Keywords : Indoor radio; Optimisation ;Search problems; Network planning; Wireless LAN; Indoor positioning system; Optimization model; WLAN infrastructure; Network deployment; Positioning error reduction; Mono-objective algorithm; Indoor positioning optimization; Quality of service; Accuracy; Connectivity; Communication quality demand
Master IoT : 2019 – 2020
Students :
GADEGBE Patrice working on Software Engineering
GOUDIABY Fatou working on Network and Telecommunication
Tutor :
BAALA Oumaya
University of Technology of Belfort Montbéliard
Global view
The objective of this project is to study the positioning quality indicators which currently exist in order to propose an optimization of current indoor localization systems. This study will focus on indoor localisation systems that are based on WLAN.
Today, several indoor positioning solutions are proposed in order to solve the problem of positioning and indoor navigation. Among these solutions there is an indoor positioning solution which is based on WLAN. This solution is by far the most popular in view of the advantages they offer compared to other interior positioning systems. This type of positioning system does not require the installation of additional infrastructure if a WLAN infrastructure installation already exists. In addition, most of today’s smartphones natively support WLAN technologies. Those smartphones can easily obtain and read the Received Signal Strength (RSS) of an Access Point (AP). Furthermore this solution attract researcher who done much substantial work on this field.
1 – How WLAN-based indoor positioning system work ?
A common approach in the WLAN-based indoor positioning system is to use a map of the measured radio signal strength of nearby access points.
The work area is mesh , Mesh of the area is called a Marking Position (MP), every Marking Position have a fingerprint based on the RSS of all the visible access points.
Every fingerprint is associate to a location of the area with is linked to a Marking Position.
The client app send the location request to a server which return the location of the user according to his fingerprint database.
In order to overcome the this training process to make the map of the fingerprint of each Marking Position and according to some previous work we will use a propagation model based method to get the propagation of the radio signal of each access point.
 |
| Figure 1: WLAN-based indoor positioning system |
2 – WLAN-based indoor positioning system constraints
The quality of indoor positioning based on WLAN is subject to several external constraints such as: the architecture of the building, the architecture of the network, and the propagation characteristics of the radio signal.
In this approach there is the positioning error indicator which researcher will try to reduce in order to have a good accuracy of the positioning system. There are many factors which influence the positioning error. A first factor which is almost impossible to predict comes from the environment changements such as Access Point Orientation, the number of people inside the area, the architecture of the building, those factor create an random positioning error. The other and second factor is due to a inconvenience of the technique of fingerprint.
There can be several marking point which will have the same RSS fingerprint these phenomena make the positioning based on fingerprinting technique become invalid.
A way in propagation models to reduce this error is to change the position of the access points or increase the number of access points this will decrease the number of Marking Position which make this effect.
At a first glance this approach to reduce the positioning error sound good but neglect the fact that the main objective of a wlan in a space is to provides the necessary connectivity to users. In fact by increasing the number of access point we increase the localisation system accuracy too but the quality of communication will decrease due to frequency interference and the installation cost will increase too.
 |
| Figure 2: WLAN-based indoor positioning system constraint |
3 – The problematic
Nowaday, indoor localization systems incorporate positioning indicators to improve the accuracy of the localization service provided. But with these methods used there was a lack of accuracy and frequent errors due to the neglect of network connectivity quality of service.
So the problematic is how to get a network configuration which will guarantee the required connectivity for user and reduce the positioning error.
To achieve this target we will see the problem as an optimization problem and this optimization will be done during the wlan network planning process.
The advancement of new technologies and research work will allow us to better understand and improve the existing indoor system to obtain better results.
4 – Existing Works
Our work is based on the work of : Y. Zheng, O. Baala and A. Caminada, « Efficient design of indoor positioning systems based on optimization model, » 2010.
In addition with this excellent research work we will try to improve the research by doing the second approach of the optimization which relies on a multi objective search strategy and looks for a set of pareto-optimal solutions, each solution representing a different trade-off between the objectives that is involved. And take into account the financial requirements.
5 – Quality indicators
So we will work on two indicators which will evaluate the actual network QoS and the positioning error. These two indicators will be calculated at each Marking Point.
5.1 – QoS indicator
Cause the interferences are not uniform , the perceived QoS of an user can change inside the same Access Point coverage.
The indicator we will use to measure the interference is the Signal To Interference Plus Noise Ratio (SINR).
5.2 – Positioning error indicator
We will use the Refined Specific Error Ratio (RSER)
6 – Our work
our work is to write a program which take in account the topology of the building, access points parameters as : the allocated frequencies , the emission power and the azimuth; the installation cost, the number of maximum access point to install and the program will output the best configuration for network deployment.
 |
| Figure 3: Global view of the project |
It therefore seems necessary to propose a solution in order to find a compromise between communication and the quality of positioning provided by the system.