Elsevier

Computer Networks

Volume 182, 9 December 2020, 107508
Computer Networks

Application of Non-Orthogonal Multiple Access for machine type communication in sub-terahertz band

https://doi.org/10.1016/j.comnet.2020.107508Get rights and content

Abstract

There is a compelling need to address the problem of accommodating the rapidly growing number of machine type devices engaging in communication within the scarce radio spectrum, in current and forthcoming generations of wireless networks. The coming decades will also witness various novel applications, demanding wireless communication at very high data rate and energy efficiency, which currently favored technologies and techniques are inadequate to support. Because of these reasons, it seems inevitable that wireless communication will have to be facilitated by the efficient utilization of larger bandwidth at higher frequencies. Non-Orthogonal Multiple Access (NOMA) and shifting to higher frequency ranges such as the lower terahertz band, have been considered as solutions to provide space to the increasing number of communicating devices while meeting performance requirements. In this paper a machine type communication system model, utilizing the promising technique NOMA and operating in the sub-terahertz band, is developed and analyzed for its performances of data rate, spectral efficiency and energy efficiency. Thereafter the energy efficiency of the proposed model is optimized by using unconstrained optimization and then constrained (by threshold of data rate) optimization, the latter solving a multi-objective problem by the ϵ-constraint method. The optimal transmission power is also derived for both optimization techniques. Finally, simulation results demonstrate that the energy efficiency can be maximized by utilizing unconstrained optimization and the spectral efficiency can be maximized by utilizing constrained optimization for both machine-to-machine and base-to-machine communications.

Introduction

As an exponentially growing number of devices are participating in the Internet of Things (IoT), its impact on individuals, businesses, economies and societies is going to be massive. With estimations of a trillion sensors and billions of users to be connected to the Internet by the year 2021[1], wireless data traffic, a crucial facilitator of IoT, is thus expected to increase in several folds by 2021. Furthermore, we are witnessing the advent of applications such as virtual and augmented reality, self driving cars, telesurgery, wireless backhaul, etc. In fact, within the next decade there could be even newer applications not envisioned yet. Almost all of these applications will have stringent requirements of low latency, high data rate, spectral and energy efficiency, at a level greater than even fifth generation (5G) networks can provide[2].

Given such a scenario, there is urgency for suitably accommodating users and sensors within the available yet limited frequency spectrum in order to enable the required level of communication. Various technologies and techniques are being investigated to prevent wireless spectrum saturation, e.g.,by shifting to higher frequency ranges to obtain required bandwidths or using more efficient means of utilizing the already available frequencies. In[2], Rappaport etal. explain that required data rates such as those in the order of 100 Gb per second may only be available at frequencies beyond the 100GHz mark.

The frequency spectrum between 100GHz and 10 THz is identified as the Terahertz (THz) band, while the frequency spectrum in the range of 100GHz–300GHz is referred to as the sub-THz or sub-mmWave band[2], [3], [4]. Its feasibility of accommodating the increasing number of wireless communication devices and supporting the new applications, is currently being examined. Along with providing vastly greater bandwidth and data rate support, operating in this band is advantageous in terms of robustness in various unfavorable environments, resisting noise from optical sources, supporting strong communication security, etc. without posing any health concern[5]. One of the major drawbacks of the technology, however, is that operating in frequencies above 1 THz results in considerable propagation losses, thus severely limiting the range of communication. Considering how operating in the lower THz bands has enabled covering a reasonable range of distance[6], and in an effort to deal with this challenge, this paper has carried out analysis in the lower part of this spectrum. greater than that of URLLC.

