MmWave M2M Networks: Improving Delay Performance of Relaying

Abstract

For future wireless networks, applications such as Industrial Internet of Things are strictly delay-sensitive. Meanwhile, millimeter-wave (mmWave) communication is a promising means to provide ultra-high data rate and ultra-low latency services to massive number of devices. In order to minimize uplink end-to-end delay in such machine-to-machine (M2M) mmWave communications, we investigate buffer-aided multi-hop relaying networks and formulate the problem as a multi-tier queueing system. We propose a Minimum-Delay relaying scheme, and by leveraging stochastic geometry, we present a tractable analytical framework to investigate the signal-to-interference-plus-noise-ratio (SINR) distribution of devices at each-tier, thereby computing the expected delay and delay outage probabilities using Lagrange optimization. A state-of-art max-SINR relaying scheme is analyzed for comparison, and the performance of Minimum-Delay relaying in 3-tier architecture is further analyzed. The derived average delay and delay outage probability are validated through simulations based on multiple cells in a dense urban scenario. Numerical results show that the proposed Minimum-Delay relaying scheme achieves significant lower average end-to-end delay than direct association or the max-SINR relaying scheme. Furthermore, results for Jain's fairness and spectral efficiency reveal that the Minimum-Delay relaying scheme has even greater performance improvement under high traffic loads.