Massive MIMO Channel Measurement and Characterization for Manufacturing Scenario

One of the main differences between 5G and previous generations of cellular networks is that 5G supports not only mobile broadband enhancement, but also unprecedented reliability and very low latencies. This is beneficial to new applications in manufacturing scenario. In order to design a feasible wireless solution for manufacturing scenario, the particular characteristics of manufacturing environments need to be considered. In this paper, the massive MIMO channel measurement in factory was conducted, and the channel propagation in the spatial and frequency domains were analyzed and compared with a general indoor scenarios. Due to more metal reflections and big machine deployment, it was found that there were a lot of dense multipath so that the delay and angular spreads are larger than the ones in office scenario. Moreover, the effect of particular channel characteristics on communication system in a manufacturing scenario, was investigated in terms of CP (Cyclic Prefix) length and MIMO rank.

I. Introduction

5G enables communication with unprecedented reliability, very low latencies, and massive connectivity, many applications in different vertical domains, including the automotive, healthcare, agriculture, and manufacturing, have been thus considered in 5G networks [1]. For the industrial production, the fourth stage of the industrial revolution aims (c) at improving the flexibility, versatility and efficiency of future smart factories. Hence, 5G wireless communication is a promising solution to achieve such requirements. As the foundation of the communication system design and evaluation, the reasonable wireless channel characteristics and model are needed for each environment. Especially for industry scenario, there are many metallic devices and walls so that it is quite different from a general indoor scenario. Some large-scale and small-scale channel measurements have been conducted in [2]–[12]. In [2], not only path loss and delay spread are analyzed, but also the electromagnetic interference caused by electric and machinery is involved. In [3] and [4], the path loss measurement is conducted, and the path-loss model is derived in manufacturing scenario. In [5], both of the path loss and temporal channel fading characteristics in manufacturing scenario are compared from different frequency bands. In [6], the path loss models in industrial and office environments are compared for different frequency. In [7]–[10], the temporal and frequency domains’ channel variations are measured and analyzed with the mechanical robot arm’s movement in factory. The multi-frequency from 2G to 6G channel measurement and analysis are conducted in [11] and [12]. Through the literatures results in [2]–[12], it can be seen that the factory has more dense multipath than traditional indoor scenario, and all of the literatures mainly focus on the path-loss model, frequency, and temporal characteristics in industrial environments. However, there is a lack of the massive MIMO channel analysis in spatial domain, which are cruel for the transmission scheme design. Fig. 1.

(a) Photo of automobile assemply factory. (b) Sketch of measurment setup in factory. (c) Sketch of measurment setup in office