A Novel Radio Frequency Identification Collision Resolution Method Based on Statistical Learning

In the application scenarios of radio frequency identification technology, there are many situations where a large number labels respond to reader at same time, resulting in not being able be identified for long time. order address label collision problem identification, this paper studies impact statistical learning method on resolution and decoding labels, proposes novel clustering using maximum posteriori probability estimation based Monte-Carlo. Unlike traditional algorithms, proposed does require prior knowledge clusters need constantly iterate. addition, has low complexity ensures both accuracy robustness while quickly finding cluster centroids. Finally, performance is evaluated simulation experiment field experiment, resolved signals decoded matched filter phase jump. Overall, effectiveness our demonstrated through comparisons with different metrics benchmark methods, including bit error rate, efficiency, throughput, error, time complexity.