Optimization methods for receiver location

The following problem arose in a project run by Schlumberger Consulting for Illinois Power: Given a geographic region, place the minimum number of receivers (ampliiers) on existing utility poles so that all the region's wireless meters can transmit their reading to at least one. Due to the growing popularity of wireless meter reading, the size of the geographic regions involved, and of the amount of necessary hardware, the receiver location problem is expected to become more and more important in the following years. We use simple heuristics, preprocessing, an integer programming formulation, and commercial software to tackle two real life instances. On the rst one, we obtained the optimal solution within 2 minutes of computing time, as opposed to the previously used manual approach taking about 200 man-hours. On the second, large-scale instance we found a solution within 4% of the optimal, within an hour.