Skip to content

Discovery and routing

Not completed yet. Work in progress....

The problem of routing

Let's consider a network like the one in the following figure:

simple network

We want to send a receive data between Base ans Node4. The two nodes are not in RLOS (Radio Line of Sight) that means that radio signals from Base can reach the Node4 and the other way around.

To reach the target we have to use intermediate nodes a repeaters of the radio signals broadcasted from one Node to reach the Other.

As an example the signal from Base can be pick up and repeated from Node2 then Node3 and eventually reach Node4. Then Node4 can follow the path in reverse to reach the node Base.

So the new problem is how to select a optimal sequence of node to act as repeaters for the communication between Base and Node4.

Let's take the previous network and assign a cost to each edge. The cost maybe directly correlated to the quality of the radio link channel between two nodes.

Using the RSSI value provided byt the device is normalized between 0.0 (min RSSI) and 1.0 (max RSSI) the normalized RSSI is rectified by the following function:

$cost=1.0-((log_2(rssi+0.2)+1.6)*0.56)$

logarithmic cost