Algorithm 4 Simulated tracking

1:procedure Tracking($profile$, $RSSI$, $filter$, $freq$)   2:      Group all readings by frequency                                 ▹ 3 s, 5 s, or 10 s   3:      for each frequency group i do   4:            if Group i not Empty then   5:                  Group $RSS\left[i\right]$ by $Beacon\text{\_}id$   6:                  for each Beacon group j do   7:                        if $filter==TRUE$ then   8:                             $RSSI$ = MovingAvg($RSSI$, 2)   9:                        $fp$[j] = getCurrentFingerprint($RSSI$) 10:                  $prediction$[i][‘location’] = nearestNeighbor($fp$, $profile$) 11:                  $prediction$[i][‘true_loc’] = round($group$[‘true_loc’].mean()) 12:      return $prediction$