Location-aware Maps based on Service Domains

Abstract

We have developed a conceptual model of spatial/geographical service-domains for ambient services [1]. This model facilitates the mapping of web services to spatial contexts and delivers services that are applicable and available based on user location, preferences and priviledges. This project aims to apply the model for guiding users around a university campus in an intelligent and context-aware manner. Thus, we aim to use the conceptual notion of service domains to associate video clips and visualisation of space as the basis for guiding users. A demonstration prototype of such a system will be build for Level 5, C Block that is within range of the Ekahau positioning server.

[1] Loke, S, W., Krishnaswamy, S., and Naing, T, T., (2004), Service Domains for Ambient Services: Concept and Experimentation, ACM/Kluwer Mobile Networks and Applications (MONET) - Special Issue on Mobile Services, (eds.) Qusay Mahmoud and Upkar Varshney, To Appear.

Investigators: Seng Wai Loke, Shonali Krishnaswamy, Wei Yang Ng

Summer 2004-05

Table of Contents

1. Introduction

2. Project Scope

3. The Existing System

4. The Proposed System

4.1 Positioning Client Application (PCA)

4.2 Positioning Proxy Application (PPA)

4.3 Video Playback and Java Media Framework (JMF)

5. Screenshots

1. Introduction

The primary aim of this project is to demonstrate the association of video clips and visualization of space as the basis for guiding users. In order to fulfill this requirement, an application has been developed for the IPAQ handheld pc that receives information from the Ekahau Positioning Engine (EPE) through a proxy application. This document aims to explain the architecture of this system and concludes by presenting the outcomes and findings of this project.

2. Project Scope

Here, the scope of the project is as follows:

  • Ability to track the user’s current location graphically with the use of a mobile device.
  • Be able to track other devices.
  • Auto-scrolling of the map as the tracked device moves.
  • Ability to select key map areas (logical areas) and display information of the area.
  • Association of a video clip with each area. Video playback is performed on the handheld device.
  • System must be able to support multiple users.

3. The Existing System

Fig.1 Existing Setup of the Ekahau Positioning System

The EPE is third party proprietary software developed by, Ekahau Inc. (www.ekahau.com ), which utilizes 802.11 wireless access points to track mobile devices such as laptops and PDAs installed with 802.11 compliant wireless cards. The mobile device is installed with an Ekahau client that sends access point signal strength information to a calibrated EPE. The EPE than computes these raw data into X and Y axis location coordinates. The diagram above shows such a setup within the office located in Level 5, C Block, Monash Caulfield Campus.

A typical installation of the EPE comes with a Java software development kit (SDK) and example Java programs that demonstrates the usage of the classes within the SDK. Running the supplied “TrackingView” example Java program displays an image of a floor plan (stored on the EPE) with dots representing each tracked device along with their IP address. This example program establishes a direct connection with the EPE for tracked devices information.

4. The Proposed System

This section presents the proposed system that aims to meet project scope requirements as outlined earlier in this document.

Fig.2 The Proposed System

Here is a list of factors that influenced the design of the proposed system:

  • The Java SDK supplied with the EPE is not supported by the Jeode EVM.
  • There must be a unique association of each video clip with the corresponding logical area.

From the first factor, it is not possible for the Java application residing within the mobile device to directly establish a direct connection with the EPE. Although the EPE supports the YAX protocol, which allows developers to develop handheld applications where the use of the SDK is not possible, the use of Java RMI has advantages over socket based programming. The use of RMI allows an O-O approach when designing the client-server application.

4.1 Positioning Client Application (PCA)

The proposed system still requires an installation of the Ekahau client application on the mobile device. This allows the device to be tracked by the EPE as mentoned earlier. The PCA is essentially an AWT GUI interface written in Java that displays the location of devices tracked by the EPE. (Note: Any device with a supported wireless card and installed with a running version of the Ekahau client within range of the EPE’s wireless network is considered a tracked device)

4.2 Positioning Proxy Application (PPA)

The PPA is an RMI server that feeds the PCA with real time location information. This is possible by exporting the PCA’s stub to the PPA during a PCA startup. This allows the PPA to invoke the PCA’s methods thus eliminating the need for the mobile device to constantly poll the server for location updates.

At the core of the PPA, is the device tracking class that utilizes the SDK classes supplied with the EPE to connect and receive tracked devices location data from the EPE. A listener is used within this device tracker to send location updates to each PCA.

4.3 Video Playback and Java Media Framework (JMF)

In addition to device tracking on the PCA, JMF is used for video playback on the mobile device. Video clips reside within an Oracle ® database through the use of BLOB type data columns. Users only need to select the hotspot areas within the PCA’s GUI interface to display a panel with relevant information of the area including a video clip.

5. Screenshots

Fig.3 The Main Map

Fig.4 A video clip of the area

Figure three displays a floorplan of the DSTC office. The two dots represent devices tracked by the EPE. The font colored in red shows the current selected device. Checking the checkbox hidden by the “tooltip” enables the follow mode, the screen will then automatically scroll to keep the selected device in view. The areas enclosed by red borders depict logical areas. Clicking on them would display a panel similar to the one shown on figure 4. The top panel displays a video clip associated with the area and the bottom displays textual information of the area.

Back to top