Development and Performance Evaluation of Multi-Objective Approach-Based Spectrum Allocation in Cognitive Radio Network

doi:10.21203/rs.3.rs-3347470/v1

Download PDF

Research Article

Development and Performance Evaluation of Multi-Objective Approach-Based Spectrum Allocation in Cognitive Radio Network

https://doi.org/10.21203/rs.3.rs-3347470/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

For effective utilization of radio frequency, Spectrum Sensing (SS), Spectrum Decision (SD), Spectrum Allocation (SA), and Spectrum Management(SM) are some recent research directions in Cognitive Radio Networks (CRNs). Out of these sub-areas, the SA is an important task in CRN. In the past, researchers have addressed SA as a Single Objective Optimization (SOO) problem and have solved it by using bio-inspired techniques. But in reality, the SA has a multi-objective optimization problem. In the recent past, the SA problem has been addressed as a two objectives optimization problem and has been solved by Multi–Objective Bio-inspired Approach (MOBA). In the present paper, we have also considered spectral and energy efficiencies of the SA task as two independent objectives and then solved this problem using two different Multi-Objective Optimization (MOO) techniques. These two are the Multi-Objective Ant Lion Optimization Algorithm (MOALO) and the Multi-Objective Grasshopper Optimization Algorithm (MOGOA). The spectrum utilization performances achieved by the two approaches have been compared. In addition, the results obtained by the already reported Superior Population Genetic Algorithm (SPGA) based technique have also been compared. The analysis of simulation-based experimental results obtained from the three methods reveals that the proposed MOALO algorithm provides superior performance in terms of Pareto front and network capacity compared to the other two approaches. The proposed multi-objective methods can also be applied when the SA task is viewed as a three-objective optimization problem.

Spectral efficiency

Energy efficiency

Cognitive Radio Network

Multi-objective Optimization

MOALO

MOGOA

SPGA

In recent times, the use of wireless devices has increased tremendously. Effective utilization of radio resources is a significant challenge to overcoming spectrum scarcity issues because radio spectrum is a finite resource [1]. The CRN utilizing the Dynamic Spectrum Allocation (DSA) technique is a promising technology to overcome spectrum scarcity issues as well as efficient utilization of radio spectrum [2]. Several performance criteria as objective functions are used for technological preparation for dynamic spectrum allocation such as network capacity, spectral efficiency, energy efficiency, fairness among users, and overhead reduction. These attributes and their performance are correlated. The performance criteria cannot be used individually [3]. The main objective of DSA is to enhance spectral efficiency through the effective utilization of underutilized spectrum bands; however, more energy is required for dynamic spectrum allocation [4]. In our work trade-offs between Spectral Efficiency (SE) and Energy Efficiency (EE) are considered which are optimized using MOALO and MOGOA optimization techniques. ALO algorithm is a meta-heuristics optimization technique that is inspired by the hunting mechanism of antlions [5]. A SOO algorithm can achieve only one best solution. The trade-off between spectral and energy cannot be determined by using an SOO algorithm. So MOALO has been proposed in our work to find the trade-off between them. Non-dominated Pareto optimal solutions have been archived in a repository. The roulette wheel mechanism is used to find the solution from the repository [6]. MOGOA is also a nature-inspired optimization algorithm that is inspired by the swarming behavior of grasshoppers [7]. This algorithm is also employed in our work for finding optimal solutions between two independent objectives.

Spectral and energy efficiencies are formulated as MOO problems to maximize both these objectives and to find trade-offs between them for solving SA problems in CRN.
MOALO and MOGOA are used for optimization of two objectives and the Pareto front of the MOALO has been investigated for different SINR thresholds.
The convergence rate of the proposed algorithm is also achieved for the two objectives and network capacity varying with the number of Secondary Users (SUs) and average energy efficiency varying with the number of SUs are also observed for different SINRs.
SPGA, MOGOA, and MOALO algorithms are all compared in terms of their trade-off analysis between SE and EE.

Several articles related to both SO and MO optimization algorithms used for spectrum allocation have been reviewed. The single objective cannot achieve the trade-off between multiple conflicting objectives; additionally, it deteriorates the performance of other objectives. To achieve balanced performance between the two objectives, the researcher formulated the multiple objectives as a MOO problem. Some of the works in this area are given in this section.

