Outage Analysis of Relay-Assisted mmWave Cellular Systems Employing JSDM

Abstract

In this paper, the outage performance of relay-assisted millimeter wave (mmWave) cellular systems employing joint spatial division and multiplexing (JSDM) is investigated. It is assumed that the macro base station (BS) equipped with a large number of antennas serves the single antenna pico BS (as a relay) and users simultaneously, and that the pico BS is located at the edge of the macro cell.Theoretical analysis of the signal-to-interference-plus-noise ratio (SINR) outage probability of each user is first obtained. The cell SINR outage probability is then derived. Under the noise-limited assumption, simplified closed-form expressions of the outage probability are given as well. Simulation results demonstrating the performance improvement due to the relay introduced by the pico BS are provided.Overall, the impact of deploying pico BS as a relay in the mmWave cellular systems is characterized.The authors are with the offer an order of magnitude increase in achievable rate compared with current cellular systems and play an important role in future cellular networks [5].Despite the great potential of high data rates with mmWave systems due to the richness in bandwidth, the transmission range is generally limited due to the high free-space path loss and poor penetration in high frequency [3]. Thus, it is highly possible that some users may fall in outage in mmWave systems, and hence the coverage analysis of mmWave systems have attracted much attention recently (see e.g., [8]-[14] and references therein). For instance, the authors have proposed a stochastic geometry framework for analyzing the coverage and rate of mmWave systems assuming pencil beams for the users in [8]. They have shown that there is an optimal relative base station (BS) density for the signal-to-interference-plus-noise ratio (SINR) and rate performance beyond which the performance doesnot improve in a dense mmWave network. An analytical framework which computes coverage probabilities and rate of mmWave cellular networks has been proposed in [9], where path-loss and blockage models based on empirical data for mmWave propagation have been taken into account. Relying on the noiselimited assumption for modeling mmWave cellular systems, simple and closed-form formulas for computing the coverage probability and the average rate have been obtained. In [10], the authors have made a further step and investigated the coverage in heterogeneous mmWave networks with homogeneous Poisson point process (PPP) models of the BS and user distributions, where beamforming with pencil beam at the BSs is assumed. It has been shown that biasing towards the small cells in user association can improve both the coverage probability and the rate. Also, the authors have assumed homogeneous PPP model for the macro cells while Poisson hole process (PHP) model for small cells and studied the coverage of the proposed non-uniform mmWave heterogeneous cellular network in [11]. They have also shown that there exists an optimal density of the small cells to achieve the best coverage pro...