Distributed Pilot Design for Massive Connectivity in Cellular Networks.

Xihan Chen,An Liu,Wei Yu,Hei Victor Cheng,Kaiming Shen,Min-Jian Zhao

DOI: https://doi.org/10.1109/globecom38437.2019.9013978

2019-01-01

Abstract:Massive connectivity is regarded as a key requirement for future networks to support new communication paradigms, where the human-type communications coexist with machine-type communications. Owing to the limited coherence time but the huge number of potential devices, it is impossible to allocate mutually orthogonal pilot sequence for all potential devices, which may impose severe interference on the device activity detection and channel estimation. Existing nonorthogonal pilot design methods for conventional cellular network are not suitable for the massive connectivity regime. To overcome this challenge, we first formulate the pilot sequences design as an optimization problem to minimize the average mean square error (MSE) of channel estimation under the individual power constraint. The proposed optimization problem is nonconvex and highly coupled. By exploiting some approximation techniques, we convert the problem into a more tractable form and subsequently develop a distributed algorithm based on the matrix fractional programming (FP) and the alternating direction method of multi-pliers (ADMM) methods. Simulations validates that the proposed scheme not only achieves significant gains in channel estimation over state-of-the-art baseline schemes, but also improves the device activity detection performance.