Pattern Division Random Access (PDRA) for M2M Communications With Massive MIMO Systems

Abstract

In this work, we introduce the pattern-domain pilot design paradigm based on a superposition of orthogonal-building-blocks with significantly larger contention space to enhance the massive machine-type communications (mMTC) random access (RA) performance in massive multiple-input multiple-output (MIMO) systems. Specifically, the pattern-domain pilot is constructed based on the superposition of L cyclically-shifted Zadoff-Chu (ZC) sequences. The pattern-domain pilots exhibit zero correlation values between non-colliding patterns from the same root and low correlation values between patterns from different roots. The increased contention space, i.e., from N to C_N^L, where C_N^L denotes the number of all L-combinations of a set N, and low correlation values lead to a significantly lower pilot collision probability without compromising excessively on channel estimation performance for mMTC RA in massive MIMO systems. We present the framework and analysis of the RA success probability of the pattern-domain based scheme with massive MIMO systems. Numerical results demonstrate that the proposed pattern division random access (PDRA) scheme achieves an appreciable performance gain over the conventional one, while preserving the existing physical layer virtually unchanged. The extension of the superposition of orthogonal-building-blocks scheme to superposition of quasi-orthogonal-building- blocks is straightforward.