Random Coding Error Exponent for the Bee-Identification Problem

Anshoo Tandon, Vincent Y. F. Tan, Lav R. Varshney

Abstract: Consider the problem of identifying a massive number of bees, uniquely labeled with barcodes, using noisy measurements. We introduce this “bee-identification problem characterize the random coding exponent, and derive efficiently computable bounds for this exponent. We demonstrate that joint decoding of barcodes has much better exponent than separate decoding followed by permutation inference.