Rate Matching Pattern Estimation

Abstract

Rate matching in 3GPP standard is a process of getting the right amount of bits out of full-rate channel encoded bits, either by repetition or by puncturing, at the transmitter. The rate matching parameters may not be available at the receiver in various non-cooperative communication contexts, and the estimation of the rate matching pattern is required before any channel decoding attempt. A generic rate matching pattern estimation problem is formally formulated for repetitive rate matching. A set of ML estimators are derived based on the conditional probability of the observed bit-repetition index over a perfect channel as well as over BSC channel. Their performances are explored through simulations. It was found that the detection error rate of the proposed ML estimator was relatively high when very few bits are repeated, but was reduced by our ML estimator exploiting multiple observations.