The algorithm for predicting the RSSI is written in Matlab,
and it operates based on the data in the log file. Therefore, the
current implementation of the algorithm does not operate in
real time3. It divides the data in the log file into two parts. The
first part is used to determine the parameters of the clusters,
Markovian model, and MAP detector. We refer to the first
part as the training data. The second part is used to validate
the RSSI predictions, i.e., the RSSI predictions are compared
to the RSSI measurements in the second part in order to
determine the error and duration of the predictions. We refer
to the second part as the validating data