This post cover explanation about how wireless localization technique work, and we will give some example of common localization technique in wireless system.
How Most Localization Technique Work?
In order for wireless localization technique to works we need to have a reference points which location is known. The best practice is follow n+1 reference points for n dimension coordinate. So when we want to predict location in 2-dimension coordinate (x,y coordinate) we need at least 3 reference points and if we want to predict location in 3-dimension coordinate (x,y,z coordinate) we must have at least 4 reference points. These reference points should not located inline one and another for best location accuracy.
Illustration above describes how localization technique work, We want to find out x and y location of a target. The first thing we should do is to make each reference points (RPs) receive signal from the target. Then using signal from target we determine the how far is the target from each reference point (i.e : R1,R2,R3). In each RP point of view, a target can be located at any position in the area of a circle with radius equal to how far target form RP. This expression can be written as equation below.
Now, why do we need 3 RP for predicting location in 2-dimension coordinate? and why do we need RPs no to be located inline one and anothers?
If we have only 2 RPs like shown in illustration above. There are 2 intersections of the these 2 circles 1 is real target position and the other is not. If we take estimation from these 2 position the result is more likely in between these 2 position and this is not the correct position of the target. Thiis ambiguous solution also happens if the RPs are located inline to each other. So we need to that perquisite requirement to get best result.
Common Localization Technique
The differences in many localization technique are in how it to measure distance between in RP and target ( R) and in the method solving ambiguous solution. In this post we only describe method how to find R.
In this method we measure R by finding how long does it take for the signal to reach RPs which we call it signal travelling time(t). we can find R by multiplying t with electromagnet travelling speed(c).
Its look easy. However, to find t we need to have time synchronization between target (A) and RP (B), See illustration above. The drawback of this method is it is difficult to synchronize target and RP. If there is 1 second miss in calculation of t, because of c=3x10^8m/s. It can cause predicted R 3x10^8 meter miss from the real target position.
2. Time Difference of Arrival Method(TDOA)
To overcome ToA synchronization problem TDOA was made. In the picture above 1 of the RB (in our examplew e denote it as Rx2) acting as time reference. Most of RP (e.g Base Station and WiFi access point) normally is synchronize to each other. So, upon receiving signal from target (Tx) we obtain distance difference (R1-R2 or R2-R1) using 1 of the RP as time reference e.g RP2 as reference.
Then after solving this problem We can get the target location.
3. Receive Signal Strength (RSS)
In this method we calculate R by using path loss. We use strength of received signal in RPs given the Tx transmitting power.
Other methods are combination between these 3 methods to achieve higher accuracy.
Reference
- Krishnakumar, Krishnan - 2005 - The Theory and Practice of Signal
Strength-Based Location Estimation
- Ghannouchi, Wang, Tiwari - 2012 - Accurate Wireless Indoor Position Estimation
by Using Hybrid TDOA RSS Algorithm
- http://www.exploregate.com/Video.aspx?video_id=59#.Uaa77UAr0Wx
.
No comments:
Post a Comment