How Can I Assign Probabilities Based on Latency Proximity?

[optimistbs]

Hello All,
I calculated an average latency value let say 3 units
then after observing the network I found that different nodes have the fallowing different latency i.e. 4, 5, 6, 7, 10 units.

now I need to assign probability to each latency considering that the value which is more closer to average latency 3 has high probability and the more far has low probability, so in this case 4 has high probability and 10 has low probability, so my question is how can I assign probabilities which should be logical and practicle.

kind regards

 

[Stephen Tashi]

now I need to assign probability to each latency

When applying probability theory, probabilities are assigned to events. It isn't clear what event you are talking about.

For example, if we say we have a probability distribution for the weights of 20 year old females, the probability "of a weight" refers to the probability of that weight being the weight of a 20 year old female chosen at random from the population.

If this is some sort of textbook problem then one guess at the meaning of the question is that we are to assume the population of nodes has a normal distribution of mean 3. Then we estimate the standard deviation of this normal distribution from the given latencies. Then we apply the PDF of the estimated distribution to say something about the probability of a randomly selected node. (Strictly speaking, the value of a continuous PDF at a single point (like 4) does not give the probability of exactly realizing that value in a randomly drawn sample.)

If this is a practical problem, you'll have to do a better job of explaining it.

 

[optimistbs]

Thanks for the reply.

actually these latency comes randomly , there is no mean.
The problem is not book related , it is part of my research.
to explain it more, I would say.

I want to launch probabilistic attack on node to find who is the receiver of the message based on already calculated average latency, for this reason i calculate the latency i.e. 4 unit.

now if attacker observe the network( different node) and he find latency of 6 nodes as 4, 5, 6, 10, 11, 2. now attacker want to assign probabilities to each latency node. I want to assign higher probability to those latency which is either exactly equal to average value 4 or closer to it and low probability to latency which is either more larger or smaller then average value.

I hope you understand my problem.

kind regards

 

[Stephen Tashi]

Your description is still not clear.

See if this describes what you are doing:

You send a test message. The nodes receive the test message and send back an acknowledgment. For the test message, you can identify the nodes that replied. You measure the time between when the test message was sent and when the acknowledgment from the node was received. This is the "latency" of the node.

Later you send another message and get replies that do not identify which node replied. You are trying to identify which node replied based on time that elapsed between when you sent a message and when you got the acknowledgment.