Elsevier

Pervasive and Mobile Computing

Volume 41, October 2017, Pages 106-131
Pervasive and Mobile Computing

Review
Multi-objective optimization for spectrum sharing in cognitive radio networks: A review

https://doi.org/10.1016/j.pmcj.2017.07.010Get rights and content

Abstract

Due to advancements in the wireless networking technologies, a lot of new wireless services are demanding the radio spectrum for their usage. The radio spectrum is considered as a limited natural resource for radio communications, and it became crowded. In fact, this resource is not limited, rather the scarcity of the radio spectrum arises because of inefficient and fixed utilization of this valuable resource. This inefficient and under-utilization of radio spectrum has diverted the attentions of the researchers towards the different techniques in order to access the spectrum dynamically and adaptively. The concept of Cognitive Radio (CR) can be implemented by making the communication devices intelligent, in order to use the available radio spectrum efficiently. The efficient implementation of cognitive radios into the wireless networks gives rise to a tremendous number of optimization problems with multiple objectives. In most of the cases, these objectives are conflicting with each other and need to be addressed carefully. In this article, we will present the importance of multi-objective optimization (MOO) in cognitive radio networks (CRNs). We will discuss about the fundamental differences in single objective optimization (SOO) and MOO techniques, in order to solve MOO problems associated with CRNs. We will differentiate between the possible relations such as, conflicting, supporting and design dependent, among these objectives. A comparative discussion will also be provided about optimization types, solution approaches/algorithms and different combining techniques to combine multiple objectives in order to address MOO problems those comes with CRNs.

Introduction

Cognitive radio (CR) plays a significant role in utilizing the existing radio spectrum more efficiently and adaptively, which has been receiving a growing attention from the researchers in the past decade [1], [2], [3], [4]. The designing, implementation and successful operation of CRs into the legacy wireless networks, give rise to an enormous number optimization problems with multiple performance objectives which need to be addressed simultaneously. Multi-objective optimization is vital in order to address these objectives related to CRNs. Cognitive radios can be programmed and reconfigured dynamically. It has the capability to sense and analyze its surrounding radio environment in order to find out the white spaces or spectrum holes for the transmission of cognitive users. On the basis of interaction with the environment and gathering knowledge of the availability of spectrum holes, a CR can reconfigure its transmission parameters in an automatic and dynamic manner [5]. Fig. 1 shows the idea of spectrum holes. There are two categories of wireless radio users, licensed or primary users and unlicensed or secondary users. The cognitive users are referred to as unlicensed or secondary users. The primary users have paid to use radio spectrum for their services, and on the other hand the secondary or cognitive users have no license for their services.

Multi-objective optimization (MOO) can be used to address optimization problems related to CRNs in many ways. A multi-objective optimization approach based on multi-objective evolutionary algorithm is proposed in [6], where the author considered five most common communication objectives such as, to minimize bit-error rate, maximize throughput, minimize power consumption, minimize spectral interference and maximize spectral efficiency. Naeem et al. [7] provided a comprehensive survey on Resource Allocation (RA) in cooperative CRNs, where the authors also presented the taxonomy of different objectives and protocols used for resource allocation in cooperative CRNs. Ahmad et al. [8] proposed Radio Resource Allocation (RRA) in Cognitive Radio Sensor Networks (CRSNs), which is essentially an MOO problem. Currently, Wireless Sensor Networks (WSNs) are using the unlicensed Industrial, Scientific and Medicine (ISM) band for their transmissions. However, because of the advancements in the wireless technologies and their increased demands, the unlicensed ISM band becomes crowded. This problem is to be addressed by incorporating the intelligent cognitive radio devices instead of using the legacy wireless devices in existing wireless networks.

