The physical and engineering requirements of scalable, decentralised, distributed, large-scale MIMO

Abstract

In recent papers [1], [2] we introduced a novel approach to the deployment of wireless infrastructure by scalable, distributed, decentralised MIMO. SDD-MIMO is a novel wireless network architecture specifically intended for wireless broadband infrastructure whose aggregate capacity and range scale as service antenna nodes are added to the network. We previously considered the physical and engineering constraints of such a system [1] and produced a compelling deployment simulation of wireless broadband for an over-the-horizon transoceanic troposcatter link of 155 kms extent [2]. The key issue is how to achieve scalable broadband over a long range. In this paper we discuss the engineering and physical constraints from a practical perspective. We take a close look at how the beam-forming process works. We propose a novel definition of capacity expressed in terms of the correlation between the received and transmitted signals. We demonstrate through Monte-Carlo simulations that this capacity agrees well with that previously derived for cellfree massive MIMO [3]. Moreover, the Monte-Carlo simulation allows us to explore the fluctuations in capacity versus the size of the network. We show that the capacity per user converges to the theoretical value as the network increases in size. For the SDD-MIMO network work to function, one must also solve the problem of global synchronisation of radio carrier frequencies and information symbols. We show that a well known low-cost GPS technique may be used to achieve this goal. Finally we take a look at what is probably the main constraint of distributed MIMO: the communication of beam-forming information across the Internet back-haul. We present some preliminary results demonstrating that adequate throughputs and latencies can be expected