In order to combat spectral saturation, which is a hindrance to the requirements of massive connectivity, researchers have recently demonstrated their interest in non orthogonal multiple access (NOMA). Multiple access (MA) techniques can be broadly classified into orthogonal multiple access (OMA) and NOMA. Although, OMA has been utilized in the present and earlier generations of networks, currently there is consensus that NOMA is essential, and potentially more effective than OMA for the applications and requirements of 5G networks and beyond. NOMA brings better spectral efficiency and throughput to communicating systems than OMA, by assigning the same frequency to multiple users (hence non-orthogonal) while segregating in the power or code domain, thus enabling massive connectivity through the share of communication resources[7]. This is also evinced by the plot in Fig.1 where NOMA is superior to OMA in terms of higher spectral efficiency across the frequency spectrum from microwave (1 – 30GHz) to mmWave (30–100GHz) to the sub-THz range (100–300GHz) range as stated in[8]. Although spectral efficiency of sub-THz is lower than that of mmWave, however high bandwidth (i.e.,10–100GHz) has been proposed in THz frequency[9], [10], [11]. Furthermore, it also provides better user fairness, lower latency and compatibility with present and upcoming systems of communication[12]. NOMA is not a fundamentally novel technique: from an information theory perspective, the idea of sharing communication resources non orthogonally was understood previously as well. However, despite its potentials over OMA, NOMA implementation was held back primarily due to the level of signal processing and associated processing power required by receivers. For NOMA receivers to discern and detect their anticipated signals from a superposition of various others which reach them, they need to be complex enough to carry out successive interference cancellation (SIC). However, that level of complexity for receivers was deemed unacceptable in the previous generations. Today, with advances in processing power capabilities, there has been a reconsideration of SIC and hence NOMA within the research fields of both academia and industry[13]. Because the migration to higher frequency spectra within the near future seems inevitable, and NOMA appears to be the ideal MA technique, this paper investigates the use of NOMA in the lower THz band, in order to effectively and efficiently deal with the challenge of utilizing the bandwidth in this domain. Furthermore, due to the large band of THz network, which will serve huge amount of data services, it will not only cause challenges such as heavy transmission burden, but theoretically also result in significant energy cost. However, the use of NOMA in the THz band can address this by assigning multiple users to the same transmission channel. User clustering techniques with NOMA may result in significantly better energy efficiency performance than OMA as demonstrated in[14], where the authors expected the use of THz-NOMA caching systems to improve data rate performances as well. This establishes the rationale for investigating the integration of these two technologies and analyzing key performance metrics such as data rate, energy efficiency and spectral efficiency.

As well as using NOMA, this paper also makes use of orthogonal frequency division multiplexing (OFDM), a multicarrier modulation technique that forms the basis of MA in 4G and LTE systems. OFDM’s working principle is to divide the available bandwidth or resource blocks into a number of narrow band, overlapping but orthogonal subcarriers (SCs). OFDM techniques have been used in IoT networks such as downlink Narrow Band IoT (NB-IoT): a standard presented by 3GPP and designed for massive numbers of machine type communication (MTC) devices over long ranges, low rate and energy efficient links. Using narrow bandwidths, the NB-IoT standard has greater coverage than global positioning system/general packet radio service and is also able to filter out relatively more noise, thus yielding better signal to interference and noise ratio (SINR)[15]. In order to investigate how each of the above mentioned technologies integrate, our system model runs over a narrow band in the sub-THz spectrum, utilizing OFDM as well as power domain NOMA to enable each of the OFDM SCs to be shared by multiple devices, segregated by their transmission power levels[16]. Such a communication system should theoretically support a vast number of devices and higher levels of spectral efficiency, than what could have been supported by OFDM systems alone, which allows each SC to be occupied by only one device per time slot[17].