In [8], maximum energy efficiency, minimum fairness, and maximum spectrum utilization have been achieved using the Non-Sorting Genetic Algorithm (NSGA-II). This work has focused on achieving joint spectrum and power allocation in the CRNs. Multi-Objective Different Evolution (MODE) algorithm maximizes network capacity and compares with NSGA-II and MOPSO algorithms [9]. Ant Colony Optimization(ACO) algorithm is used to improve spectrum utility, and detection rate and reduce collision, detection rate, and reduced false alarm probability [10]. NSGA-II algorithm is used to find trade-offs between network throughput and spectrum utilization [11]. In [12], the NSGA-II algorithm is used for the improvement of spectrum usage by the SUs and reduction of interference and to achieve sum rate, fairness, and quality of service. Enhancement of Spectrum usage by the SUs and interference minimization is achieved using Multi-Objective Ant colony optimization (MOACO) and Deep Reinforcement Learning (DRL) with a Double Q-learning algorithm [13].In [14], the Wave-based adaptive differential evolution (WADE) algorithm is suggested to find the Pareto optimal solutions between network utility and fairness. The proposed technique has gained more than 39.77% in terms of network utility and fairness comparisons to the Differential Evolution (DE) algorithm. The Coevolution Chaotic Particle Swarm Optimization (CCPSO) algorithm achieved greater EE comparisons to the PSO algorithm [15]. To resolve channel allocation issues and improve SE and to achieve channel gain, channel allocation robustness, and maintain minimum network interference, a genetic algorithm is used [16]. Non-dominated sorting and quantum particle swarm optimization are proposed in [17] for finding a trade-off between network utilization and fairness. Non-dominated sorting quantum particle swarm optimization has a better convergence rate and accurate convergence values. Allocation of the radio spectrum to secondary users, as well as utilization of the radio spectrum, has been achieved using a Chemical reaction-based algorithm (CRO) [18]. The cooperative spectrum sensing technique is used to find the trade-off between spectral and energy efficiency in CRNs [19]. System utility and fairness are improved by using the proposed technique.NSGA-II algorithm is used to find the Pareto optimal front between SE and EE in massive multiple input and multiple output systems [20]. Yellow Saddle Goatfish Algorithm (YSGA) is a bio-inspired technique used for maximization of network capacity.YSGA algorithm performs better compared to Multi-Objective Particle Swarm Optimization (MOPSO) and Multi-Objective Differential Evolution (MODE) algorithms [21].

The following is a description of the study: Section II discusses numerous issues raised and techniques employed in the related work; the system model is described in Section III. The suggested MOALO, which is employed for finding Pareto optimal solutions between spectral and energy efficiencies, is also discussed in section IV. Section V discusses the simulation results. Section VI contains a summary of the paper's conclusion.

The dynamic spectrum allocation model consists of a primary base station (P), cognitive radio base station (S), the number of licensed users (A), the number of unlicensed users(B), channel bandwidth (W), and the channel available for communication(C). In Fig. 1, the network model consists of Primary users' Base Station (PU-BS) and Secondary Users Base Station (SU-BS) with primary users denoted as PU= {1, 2,....PU_a}, and secondary users expressed as SU= {1, 2... SU_b} which is connected with PU-BS and SU-BS respectively. C= {1, 2… C} represents a communication channel with a total bandwidth of W. Cognitive users have some knowledge regarding channel state conditions. Two factors must be considered when assigning channels: the first is to avoid interferences between Primary Users (PUs) and Secondary Users (SUs), and the second is to reduce interferences between SU and SU. Secondary users must have information about channel occupancy.

