How to find the region that is not covered by any node in a wireless sensor network?

How to find the region that is not covered by any node in a wireless sensor network?

Answer 1

[Walter Roberson]

With difficulty. You can find a beginning to a solution at " Applications of Geometric Algorithms to Reduce Interference in Wireless Mesh Network" by Hung-Chin Jang .

However, a fuller solution would require materials modeling to deduce the composition of barrier objects by (for example) studying the reduction in strength of signals as signals of varying frequency pass through the object. "Okay, so I deduce there is a tree over there. But is it just reducing the signal or is there an area where it is blocking the signal?"

You pretty much need to use the same techniques as MIT researchers recently used to locate users within centimeters by measuring interference, except that you would take the information and use it to figure out the places the signal cannot reach. It needs new hardware.

... and none of this is a MATLAB question.

This is what happens when you ask other people to do your research for you: the other people might point out why the problem is too hard to solve in any reasonable time.

[poonam tailor]

Sir i have the idea bt the detail of the question is to co/pare each and every pixal area of the field with the each node area as its range such that the not covered area will find out. The code i used for it will be too lengthy and complex with too much running time so i need a acurate and fast method of programming for this problem. In the code is I have a wireless sensor network 100*100 with 100 randomly rendered nodes, each node have its maximum range, maxr, my question is to find the black hole region where no any node have its effect to sense?

*I have applied 2 to 4 logic to solve it

1. I run the for loop j from field length to for loop k to width with n nodes ( 100*100 pixals) * (100 nodes) = 1000000 iterations = lengthy loop[03/05 9:22 pm] Poonam ankur sagar:
2. Using set operation that is limit me on only union of all node range maxr#20, Want to forward in this
3. Using recursion to call one node to another

[Walter Roberson]

There is no accurate and fast method of programming that.

For example, I have a wireless access point beside me in my house. I have a garage. The garage has concrete walls. Concrete blocks wireless signals fairly thoroughly. So there is no wireless access in my garage, right? Wrong! Because over to my right hand side is a wooden fence, which reflects a portion of the wireless signal towards the garage, so there is some bounced signal that hits the wood-core human door on the garage. Wood blocks signals less than concrete but the wooden fence blocks (bounces) some of it, and then some of that bounced signal gets through the wooden door into the concrete garage. So there are parts of the garage that have wireless signal. And there is a different part of the garage where the signal travels straight in through the glass window during the fall and winter and spring, but not in the summer because in the summer the tree between the access point and the window is in full leaf and that blocks the signal.

If you are keeping track, you will see that to model all of this you need to know the signal strength, the geometry, the attenuation coefficient of concrete, the signal reflection coefficient of wooden fences, the signal attention coefficient of wood-core doors, the signal attention coefficient of glass, the time of year, and the weather (did I forget to mention that when we get blizzards of snow in the middle of winter that the fast-driving snow blocks the signal?)

You cannot model this quickly and accurately. You can model it slowly and accurately or you can model it quickly and so inaccurately as to be mostly useless for real-world scenarios.