Machine to machine communication is an integral part of the IoT landscape, where various types of sensors and devices independently communicate among themselves, sharing information and collaborating on projects, without being governed centrally. Information that is gathered by these devices are often passed onto actuators in real time control systems. Therefore, the ability to guarantee high levels of reliability and timing may be critical for many such devices. Based on this criterion, two classes of machine type communication devices (MTCDs) exist: (i) ultra reliable low latency communication (URLLC) devices, characterized also by high communication power requirements and (ii) massive machine type communication (mMTC) devices, characterized by much lower cost, power and moderate latency and reliability requirements. The former class of devices is going to be utilized in applications such as autonomous driving, telesurgery and industry automation while the latter class will be functional as metering, monitoring and sensing devices[18]. This paper presents a system model of a wireless network of MTCDs, evaluated for its performance in the downlink scenario, using the aforementioned techniques of NOMA over OFDM SCs in the lower THz band. Hence, the main contributions of this paper are:

  • Proposing a novel system model where the BS and MTCDs utilize NOMA to share OFDM SC over a narrow, sub-THz band of frequency. This investigation of NOMA in the THz band attempts to enable massive connectivity of devices with enhanced communication performance.

  • Derivation of data rate and energy efficiency for the machine-to-machine and base-to-machine links of the proposed model.

  • Maximizing energy efficiency of the proposed model, first via unconstrained optimization and then via constrained optimization by formulating a multi-objective problem and solving it with the ϵ-constraint method.

  • Highlighting spectral efficiency and energy efficiency performance between BS and MTCDs and also MTCDs between themselves, observing its maximization with the optimization techniques used and comparing between the techniques.

The organization of the paper henceforth is as follows. Section2 discusses related works to this study, Section3 presents and explains the system model, along with equations of SINR as well as channel gain, Section4 discusses the performance metrics such as data rate, energy efficiency and optimization techniques for maximum energy efficiency, Section5 consists of the results from numerical simulations and discussions of the plots and finally Section6 concludes the paper with the best scheme for energy efficiency of proposed model and suggestions for future studies.

Section snippets

Related works

In this section, some of the major works related to the techniques and schemes are briefly highlighted (see Table1).

In order to unify the existing knowledge on THz communication research,[6] made a comprehensive overview of the technology, highlighting its potential benefits (such as terabit-per-second capacities of communication, smaller transceivers and improved energy efficiency) and the important research challenges to be addressed (e.g.,barriers to designing THz electronics and media

Network descriptions

Consider a landscape with a BS1 and several MTCDs in downlink scenario, where machine-to-machine communication (between two MTCDs) and base-to-machine communication (from BS to MTCD) is shown in Fig.2. For the given network model, two main types of communication are possible in this landscape:

  • (1)

    The BS communicates with MTCD. So the communication between MTCD and the BS is going to be denoted as “base-to-machine (B2M)”.

  • (2)

    MTCDs communicate between

Evaluating performance metric

In this section, we analyze the performance metric such as the data rate, energy efficiency and develop optimization techniques for the various M2M and B2M communications of our system model.

Results and discussion

Using the equations for the data rate, energy efficiency and required optimum power of the communication links that are considered in the system model, we have carried out a number of simulations to analyze performance in this section. In these simulations, spectral efficiency (bps/Hz) is calculated by the ratio of data rate to bandwidth. Energy efficiency (bps/Hz/W) is calculated by the ratio of spectral efficiency to Pidle+E[pt].

Conclusion and future directions

To conclude, in this paper, an application of NOMA for MTC systems in the sub-THz band was shown. A system model was presented and equations for the data rate, energy efficiency and optimal transmission power for its different communication links were provided. The performance of the model in terms of data rate, optimal transmission power, spectral efficiency and energy efficiency was evaluated and the plots from simulations were given. The optimization of the transmission power for maximum

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Saifur Rahman Sabuj was born in Bangladesh in 1985. He received a B. Sc in Electrical, Electronic and Communication Engineering from Dhaka University, Bangladesh in 2007, an M. Sc Engineering in the Institute of Information and Communication Technology, Bangladesh University of Engineering and Technology, Bangladesh in 2011, and a Ph.D. degree in the Graduate School of Engineering, Kochi University of Technology, Japan in 2017.

From 2008 to 2013, he was a faculty member of Green University of

References (45)

  • APT Report on Technology trends of Telecommunications above 100 GHz

  • HanC. et al.