The Signal interference and noise ratio (SINR) for b^th secondary users represented by s^th cognitive base station occupying c^th spectrum band is given by:

$$\text{S}\text{I}\text{N}\text{R}=\frac{{g}_{s,b}^{c}{P}_{s,b}^{c}}{{\sigma }^{2}+{i}_{b}^{c}}$$

Here

${g}_{s,b}^{c}$ is the channel gain between s^th SU-BS and b^th cognitive radio on c^th spectrum band

${P}_{s,b}^{c}$ is the transmission power of s^th SU-BS assigned to b^th secondary users on c^th spectrum band

${\sigma }^{2}$ represents noise power

${i}_{b}^{c}$ Interference between the PUs and SUs

Interference between the PUs and SUs is given by:-

$${i}_{b}^{c}=\sum _{a=1}^{A}{P}_{p,a}^{c}{g}_{p,a}^{c}+ {\sum }_{i=1,i\ne b}^{b}{\sum }_{l\in U/\left\{u\right\}}{P}_{l,i}^{c}{g}_{l,i}^{c}$$

Here

${P}_{p,a}^{c}$ is denoted as transmitted power from p^th PU-BS allocated to a^th PUs on c^th spectrum band.

${g}_{p,a}^{c}$ channel gain between p^th primary base station allocated to a^th primary users on c^th spectrum band.

${P}_{l,i}^{c}{g}_{l,i}^{c}$ is represented as interference due to other cognitive users.

The capacity acquired by each secondary user associated with s^th cognitive base station over c resources expressed as:-

$${C}_{b}=\sum _{c\in C}\frac{W}{c}{a}_{s,b}^{c}{\text{log}}_{2}(1+SINR)$$

Here w represents bandwidth, the number of resources denoted by c, and ${a}_{s,b }^{c}$represents the spectral band allocated to secondary users.

The total capacity obtained by all SUs per unit bandwidth, represented as a system's spectral efficiency (in bps/Hz), is formulated as an objective function 1.

It is denoted as

Function1:$SE=\sum _{b\in B}\frac{{C}_{b}}{W}$ (4)

Energy efficiency (in bps/Hz/W) is expressed as the total system capacity obtained by all CRs per total system power, formulated as an objective function 2.

Function:$EE=\sum _{b\in B}\frac{{C}_{b}}{{P}_{T}}$ (5)

The multi-objective function can be formulated as

F1:MaxSE (6)

F2:MaxEE (7)

Subject to:

$${ C}_{1}:-\sum _{c\in C}{a}_{s,b}^{c}=1 ,s\in S,b\in B$$

$${ C}_{2}:- {a}_{s,b}^{c}\in \left\{\text{0,1}\right\} c\in C$$

$${ C }_{3}:- {a}_{s,b}^{c}{\text{log}}_{2}(1+SINR) \ge {C}_{b}$$

The main objective of our work how to improve both the independent objectives by utilizing the unutilized spectrum band for dynamic spectrum allocation. Dynamic Spectrum Allocation (DSA) is also an energy-consuming operation that also reduces energy efficiency. So efficient dynamic spectrum allocation technique can be achieved by maximizing spectral and energy efficiency in CRNs.

The two independent objectives cannot improve the system's performance individually. The trade-off analysis between the two conflicting objectives plays a vital role in performance improvement. MOO problem optimizes. M no of objectives maximize simultaneously to achieve Pareto optimal solutions.

Maximize y = F(x)= [f₁(x),f₂(x)…f_m(x)]^T (8)

Subjected to g_k(x) ≤ 0, k = 1, 2,…N

and x =[x₁,x₂……………x_p]^T

y is known as the objective vector and g_k represents a constraint.

3.1 MOALO Algorithm Employed for Spectrum Allocation In CRN

SOO techniques are mostly used for finding one of the best solutions. MOO techniques are mostly used for finding more than one solution and MOO has multiple trade-off solutions. These techniques can generate multiple optimal solutions. Pareto solutions are the optimal solution for a given optimization problem. The multi-objective optimization problems are solved by MOALO using roulette wheel mechanisms. The significant characteristics of the algorithm are explained as the random walk of ants, entrapment in an antlion pit, constructing a pit, Sliding ants towards antlions, catching prey, and reconstructing the pits and elitism.ALO has some distinct features like convergence speed and diverse Pareto optimal front

Step 1:-

The algorithm starts with the random values

Y(t)=[0,cumsum((2k(t₁))-1),cumsum((2k(t₂))-1)… cumsum((2k(t_n))-1)] (9)

Cumsum is used for finding the cumulative sum, maximum iteration, and the no of steps represented by n and t.

$$\text{k}\left(\text{t}\right)=\left\{\begin{array}{c} 1 random number>0.5 \\ 0 random number \le 0.5 \end{array}\right.$$

