Energy Efficiency Maximisation in Large-Scale MIMO Systems

Analysis of Energy Efficiency Maximisation in Large-Scale MIMO Systems

 

Introduction and Motivation

1.1 Background

The development of smart terminals and their application, the need for multimedia services rapidly increases lately [1]. The capacity of wireless the Quality of Service necessities of mobile applications of wireless communication networks is increasing exponentially [1].

Bandwidth Efficiency is typically one of the important metrics to Systems [1], [1]. Energy Efficiency become a metric for assessing the performances of wireless communications systems with some BE restrictions [1] – [1].

1.2 Research Motivations

An accurate modelling of the total power consumption is the primary of (BS) antennas and number of active (UEs) for LS-MIMO systems [15].

1.3 Research Aim and Objectives

The research objectives which are briefly explained and summarized as below:

1. To compare the performance of the proposed uplink and downlink of LS-MIMO systems for ZF, MRT/MRC, and MMSE processing schemes at BS.
2. To implement a new refined model of the total power consumption for LS-MIMO system.
3. To derive closed-form EE-maximal values of number of (BS) antennas, number of active (UEs), and the transmit power using ZF processing in single-cell system and new refined model of the total power consumption when the other two are fixed.
4. To evaluate analytic results for ZF processing scheme with perfect CSI.
5. To measure numerical results for ZF, MRT/MRC, and MMSE processing schemes processing schemes with perfect CSI in a single-cell scenario.
6. To measure numerical results for ZR processing schemes with imperfect CSI, and in a multi-cell scenario.

1.4 Main Contributions

This thesis has contributions to knowledge in three research issues for LS-MIMO system, which are the new refined circuit power consumption model, energy efficiency maximisation with ZF processing scheme, and deployment of imperfect CSI case and symmetric multi-cell scenario. Those main contributions of this thesis are summarized and elaborated more detail as follows:

1. The circuit power consumption is the sum of the power consumed by different analogue components and digital signal processing. The new refined model of the total power explicitly described how the total power consumption depends non-linearly on number of number of UEs, number of BS antennas, and transmit power.
2. The closed-form EE expression under the assumption of ZF processing scheme is employed in the uplink and downlink for optimal number of UEs, number of BS antennas, and transmit power for a single-cell scenario with perfect CSI. This option is driven by analytic convenience and numerical results likewise which are close to optimal.
3. Analysis of imperfect CSI case and symmetric multi-cell scenarios deployment are extended using the same method above.  A New achievable rate derived for symmetric multi-cell scenarios with ZF processing.

1.5 Research Methodology

In the first stage of the research, literature review of past and current works on the area of MIMO, MU-MIMO, and LS-MIMO are extensively conducted to broaden the perspective on such areas of study. Furthermore, state of the art related to those addressed issues are deeply studied and intensively explored during this period.

Following the literature review phase, implementation starts with formulating the EE maximisation problem.  A new refined circuit power consumption model is proposed. All this then used to compute closed-form expression for the optimal number of UEs, number of BS antennas, and transmit power under the assumption of ZF processing scheme.

The testing stage starts with simulation. All the simulations were performed using Monte Carlo Simulation techniques in Matlab. Monte Carlo simulation can handle very complex and realistic. Monte Carlo Simulation were executed for all the investigated schemes with perfect CSI, for ZF with imperfect CSI, and in a multi-cell scenario

In the validation stage, numerical results are used to authenticate the theoretical analysis and make comparison amongst different processing schemes.

1.6 Thesis Structure

This thesis comprises of six chapters, where each chapter is inter- dependent.

Chapter 1 Introduction:

Chapter 2 LS-MIMO-An overview: This chapter presents an overview of the LS-MIMO concept.

Chapter 3 Literature Review- Energy Efficiency Maximisation in LS-MIMO:

Chapter 4 Techniques to Maximise Energy Efficiency: The simulation procedures will be explained in this chapter.

Chapter 5 Result and Analysis: This chapter describes description and evaluation for this investigation of LS-MIMO .

Chapter 6 Conclusion & Further Work: This chapter concludes the results of the implementations, and recommendation of developing revised model for LS-MIMO systems.

LS-MIMO – An Overview

2.1 Introduction to LS-MIMO

Wireless communication is one of the most successful technologies is one of the most successful technologies in modern years, given that an exponential growth rate in wireless traffic (known as Cooper’s law) [1]. This trend will certainly drive by; for example, augmented reality and internet-of-things [1].

Figure 2-1:[6]

2.2 Antenna configurations

Radio-Frequency (RF) circuit is usually connected to its physical antennas through an RF cable in a passive AA. A Remote Radio Unit (RRU) in with a Baseband Unit (BBU) has become a preferred configuration recently [1].

2.3 Channel Measurements

Realistic channel measurements have been carried out in in an effort to identify the main characteristics of LS-MIMO channels [15]

2.4 Channel Model

Three types of channel models have been used for evaluating the performance of wireless communications systems, namely the Correlation-Based Stochastic Model (CBSM), the Parametric Stochastic Model (PSM) and the Geometry- Based Stochastic Model (GBSM) in [1].

2.5 Processing Schemes

Precoding LS-MIMO is based on linear processing at the BS. BS has observation of the multiple access channels from the terminals [6]. The BS applies linear receive combining to discriminate the signal transmitted [6]. The simplest choice is maximum ratio (MR) combining by adding the signal components coherently. In [6], this result signal amplification proportional to.

Energy Efficiency Problem – Literature Review

3.1 System and Signal Model

The uplink and downlink of a single-cell multiuser MIMO system operating is considered over a bandwidth of B Hz [15].