An overview of existing surveys intended for the CRNs is furnished in Table 1. It is apparent from the table that, majority of the authors have focused on different applications, regarding cognitive radio based wireless networks. Many authors proposed different types of algorithms/solution approaches, in order to address the optimization problems concerned with CRNs. Some of the authors discussed about multiple objectives related to CRNs, and focused on different aggregation methods to combine multiple objectives into a single objective function, in order to solve MOO problems using single objective optimizer. Whereas no work is presented which explores the actual complexities involved in multi-objective optimization problems related to CRNs, especially with multiple conflicting objectives. The wireless networks equipped with CR devices are called CRNs or secondary networks. These CRNs are capable of sensing the environment intelligently, in order to locate and identify the spectrum holes or white spaces, and accordingly adjusting their transmitting and receiving parameters adaptively for their transmissions. Optimization plays a significant role in all areas, such as from designing to successful implementation of cognitive radio networks. The optimization problems concerned with CRNs can be broadly categorized as single objective optimization (SOO) problems as well as multi-objective optimization (MOO) problems. In case of SOO, the main goal of the optimizer is to find the lone optimum solution in order to minimize or maximize only one objective function under various constraints. Whereas in case of MOO, the main goal of the optimizer is to find out multiple tradeoff optimal solutions in order to optimize multiple objectives simultaneously. Most of the real-world optimization problems related to CRNs call for multiple objectives. Since all the objectives are important, they need to be optimized simultaneously. This gives rise to the importance of MOO in the field of CRNs.

The rest of the paper is organized as follow. Section 2 provides the basic concepts of cognitive radios. A detailed discussion on preference-based MOO as well as real MOO is provided in Section 3, where a review on MOO in CRNs is also provided. Different types of optimization and programming techniques, used to formulate and program the optimization problems regarding to CRNs, are discussed in Section 4. Section 5 provides a detailed discussion on algorithms and solution approaches used to address optimization problems related to CRNs. In Section 6, different types of applications regarding CRs are elaborated. Finally, Section 7 concludes the paper and provides future directions.

Section snippets

Basics of cognitive radio

The concept of CR was first conceived by Joseph Mitola in order to make the software defined radio (SDR) more personal [39]. The conceptual view of CR is depicted in Fig. 2. CR is an intelligent radio communication system in which the “Intelligence” is integrated with SDR. An SDR is a programmable and reconfigurable signal processing chip, which is usually developed using digital signal processor or a general purpose microprocessor.

There are three major models of DSA, namely (1) common-use,

Multi-objective optimization

Optimization can be defined as a process to achieve the best outcome of a physical system, or some other situation under some predefined criterion. Every optimization problem calls for one or more objectives to optimize. Generally, most of the real world optimization problems involve more than one objective and therefore called as MOO problems. In an MOO problem, different solutions provide tradeoffs between different objectives. For the optimization problem having only one objective function,

Optimization types

Different types of optimization and programming techniques used to formulate and program the optimization problems in CRNs, are presented in this section. Fig. 8 depicts the taxonomy of optimization types proposed by different authors in their literature related to CRNs. These types include linear programming (LP), mixed integer linear programming (MILP), non-linear programming (NLP), mixed integer non-linear programming (MINP), convex optimization, non-convex (NC) optimization, combinatorial

Solution approaches

Optimization problems can be formulated in many different ways, and can be solved using different types of algorithms and solution approaches. A classification of solution approaches, proposed by different authors in the previous literature in order to address the optimization problems coming to cognitive radios, is depicted in Fig. 9. This taxonomy of solution approaches can be categorized as, Evolutionary, Heuristic, Analytical/Iterative, Game Theoretic and Stochastic.

Applications

In this section, we will discuss about the several application of CRs in the real world scenarios. Fig. 10 shows different types of applications related to CRs. These applications can be characterized as spectrum sensing, spectrum sharing, CR adaptation, resource allocation, learning and reasoning, and many more discussed under miscellaneous category.

Conclusion and future directions

Cognitive radio plays a significant role in order to use the radio spectrum more efficiently and adaptively as well as to address the challenges concerned with spectrum scarcity. The design, deployment, successful implementation and operation of cognitive radios into the current wireless networks, give rise to a large number of optimization problems. The optimization problems related to cognitive radio networks, exposed off with multiple and diverse performance objectives in nature, which need

References (113)

  • ZhangJ. et al.