The fitness values of every ant are calculated using objective Functions in each iteration

To normalize the ant's random walk, the following equation is used.

$${\text{Y}}_{\text{i}}^{\text{t}}=\frac{\left[{\text{Y}}_{\text{i}}^{\text{t}}-{v}_{\text{i}}^{\text{t}}\right]\times \left[{\text{e}}_{\text{i}}^{\text{t}}-{\text{q}}_{\text{i}}^{\text{t}}\right]}{\left[{\text{u}}_{\text{i}}^{\text{t}}-{\text{v}}_{\text{i}}^{\text{t}}\right]}$$

Here ${q}_{i}^{t}$ and ${\text{e}}_{\text{i}}^{\text{t}}$ represent the minimum and maximum of i^th variable at t^th iteration. The variables${\text{v}}_{\text{i}}^{\text{t}}$ and ${\text{u}}_{\text{i}}^{\text{t}}$ represent the minimum and maximum random walks, respectively.

Step 2:-

ALO mimics the trapping of ants in antlion pits by altering the random walks in their vicinity.

${\text{q}}_{\text{i}}^{\text{t}}={\text{a}\text{n}\text{t}\text{l}\text{i}\text{o}\text{n}\text{s}}_{\text{j}}^{\text{t}}+$ q^t (11)

${\text{e}}_{\text{i}}^{\text{t}}={\text{a}\text{n}\text{t}\text{l}\text{i}\text{o}\text{n}\text{s}}_{\text{j}}^{\text{t}}+$ e^t (12)

q^t represents the minimum of all variables at i^th iteration and e^t indicate the maximum of all the variable at i^th iteration.${\text{A}\text{n}\text{t}\text{l}\text{i}\text{o}\text{n}\text{s}}_{\text{j}}^{\text{t}}$represents the position of the selected antlion at i^th iteration.

Step 3:-

To improve their chances of survival, larger antlions dig deeper pits. Stronger antlions attract more ants using the roulette wheel mechanism.

Step 4:-

The boundaries of the random walk are decreased to mimic the sliding ants toward antlions.

$${q}^{t}=\frac{{q}^{t}}{I}$$

$${e}^{t}=\frac{{e}^{t}}{I}$$

The final step of the algorithm is catching the ants and reconstructing the pits.

Step 5:-

The last operator in ALO is elitism. Fitter antlion formed during optimization is stored in an archive. Random walks on antlions gravitate towards selected ant lions and the elite antlion.

$${\text{a}\text{n}\text{t}\text{l}\text{i}\text{o}\text{n}\text{s}}_{\text{i}}^{\text{t}}=\frac{{R}_{A}^{t}+{R}_{E}^{t}}{2}$$

3.2 SPECTRUM ALLOCATION USING THE PROPOSED MOALO ALGORITHM

The spectrum band ${a}_{s, b}^{c}$ =1 when it is allocated to SUs and 0 otherwise. The spectrum band is allotted to SUs sequentially. Non-dominated Pareto solutions are obtained. The Pareto solutions are ensured with ranks using a roulette-wheel mechanism. The proposed algorithm used for the improvement of performance metrics such as SE and EE, which is given in Table 1.

Table 1

MOALO Algorithm for Spectrum Allocation
Algorithm 1:	MOALO optimization algorithm used for finding a trade-off between the two independent objectives(SE and EE)
Step 1: Locate A no of primary users and B no of secondary users using channel C for communication having bandwidth W. Step 2: Calculation of Signal interference and noise ratio for b^th secondary users of s^th cognitive base station occupying c^th spectrum band. Step 3: Define the objective function for optimal spectrum allocation (SE (objective 1) and EE (objective 2)). Step 4: Initialization of MOALO parameters (max_iter, no of population, archive Max Size, and number of objectives) Step 5: Initialization of ant positions and select the elite from the archive Step 6: Random walk around the antlion and its normalization Step 7: Fitness function evaluation Step 8: To find a Pareto front between two objectives, each solution compares with the rest of the solutions in the populations. Either a set of solutions dominates each solution or each solution dominates the domination count. Step 9: Update the archives Step 10: Ranking process(First non-dominated Pareto front considered as a Rank 1) Step 11:Delete some solutions using the Roulette wheel mechanism and from the archive to store new solutions Step 12:Pareto optimal solutions achieved between two objectives

The flow chart of spectrum allocation using the proposed MOALO algorithm is presented in Fig. 2.In this paper we have also employed MOGOA algorithm for efficient allocation of spectrum. However, its flowchart is not provided due to space limitations.