3.2 Channel Model and Linear Processing

The M antennas at the BS are spaced apart such that the channel components between the BS antennas and the single-antenna UEs are uncorrelated [15]. The channel describes propagation channel between antenna at the BS and the UE. We assume small scale fading distribution [15].

3.3 Uplink

In [15], under the assumption of Gaussian, linear processing, and the perfect CSI, the achievable uplink rate of the th UE is

                                                                 (3.6)

the pre-log factor accounts for pilot overhead and is the

fraction of uplink transmission [15]. In addition,

                                 (3.7)

3.4 Downlink

A normalized precoding vector and the downlink signal to the kth is assigned a transmit power of . In [15], assuming Gaussian codebooks and perfect CSI the achievable downlink rate of the kth UE with linear processing is

                                                  (3.13)

3.5 Problem Statement

The EE of a communication system is measured in bit/Joule and the average total power consumption (in Watt = Joule/second) [15].

The total EE of the uplink and downlink is

                                                              (3.20)

Energy Efficiency Maximisation-Techniques

4.1 Realistic Circuit Power Consumption Model

The sum of the power consumed by different components and signal processing is the circuit consumption is [15]. A power consumption model is proposed

                                 (3.22)

4.2 Energy Efficiency Maximisation with ZF Processing

The EE maximisation problem is resolved under the assumption that ZF processing is employed. This solution is driven by analytic and the numerical results [15].

For ZF processing, Problem 1 reduces to

                                                              (3.30)

4.3 Extension to Imperfect CSI and Multi-Cell

The analysis is prolonged to single-cell scenarios with imperfect CSI. A new achievable rate is derived with ZF forcing processing. The achievable user rates in single-cell scenarios with imperfect CSI [15].

                                                                 (3.52)

 

Simulation Setup and Numerical Results

5.1 Simulation Setup

Simulations used to validate the system design guidelines under ZF processing and to make comparison with other processing schemes [15]. Numerical results provided under both perfect and imperfect CSI, and for single-cell and multi-cell scenarios [ 15]. For stimulating ZF, and MRT analytic results were executed and MMSE, and Monte Carlo simulations were performed to maximise EE [15].

5.2 Single-Cell Scenario

The chosen deployment model validated.

5.3 Multi-Cell Scenario

A lot of studies have been carried out.

Conclusions and Future Research

6.1Conclusions

This thesis focuses on the energy maximisation improvement of the LS-MIMO systems to cope with energy maximisation problem. The thesis has three main contributions; all are elaborated in detail.

6.2 Future Research

Several recommendations, which may guide to the future research directions on LS-MIMO systems.

 

Bibliography

[1]   K. Zheng, L. Zhao, J. Mei, B. Shao, W. Xiang and L. Hanzo, “Survey of

        Large- Scale MIMO Systems,” in IEEE Communications Surveys &

       Tutorials, vol.17, no. 3, pp. 1738-1760, third quarter 2015.

[2]  D. Feng et al., “A survey of energy-efficient wireless communications,” IEEE Commun. Surveys Tuts., vol. 15, no. 1, pp. 167-168, 1st Quart. 2012.

[3]   T. Kailath and A. J. Paulraj, “Increasing capacity in wireless broadcast  

       systems using Distributed Transmission/Directional Reception (DTDR),” U.S. Patent 5 345 599, Sep. 6, 1994.

[4]    E. G. Larsson, F. Tufvesson, O. Edfors, and T. L. Marzetta, “Massive MIMO for next generation wireless systems,” IEEE Commun. Mag., vol. 52, no. 2, pp. 186-195, Feb. 2014.

[5]   Views on Rel-12 and Onwards for LTE and UMTS, 3GPP RWS-120006, HUAWEI and HiSilicon, 2013.

[6]   E. Björnson, E. G. Larsson and T. L. Marzetta, “Massive MIMO: ten myths

        and One critical question,” in IEEE Communications Magazine, vol. 54,

        no. 2, pp.114-123, February 2016.

[7]  S. Tombaz, A. Västberg, and J. Zander, “Energy- and cost-efficient ultra- high-capacity wireless access,” IEEE Wireless Commun. Mag., vol. 18, no. 5, pp. 18-24, Oct. 2011.

[8] E. Björnson, M. Kountouris, and M. Debbah, “Massive MIMO and small

        cells: Improving energy efficiency by optimal soft-cell coordination,” in

        Proc. ICT, 2013, pp. 1-5.

[9] E. Björnson and E. Jorswieck, “Optimal resource allocation in

        Coordinated multi-cell systems,” Found. Trends Commun. Inf. Theory,

         vol. 9, no. 2/3, pp. 113-381, 2013.

[10] Y. Wu, R. Zhou, and W. Zhang, “Active antenna system: Utilizing the full

         potential of radio sources in the spatial domain,” Huawei, Shenzhen,

        China, 2012.

[11] S. Payami and F. Tufvesson, “Channel measurements and analysis for very large array systems at 2.6 GHz,” in Proc. 6th EUCAP, Prague, Czech          Republic, Mar. 2012, pp. 433-437.

[12]  Further Advancements for E-UTRA Physical Layer Aspects (Release

         9),3GPP TS 36.814, Mar. 2010.

[13]  H. Boche and M. Schubert, “A general duality theory for uplink and

         downlink beamforming,” in Proc. IEEE VTC-Fall, 2002, pp. 87-91.

[14]  R. Kumar and J. Gurugubelli, “How green the LTE technology can be?”in

       Proc. Wireless VITAE, 2011, pp. 1-5.

[15] E. Björnson, L. Sanguinetti, J. Hoydis and M. Debbah, “Optimal

   Design of Energy-Efficient Multi-User MIMO Systems: Is Massive

        MIMO the Answer?,” in IEEE Transactions on Wireless

        Communications, vol. 14, no. 6, pp. 3059-3075, June 2015.