    Cognitive Radio adaptation decision engine based on binary quantum-behaved particle swarm optimization

  • BalieiroA. et al.

    A multi-objective genetic optimization for spectrum sensing in cognitive radio

    Expert Syst. Appl.

    (2014)
  • PradhanP.M. et al.

    Cooperative spectrum sensing in cognitive radio network using multiobjective evolutionary algorithms and fuzzy decision making

    Ad Hoc Networks

    (2013)
  • PradhanP.M. et al.

    Pareto optimization of cognitive radio parameters using multiobjective evolutionary algorithms and fuzzy decision making

    Swarm Evol. Comput.

    (2012)
  • KhanU.U. et al.

    Fairness in cognitive radio networks: models, measurement methods, applications, and future research directions

    J. Netw. Comput. Appl.

    (2016)
  • JondralF.K.

    Software-defined radio: basics and evolution to cognitive radio

    EURASIP J. Wirel. Comm. Netw.

    (2005)
  • ToshD.K. et al.

    Cognitive radio parameter adaptation using multi-objective evolutionary algorithm

  • NaeemM. et al.

    Resource allocation techniques in cooperative cognitive radio networks

    IEEE Commun. Surv. Tutor.

    (2014)
  • A. Ahmad, S. Ahmad, M. Rehmani, N. Ul Hassan, A survey on radio resource allocation in cognitive radio sensor networks,...
  • HeA. et al.

    A survey of artificial intelligence for cognitive radios

    IEEE Trans. Veh. Technol.

    (2010)
  • BkassinyM. et al.

    A survey on machine-learning techniques in cognitive radios

    IEEE Commun. Surv. Tutor.

    (2013)
  • ElSawyH. et al.

    Stochastic geometry for modeling, analysis, and design of multi-tier and cognitive cellular wireless networks: A survey

    IEEE Commun. Surv. Tutor.

    (2013)
  • ZhaoQ. et al.

    A survey of dynamic spectrum access

    IEEE Signal Process. Mag.

    (2007)
  • ModyA.N. et al.

    Survey of IEEE standards supporting cognitive radio and dynamic spectrum access

  • LiJ. et al.

    A survey of security issues in cognitive radio networks

    Commun. China

    (2015)
  • YücekT. et al.

    A survey of spectrum sensing algorithms for cognitive radio applications

    IEEE Commun. Surv. Tutor.

    (2009)
  • RubyE. et al.

    A survey on distributed channel selection technique using surf algorithm for information transfer in multi-hop cognitive radio networks

  • KhasawnehM. et al.

    A survey on security in cognitive radio networks

  • FragkiadakisA.G. et al.

    A survey on security threats and detection techniques in cognitive radio networks

    IEEE Commun. Surv. Tutor.

    (2013)
  • AkyildizI.F. et al.

    A survey on spectrum management in cognitive radio networks

    IEEE Commun. Mag.

    (2008)
  • ArianandaD.D. et al.

    A survey on spectrum sensing techniques for cognitive radio

  • WangB. et al.

    Advances in cognitive radio networks: A survey

    IEEE J. Sel. Top. Signal Process.

    (2011)
  • GhafoorS. et al.

    Cognitive radio for disaster response networks: survey, potential, and challenges

    IEEE Wirel. Commun.

    (2014)
  • NiyatoD. et al.

    Cognitive radio for next-generation wireless networks: An approach to opportunistic channel selection in IEEE 802.11-based wireless mesh

    IEEE Wirel. Commun.

    (2009)
  • LetaiefK.B. et al.

    Cooperative communications for cognitive radio networks

    Proc. IEEE

    (2009)
  • WangJ. et al.

    Emerging cognitive radio applications: A survey

    IEEE Commun. Mag.

    (2011)
  • ZiaM.T. et al.