The performance analysis of the suggested technique for optimum spectrum allocation is explained in this section. In our simulation study 5 numbers of PUs, 6 numbers of SUs, and 7 numbers of channels have been considered. The system and algorithm parameters used for simulation are given in Table 2.

Table 2

System and algorithm parameters
Parameter	Value
The transmission power of the cognitive base station	20dBm
SINR threshold	10dB
Noise power	0.001dBm/Hz
Interference threshold	5*10^− 12W
Number of resources	10
Bandwidth	10MHz
Population size	100
Iteration	100
Dimension	2
Archive Size	5
The upper bound, the lower bound	0.1,1

In Fig. 3, it is evident from the graph when SE increases from 0 to 3*10^8 the corresponding EE decreases very fast in the beginning and almost remains constant after 0.5*10^8 at SINR 5 dB. The Pareto solutions achieved between the two objectives are conflicting because the SE can be further increased with the sacrifices of EE. The comparative plot between the two objectives at three different SINRs (5dB, 10dB, 15dB) has been obtained from a simulation study, which is shown in Fig. 4, the observation of this plot demonstrates at a given SE the EE become higher as the SINR decreases and it is also noticed from the plot, the length of the Pareto front decreases with the increasing value of SINR thresholds. Interference between the channels depends upon the SINR threshold.

In Fig. 5 and Fig. 6, the convergence plot between SE and iteration and EE and iteration obtained during the optimization process are shown, these two plots remain constant w.r.t iteration and become close to zero, when the training is stopped.

The comparative plot between the number of secondary users and Network capacity at three different SINRs (5dB, 10dB, 15dB) has been shown in Fig. 7, it is observed that the network capacity attains a steady state value when the number of secondary users becomes more than 10, further, it is noticed the network capacity achieves the highest value at high SINR value.

Figure 8, displays a comparative performance plot between average energy efficiency and the number of secondary users under different SINRs (5dB, 10dB, 15dB). It is found from the simulation study; that average energy efficiency initially decreases gradually and then remains constant after 10 numbers of secondary users. It is also noticed that at high SINR, the average energy efficiency attains a higher value compared to low SINR.

Comparison of Network Capacity(NC) for two different SINRs (15dB, 10dB) between MOALO and MOGOA are represented in Table 3. It is noticed from the Table that the more secondary users present on the CR network, the network's capacity is also increased. The proposed algorithm achieves better network capacity at different SINRs (15dB, 10dB) in comparison to the MOGOA algorithm.

Table 3

Comparison of Network capacity for DIFFERENT SINRS BETWEEN MOALO and MOGOA
No of SUs	Comparison of Network capacity for different SINRs(15dB,10dB) between MOALO and MOGOA
No of SUs	Network capacity(bps) using MOALO(NC_15dB)	Network capacity(bps) using MOGOA(NC_15dB)	Network capacity(bps) using MOALO(NC_10dB)	Network capacity(bps) using MOGOA(NC_10dB)
1	8.5789*10⁵	7.6565*10⁵	1.3683*10⁶	1.2078 *10⁶
3	9.0856*10⁵	8.1631*10⁵	1.4443*10⁶	1.2838 *10⁶
5	9.9122*10⁵	8.9898*10⁵	1.5683 *10⁶	1.4078 *10⁶
7	1.0666*10⁶	9.7431*10⁵	1.6813 *10⁶	1.5208 *10⁶
9	1.1359*10⁶	1.0436*10⁶	1.7853 *10⁶	1.6248 *10⁶
12	1.2006*10⁶	1.1083*10⁶	1.8823 *10⁶	1.7218 *10⁶
14	1.2192*10⁶	1.127*10⁶	1.9103*10⁶	1.7498* 10⁶
16	1.2139*10⁶	1.1216*10⁶	1.9023*10⁶	1.7418*10⁶
18	1.2039*10⁶	1.1116*10⁶	1.8873*10⁶	1.7268*10⁶
20	1.1719*10⁶	1.0796*10⁶	1.8393*10⁶	1.2078 *10⁶

More CRs in a network create more interference with the primary users as well as with the neighbor CRs. So more power is required to satisfy interference. The plots between power and network capacity and power and EE are represented in Fig. 9 and Fig. 10. The network capacity of a CR network improves with the increasing value of power in a network. In a CR network, the average energy efficiency falls as the total power of the CR network rises to meet the minimum target rate.

