ENERGY-EFFICIENT MIMO/MASSIVE MIMO COMMUNICATION WITH PRACTICAL POWER AMPLIFIERS FOR INTELLIGENT TRANSPORTATION SYSTEMS

Abstract

An Intelligent Transportation System (ITS) aims to solve transportation problems with information and communication technologies (ICT). International Telecommunication Union (ITU) study group, SG5, focused on sustainable smart cities and mentioned ICT to remedy economic/environmental problems in urban localities. The concept of a smart city is to have an underlying digital infrastructure that includes a network of data-collecting sensors and devices, comprehensive broadband, and wireless systems with cloud platforms for sharing and storing data. The emergence of 5G and beyond technologies may result in more complex and interconnected networks. These technologies will help in facilitating the evolution of highly connected modes of transport, which may exchange information with each other and in the same way with other elements such as the Internet, roadside infrastructure, pedestrians, and passengers in a smart city. Widely deployed Long Term Evolution (LTE) technology has emerged as the most promising ITS technology to fulfill the ever-increasing data rates, a superior quality of service (QoS), and enhanced network capacity demands. To respond to these demands with limited spectrum availability, multiple-input multiple-output (MIMO) and Massive MIMO technologies may introduce significant improvements. Massive MIMO architecture is considerably better than the conventional MIMO. It is an efficient means to improve spectrum efficiency and energy efficiency (EE) for ITS communication environments. However, MIMO configuration requires more hardware components. It needs radio frequency power amplifiers (PAs), couplers, and antennas on the transmitter side, whereas the receiver side needs additional antennas, couplers, and low-noise amplifiers (LNAs). Unfortunately, the practical PAs are inherently nonlinear and consume a significant share of power within the base station (BS). Besides consuming substantial power, the PA wastes about 80-90% of the total power as heat dissipation, requiring cooling facilities, thus adding even more to the energy costs. As a result, the total efficiency of a conventional practical PA ranges from 5-20% only, depending on the semiconductor material and manufacturing process selection. Now-a-days, the increasing emission of greenhouse gases in the atmosphere is causing global warming. The global energy consumption of ICT industries is about three percent of the worldwide energy consumption, which is equivalent to about two percent of the carbon dioxide (CO2) emissions globally. Wireless communication networks contribute about twelve percent of ICT emissions, whereby the most considerable fraction of energy consumption is due to BSs. Moreover, ITS users worldwide need additional BSs along newer transportation routes to support high mobility lifestyle ‘always get connected’ and avail various high-speed and reliable broadband services. The added BSs may account for a further increase in greenhouse gases. Therefore, using highly efficient and linear PA for each antenna can significantly optimize the operational costs for the mobile network operators and the power consumption within the complete communication system. Digital predistortion (DPD) is a low-cost and popular technique that compensates for the practical PA's inherent nonlinearity and makes PA operate at relatively high efficiency. The DPD introduces a nonlinear digital model before the PA to predistort the transmitted signal. The use of DPD before the practical PA helps in mitigating the distortion produced by the PA. The (DPD+PA) combination makes reliable and green communication possible with linearity perspective, providing better user experience and power efficiency perspective, providing less power waste. This energy-saving supports economically by reducing energy costs and advances practical usage by improving battery life in mobile devices and lowering the heat dissipation at BSs. Today, the availability of a plethora of consumer services gives rise to the enormous growth in wireless traffic and information processing. This data traffic growth intensifies the energy consumption in the communication network and ensures that modern wireless communication may not sustain without the provision of green radio technologies. The Green Radio research has recently started focusing on the EE study to jointly reduce power consumption and maximize the EE of radio access networks (RAN) for efficient wireless communication. MIMO-DPD combination can efficiently address the crucial issue of high energy dissipation within the BSs. Additionally, engineering innovations like Doherty PAs significantly contribute to the realization of green access networks. There is extensive literature on the usage of DPD to improve the transmitter's performance by compensating transmitter nonlinearities. However, the established literature focuses on analyzing the transmitted signal quality, which does not quantify the capability of the complete communication system. The existing literature reports the impact of linearization using DPD in terms of normalized mean square error (NMSE) and adjacent channel power ratio (ACPR). NMSE and ACPR are transmitter quality parameters that do not give any inkling for the energy saved/ throughput achieved for complete communication. In contrast to the existing research, our work quantifies the effect of PA nonlinearity and efficiency enhancement using DPD for the complete communication system, including transmitter and receiver. In addition, the thesis work includes an end-to-end mathematical analysis for evaluating a practical High-speed Railways (HSR) communication system using MIMO-DPD architecture. This analysis features the antenna effects on Block Error Rate (BLER) and throughput in terms of a signal-to-noise power ratio (SNR) over the dynamic multipath channel by adopting single-input single-output (SISO), multiple-input single-output (MISO), single-input multiple-output (SIMO), and MIMO configurations. Our work paves the way for achieving Green communications in a high mobility scenario. This research work comprises a three-way analysis covering the effect of mobility on various antenna configurations, the linearization effect at various mobilities, and the impact of different MIMO configurations. Our work reports improved BLER with MIMO-DPD arrangement, whereas the BLER and throughput performances degrade with increasing train speeds. There is a requirement for additional signal power if the nonlinearity is not compensated earlier. The energy required to drive PAs for reliable broadband communications at higher speeds is more, even with DPD. Our proposed model is not only simple and sufficiently accurate but provides an end-to-end performance in the presence of both DPD and PA. This thesis work extracts simulation parameters from existing works to authentically analyze the massive MIMO communication system. The EE optimization analysis utilizes the zero forcing (ZF) linear processing scheme for a downlink Massive MIMO system. The simulations result in choosing the optimal number of BS antennas, users, and PA power for maximizing the EE. The thesis work also finds a global optimum representing optimum EE in a given coverage area. The numerical results reveal that the Doherty PAs are the perfect choice to realize energy-efficient 5G and beyond communications. Our work represents diverse transportation scenarios, viz. pedestrian, highway, and the HSR corresponding to the femto, pico, and microcells. A brief analysis to evaluate the performance of two imperative second-order statistical performance indicators, level crossing rate (LCR) and average fade duration (AFD), is also done by carrying out the effects of transmitter speed, receiver speed, and the Rician factor on these indicators.