Location Aware Systems

Attention elsewhere

hiof — Mon, 22 May 2006 06:15:50 +0000

The last week I have done very little of what I planned. The testphones was needed by students from the other master course on a presentation they had last friday. I used this as an "excuse" to finish the main report in one of my other courses which is due friday the 26.

Today the phones will be back and it is time to implement a simple map view. Most of the code is allready written and it is merely a question of making it work.

Close to testing

hiof — Mon, 15 May 2006 00:12:34 +0000

During the last 2 weeks most of the bricks have found their positions.
Due to time issues I have choosen to implement the program solely on the telephone. This is made possible due to the significantly small graphfile containing the road network. When the amounts of road data increase a server will be the only reasonable solution, but for know it is possible to get the answers I am looking for using only the phone.

To set one thing straight first; signal strenght can be get from the phone directly and not from the location.gsm_location() method. This method only returns the cell towers range. As we see it is still possible to use it.

When it comes to my pseudocode it is still under construction. There are only details left, like how many nodes it should traverse through in the graph. However this is something that should be based on the testing later this week.

pseudo code

Most of the code is also implemented. The biggest obstacle left is using the information I allready have to fetch relevant map data from a suitable mapserver. There are many suitable examples and this should go fast and easy.

Furthermore the graph file is ready and it looks like this:
Graph.txt

The graph file containes the following information for each Node: Nodenumber, Address, cellID, connectedNodes, Latitude and Longitude, IsMarked(bool). It is easily extended with signal strenght.

So far the program have been tested with setting a current node and gettiing information about the next celltower.
Additionally tabs have been implemented and this took some time. To show how the program works, and it does , I will demonstrate by showing two lines from the graph file:

4,Glennevn,1013,3 5 8,59.127,11.369,0
5,Glennevn,1012,4 6, 59.129,11.372,0

When the program Cirt.py
is run with the node 4 as current node and cellTowerId=1012 it uses the connected nodes in node 4 and finds the correct next position.

There are three things left to improve.
1. the algorithm must be shaped according to the testing
2. graphical maps should be shown to the user
3. testing of the system to determine its accuracy.

Graphs and the roadnetwork

hiof — Tue, 25 Apr 2006 10:07:11 +0000

This document discuss the different possibilities in using graphs to represent the road network.

http://www.ia-stud.hiof.no/~johnesim/rss/Graphs.doc

References

hiof — Fri, 31 Mar 2006 09:29:47 +0000

When writing a scientific report references is a crucial element. If you like your report to be scientifically accepted good references are the keys that will help you through that door. Friday 31.03.06 our eager librarian held a lecture in this matter. To make life easier for those of you that did not attend (all of the class minus three) I will here post the most important tips and tricks.

How should the references look like?
Is it a book, a chapter or a webpage?

References should be in such a way that they are easy to comprehend.
I you refer to a textual source you are to use [].
In Norwegian for starters:
[X] Fornavnets Initialer + Etternavn
      Tittel i kursiv
      Utgitt sted
     Utgitt av utgiver
      Utgitt år, (pp) fra side- til side

Example:
[1] W.K. Chen, Lineas Networks and Systems
Belmont, CA: Wadsworth, 1993, pp 123-135

Book:
[1] W.K. Chen, Lineas Networks and Systems
Belmont, CA: Wadsworth, 1993, pp 123-135

Chapter of a book::
[2] G. O Young, "Synthetic….", in Plastics, 2nd ed.,
vol. 3, J. Peters, Ed. New York:
McGraw-Hill, 1964, pp. 15-64.

Published articles:
[3] J. U. Duncombe, "Infrared communication", IEEE Trans.
Electron Devices, vol. ED-11, pp. 34-39, Jan. 1959

[4] E.P. Wigner, "Theory of…", Phys. Rev.,
      vol. 134, pp. A635-A646, Dec. 1965.
Published in Conference reports:
[5] D. B. Payne and J.R. Stern, "Wavelength-switched passively coupled….",
      in Proc. IOOC-ECOC, 1985, pp. 585-590.
Papers presented at a conference:
[6] D. Ebehart and E.Voges, "Digital…",
      presented at the 2nd. Int. Conf. Optical fiber…, Stuttgart, Germany, 1984.
Patents and standards
[7] G. Brandli and M. Dick, "Altering current fed powersupply",
     U.S. Patent 4084217, Nov. 4, 1978.
Technical reports
[8] E. E. Reber, R. L . Mitchell, and C.J. Carter, "Oxygen absorb…",
      Aerospace Corp., Los Angeles, CA, Tech. Rep. TR-0200
      (4230-46)-3, Nov. 1968.
Electronical rexources
[10] J.Jones. (1991, May 10). Networks. (2nd ed.) [Online].
        Available: yourlinkhere.

