the distributed
top-k query processing [75] computes, in a quick and
efficient manner, the subset of most relevant answers
rather than the all answers. This prevents the transmission
of irrelevant answers and minimizes the power cost of
retrieving the huge amount of values