A scale-independent way for differential estimation in dynamic radio frequency identification systems

摘要

Cardinality estimation in radio frequency identification systems has been applied to estimate the population of tags in many applications. However, it is more meaningful to estimate the number of tags moved in and out in a dynamic radio frequency identification system, which is called differential estimation problem. Zero differential estimator is a newly proposed algorithm to solve this problem. However, the time slots consumed by zero differential estimator are relevant to the system scale under the accuracy constraint. This will result in low time efficiency when the system scale is very large. In this paper, we thus propose a scale-independent algorithm for differential estimation called zero-one differential estimator. The numbers of tags moved in and out are estimated from the idle slots in two consecutive frames. We can prove that the time slots consumed in our proposed algorithm are not relevant to the system scale under the accuracy constraint. Moreover, we conduct abundant simulations to evaluate the performance of the proposed approach. The simulation results show that the estimation error grows little as the system scale grows. It indicates that our proposed algorithm is indeed scale-independent. Copyright © 2015 John Wiley & Sons, Ltd.