data gathering for building Radio

9 days

Week 7

Week 8

2.5

Online data gathering, analysis.

10 days

Week 8

Week 9

2.6

Interim report.

6 days

Week 10

Week 10

2.7

Third meeting with project supervisor – discussion of simulation results.

1 day

Week 10

Week 10

3

DOCUMENTATION PHASE

2 weeks

Week 11

Week 12

3.1

Documentation of the project.

6 days

Week 11

Week 11

3.2

Final meeting with supervisor, discussion of Thesis documents.

1 day

Week 11

Week 11

3.3

Compilation of the Project documents.

5 days

Week 12

Week 12

3.4

Submission of draft project report

1 day

Week 12

Week12

3.5

Presentation and submission dissertation report

1 day

Week 12

Week 12

Fig 5.1 Project Schedule breakdown

5.2 Resource requirements:

Resources required for Research;

Unlimited access to GCU library.

Access to internet

Access to online resources such as Athens and IEEE Xplore.

Hardware resources required are;

A laptop with window 7 operating system installed.

Microsoft Office Suite along with Microsoft Office project.

Four Access Point’s (CISCO preferred)

PROJECT MANAGEMENT-GANTT CHART

Week#

1

2

3

4

5

6

7

8

9

10

11

1. RESEARCH PHASE

 

 

 

 

1.1 Meeting with supervisor

 

1.2 Research and comprehensive study of Indoor

 

 

positioning System using WFI Signal Strength.

 

 

1.3 Study and understanding of JAVA (J2ME)

 

 

1.4 2nd meeting with project supervisor, discussion

 

about the test bed setup for the implementation

 

2. IMPLEMENTATION PHASE

 

 

 

 

 

2.2 Designing software for Indoor Positioning system

 

 

2.2 Implementing the design in software(JAVA, J2ME)

 

2.3 Setting up test bed for the implementation of the

 

system, placing Access Point’s in fixed locations. Map.

 

2.4 Offline data gathering for building Radio

 

 

2.5 Online data gathering, analysis.

 

 

2.6 Interim report.

 

2.7 3rd meeting with Supervisor for simulation results.

 

3. DOCUMENTATION PHASE

 

3.1 Documentation of the project.

 

3.2 Final meeting with the supervisor, discussion of

Thesis documents.

3.3 Compilation of the Project documents.

3.4 Submission of draft report to supervisor for review.

3.5 Presentation and submission dissertation report

Fig 5.2 GANTT Chart

5.3 Risk Management:

Limited Time:

Completion of the project in 12 weeks poses a lower risk. But its consequences can be vital. This risk can be mitigated by project management techniques, and a week is reserved in the project schedule on the occurrence of such risk.

Development of Software using Java:

Less experience in Java can cause medium level risk. This is due to the limited knowledge about the software and may result in not developing the required software in time. So a lot of time is reserved for extensive study of tutorials and related documents to overcome this risk.

Loss of Data:

The unpredictable nature of software and hardware may cause loss of data. Although the occurrence of this risk is low but can be very daunting if this happens. This can be avoided by creating multiple backups of the data and excessive documentation.

5.4 Ethical Issues:

Separate test bed environment will be setup for the implementation of the project with separate Access Points, rather than using GCU Access Points. So no ethical approval is required.

Print Email Download Reference This Send to Kindle Reddit This

Share This Essay

To share this essay on Reddit, Facebook, Twitter, or Google+ just click on the buttons below:

Request Removal

If you are the original writer of this essay and no longer wish to have the essay published on the UK Essays website then please click on the link below to request removal:

Request the removal of this essay.


More from UK Essays

Doing your resits? We can help!