    Terahertz communications (TeraCom): Challenges and impact on 6G Wireless Systems

    (2019)
  • Technology trends of active services in the frequency range 275-3000 GHz

    (2015)
  • R. Singh, D. Sicker, Beyond 5G: THz spectrum futures and implications for wireless communication, in: Proc. 30th...
  • Y. Saito, Y. Kishiyama, A. Benjebbour, T. Nakamura, A. Li, K. Higuchi, Non-orthogonal multiple access (NOMA) for...
  • VaeziM. et al.

    NOMA: An information-theoretic perspective

  • ZhangH. et al.

    Energy efficient user clustering, hybrid precoding and power optimization in terahertz MIMO-NOMA systems

    (2020)
  • FeltrinL. et al.

    Narrowband IoT: A survey on downlink and uplink perspectives

    IEEE Wirel. Commun.

    (2019)
  • IslamS.M.R. et al.

    Power-domain non-orthogonal multiple access (NOMA) in 5G systems: Potentials and challenges

    IEEE Communications Surveys & Tutorials

    (2016)
  • A.E. Mostafa, Y. Zhou, V.W. Wong, Connectivity maximization for narrowband IoT systems with NOMA, in: Proc. IEEE...
  • JiH. et al.

    Ultra-reliable and low-latency communications in 5G downlink: Physical layer aspects

    IEEE Wirel. Commun.

    (2018)
  • ChenZ. et al.

    Optimal precoding for a QoS optimization problem in two-user MISO-NOMA downlink

    IEEE Commun. Lett.

    (2016)
  • Cited by (0)

    Saifur Rahman Sabuj was born in Bangladesh in 1985. He received a B. Sc in Electrical, Electronic and Communication Engineering from Dhaka University, Bangladesh in 2007, an M. Sc Engineering in the Institute of Information and Communication Technology, Bangladesh University of Engineering and Technology, Bangladesh in 2011, and a Ph.D. degree in the Graduate School of Engineering, Kochi University of Technology, Japan in 2017.

    From 2008 to 2013, he was a faculty member of Green University of Bangladesh, Metropolitan University, Sylhet and Bangladesh University. He has been in the position of an Assistant Professor in the department of Electrical and Electronic Engineering at Brac University, Bangladesh, since September 2017. He is currently working with Electronics and Control Engineering Department as a Postdoctoral Research Fellow at Hanbat National University, South Korea. His research interests include MIMO-OFDM/NOMA, Cooperative Communication, Cognitive Radio, Internet of things, and Machine-to-machine for wireless communications.

    A.M. Musa Shakib Khan has recently graduated from the department of Electrical and Electronic Engineering (EEE) at Brac University, Dhaka, Bangladesh. His concentration within EEE is in Communications and Computers.

    His research interest lies in technologies that facilitate 5G wireless communication (and beyond) such as multiple-access techniques, multiple input multiple output systems and means of optimizing communication for maximum performance and efficiency. He also studies software and hardware architectures of embedded systems, such as those constituting the Internet of things or having biomedical applications.

    Mr. Khan has been awarded merit and performance scholarships at Brac University and was also a recipient of scholarship at McGill University, where he had previously studied for a number of semesters.

    Masanori Hamamura received his B.S., M.S. and Ph.D. degrees in electrical engineering from Nagaoka University of Technology, Nagaoka, Japan, in 1993, 1995 and 1998, respectively. From 1998 to 2000, he was a Research Fellow of the Japan Society for the Promotion of Science.

    Since 2000, he has been with the Department of Information Systems Engineering at Kochi University of Technology, Kochi, Japan, where he is now a Professor. From 1998 to 1999, he was a visiting researcher at Centre for Telecommunications Research, King’s College London, United Kingdom, where he worked on adaptive signal processing for mobile systems. His current research interests are in the areas of signal design, wireless communications and signal processing.

    View full text