    Energy efficient cognitive radio MAC protocols for Adhoc network: A survey

  • MourougayaneK. et al.

    Intelligent jamming threats to cognitive radio based strategic communication networks-A survey

  • GavrilovskaL. et al.

    Medium access control protocols in cognitive radio networks: Overview and general classification

    IEEE Commun. Surv. Tutor.

    (2014)
  • SenguptaS. et al.

    Open research issues in multi-hop cognitive radio networks

    IEEE Commun. Mag.

    (2013)
  • YoussefM. et al.

    Routing metrics of cognitive radio networks: A survey

    IEEE Commun. Surv. Tutor.

    (2014)
  • SumathiA. et al.

    Security in cognitive radio networks-a survey

  • SansoyM. et al.

    Spectrum sensing algorithms in cognitive radio: A survey

  • ChauhanK.K. et al.

    Survey of security threats and attacks in cognitive radio networks

  • AsokanA. et al.

    Survey on cognitive radio and cognitive radio sensor networks

  • UmamaheswariA. et al.

    Survey on performance, reliability and future proposal of cognitive radio under wireless computing

  • HaykinS.

    Cognitive radio: brain-empowered wireless communications

    IEEE J. Sel. Areas Commun.

    (2005)
  • Mitola IIIJ. et al.

    Cognitive radio: making software radios more personal

    IEEE Pers. Commun.

    (1999)
  • HossainE. et al.

    Dynamic Spectrum Access and Management in Cognitive Radio Networks

    (2009)
  • Mitola IiiJ.

    Cognitive radio for flexible mobile multimedia communications

  • Cited by (20)

    • Fuzzy multi-objective programming: A systematic literature review

      2022, Expert Systems with Applications
      Citation Excerpt :

      For solving the optimization problems with one individual objective function, the goal is to get an optimal solution from a set of existing achievable solutions identified as single-objective optimization. If the problems consist of more than one objective function, the decision-maker discovers more optimal solutions identified as MOO (Ramzan et al., 2017). Finding a single solution for a MOP that meets all the objective functions is challenging.

    • A compendium of radio resource management in UAV-assisted next generation computing paradigms

      2022, Ad Hoc Networks
      Citation Excerpt :

      The problem is a single objective if the system is proposed to optimize one objective only (i.e., energy maximization, QoS maximization etc.), while multi-objective optimization collectively optimizes more than one objective. These multi-objectives may or may not be conflicting, but in some problems, they do conflict with each other [59,60]. Many optimization metrics are so far considered in the literature, i.e., QoS, QoE, latency, congestion ratio, resource utilization, and cost etc.

    • Spectrum sensing in cognitive radio networks and metacognition for dynamic spectrum sharing between radar and communication system: A review

      2022, Physical Communication
      Citation Excerpt :

      It selects a portion of significant frequency band out of the available bands, and then SUs decide either to use this band or not. Several optimization methods can be used to achieve optimal decisions based on the surrounding environments [69]. It includes several features such as spectrum characterization, spectrum selection, routing protocol, reconfiguration, link delay [70], and operating frequency.

    • A Reinforcement Learning based evolutionary multi-objective optimization algorithm for spectrum allocation in Cognitive Radio networks

      2020, Physical Communication
      Citation Excerpt :

      Consider an example, higher network capacity can be achieved by providing more transmission resources but it effects spectral efficiency as utilizes the same resource concurrently with good channel conditions which increases mutual interference and hence degrades spectral efficiency. Further, complicated signal processing techniques also obtain higher network capacity but it increases the cost and complexity of devices [3,4]. However, in practice, these conflicting objectives need careful balancing to obtain tradeoff among them to optimize the overall performance of CR networks.

    • Design of Intelligent Spectrum Sensing and Sharing Algorithm for Massive MIMO

      2023, Proceedings of International Conference on Computational Intelligence and Sustainable Engineering Solution, CISES 2023
    View all citing articles on Scopus
    View full text