Remember(1991, May 10) is when the resource is read.

A good document for more information is found here: http://wwwlib.murdoch.edu.au/find/citation/ieee.html

Presentation and evaluation

hiof — Mon, 27 Mar 2006 07:04:09 +0000

Project presentation is found here:

http://www.ia-stud.hiof.no/~johnesim/rss/GPSapp.ppt

Evaluation of Morgan Jakobsens project is found at:

http://www.ia-stud.hiof.no/~johnesim/rss/eval.ppt

Project overview

hiof — Tue, 21 Mar 2006 20:33:15 +0000

Short introduction

Cell information is a way of finding a position based on a mobile phones connection to a GSM antenna. This method is somewhat inaccurate with deviations of 50-400 metres. What I look into in this project is to add additional information into the equation to
better the position. Until today, using a mobile phone for
positioning, gives a position with a deviation of 50-400 meters.
This is by far too inaccurate when you compare it to a GPS which gives you 3-15 metres at the most. Using information we already know, in my case geographical road
information, it is possible to decrease the deviation.

The way I want to improve position is by adding a server with “knowledge” about a certain area. This server then correct information from the cell information service according to what the server knows.

Hence my problem definition is:

Can cell information accuracy, by the use of additional geographical information, be increased to an extent where it is usable as a GPS road tracking substitute?

Where do I stand:

So far I have stumbled Halden to Aremark and in this same process experimented with cell information and different programs available for this. As I have chosen Python as the programming language to use I have written code to test out a client and information transfer between the phone and an internet connected server. This is very straightforward when using the correct Python libraries. Furthermore I have structured a rouge final solution and decided where to go and how. For more information take a look in my living project report linked on the botoom of this post. Until Monday 20.03.06 this project had two participants. Today, the following day I am down to one. This changes my plans but it should be quite possible to give a conclusion when the project finishes.

New work plan:

By the end of week 12 finish the queues in my project report and have a meeting with Misund on server algorithm for positioning.

Week 13: Look into server algorithm and decide upon client functionality.

Week 14: Start server implementation and test it out against test data. SVG usage in the client and the algorithm for this, both server and clientside.

Week 15: Easter vacation. Work with report.

Week 16: Build client and start testing against server. Work more with the server as this is where most of the work needs to be done. Also training the server database.

Week 17: Continue server development while adding client functionality. The server need a client for decent testing. The client need graphics and added functionality for better usage. This will also demand more from the server.

Week 18: Finish the project report and evaluation.

Link to the current project report: http://www.ia-stud.hiof.no/~johnesim/rss/main.pdf

General system illustration

hiof — Mon, 20 Mar 2006 11:32:24 +0000

The project report is under heavy reconstruction. Due to it not being updated this morning I have made an illustration of information flow in my solution.

Using information based on stumbling results we are able to get an estimate on our position. Our own server program calculate an improved position based on records of the surrounding roads. After the improved position is calculated we request maps from the Arealis server. The correct map with positioning information is sent to the client.

Arealis map service: http://www.statkart.no/IPS/?module=Articles;action=ArticleFolder.publicOpenFolder;ID=1054

Telemathica Institute: http://client338.lab.telin.nl:8080/wasp/index.html

Ostfold Community College: http://www.hiof.no/index.php?ID=3&lang=eng

Report

hiof — Wed, 15 Mar 2006 09:58:24 +0000

Added to the report:

Design issues.

Design suggestions.

New format for the entire report.

Report

Location aware systems you say?

hiof — Tue, 07 Mar 2006 11:01:06 +0000

Take a look at these links for a better look at what context aware systems are:

PlaceLab describes their system:

http://placelab.org/publications/pubs/pervasive-placelab-2005-final.pdf

What would be really beneficial is information regarding road overlay in conjunction with cellinfo. SVG is a brilliant tool for mapping roads on mobile phone without using to many resources.

http://www.svgopen.org/2004/papers/WebServicesForSVGTinyMapsOnMobile/

If we allready have access to road-maps or road information that is insupportive to SVG, what then? Well we need to find a way to link cell info to a fixed position which again can be transfused into a position on a road. So what we are looking at primarly is a two-step solution. Firstly we need to get cell info signals linked to an area and secondly we need to link that area to a roadposition.

There are ways in which one can further develop this. For instance the roads may be given attributes. That is, along a certain road one has an estimated speed of travel and direction. But looking into that at this point is a little too early.

More to follow…

Project Description

hiof — Tue, 07 Mar 2006 08:54:43 +0000

Here follows a description of our project, the problem and methodology.

http://www.ia-stud.hiof.no/~johnesim/rss/GPSkiller.pdf

Stay alert for even more Blog updates this day.