A method and system for resolving contentions among passive identification
tags. A reader may divide a possible range of tag values into a number of
smaller ranges. For each smaller range, the reader may then poll the tags
to determine which tags have tag values that fall within the smaller
range. If more than one tag responds simultaneously, then the reader may
further divide the smaller range into even smaller ranges and poll the
tags with respect to those even smaller ranges. Once the range partition
gets small enough, the reader may then opt to poll the tags for a match of
each possible value in the range, rather than partitioning further. This
arrangement enables the reader to efficiently identify tags, by quickly
eliminating from analysis any range of tag values to which no tags match
or only one tag matches.