In Fig. 11, The simulation-based experimental results reveal that the suggested MOALO approach performs better in terms of SE-EE trade-off compared to MOGOA and SPGA algorithms. It can also be found from the simulation study that EE also increases with the increasing value of SE. In every technique, both objectives are not compromising with each other to improve another objective. In the MOALO algorithm, energy efficiency decreases after (SE = 6.5 bits/s/Hz).In other cases, EE declines at (SE = 7 bits/s/Hz) for the SPGA algorithm and (SE = 6.5 bits/s/Hz) for the MOGOA algorithm. The findings of spectral and energy efficiency for the SA problem have been compared with the MOGOA and SPGA which is represented in Table 4.

Table 4

Comparisons of two objectives ( EE and SE) for SPGA, MOGOA, and MOALO
Comparisons of two objectives using SPGA, MOGOA, and MOALO under SINR 5dB
EE and SE using the SPGA algorithm		EE and SE using the MOGOA algorithm		EE and SE using the MOALO algorithm
EE(bps/Hz/W)	SE (bps/Hz)	EE (bps/Hz/W)	SE (bps/Hz)	EE (bps/Hz/W)	SE (bps/Hz)
1.138*10⁷	3.802	1.225*10⁷	4.352	1.3501*10⁷	4.6147
1.431*10⁷	4.858	1.5184*10⁷	5.408	1.6431*10⁷	5.6707
1.619*10⁷	5.715	1.7064*10⁷	6.265	1.8311*10⁷	6.5277
1.68*10⁷	6.301	1.7674*10⁷	6.851	1.8921*10⁷	7.1137
1.609*10⁷	7.175	1.6964*10⁷	7.725	1.8211*10⁷	7.9877

In the current study, spectral and energy efficiencies associated with SA tasks have been formulated as two objectives optimization problems. To solve this problem two bio-inspired based multi-objective optimization techniques such as MOALO and MOGOA have been employed. The performances achieved by the two proposed methods have been compared with those obtained by the already reported SPGA method. The analyses of simulation-based comparative results demonstrate that the proposed MOALO algorithm-based SA approach outperforms the other two methods. The suggested MO approaches can also be employed to effectively solve the SA task when formulated as three objectives-based optimization problems.

Ethical Approval

This declaration is not applicable.

Conflict of interest

The authors have no conflicts of interest to declare that are relevant to the content of this article.

Competing interests

The authors have no relevant financial or non-financial interests to disclose.

Authors' contributions

All authors contributed equally to the study's conception and design. All authors read and approved the final manuscript.

Funding

This work is supported by the Department of Science and Technology, Govt. of India at KIIT University Bhubaneswar under DST–SERB project no-SPG/2021/001765.

Availability of data and materials

This declaration is not applicable.

Zhang, W., Wang, C. X., Chen, D., & Xiong, H. (2015). Energy–spectral efficiency tradeoff in cognitive radio networks. IEEE Transactions on Vehicular Technology, 65(4), 2208-2218.
Mili, M. R., Musavian, L., Hamdi, K. A., & Marvasti, F. (2016). How to increase energy efficiency in cognitive radio networks. IEEE Transactions on Communications, 64(5), 1829-1843.
Xu, X., Yang, W., Cai, Y., & Jin, S. (2016). On the secure spectral-energy efficiency tradeoff in random cognitive radio networks. IEEE Journal on selected areas in communications, 34(10), 2706-2722.
Kaur, A., & Kumar, K. (2020). A Reinforcement Learning based evolutionary multi-objective optimization algorithm for spectrum allocation in Cognitive Radio networks. Physical Communication, 43, 101196.
Sasikumar, S., & Jayakumari, J. (2020). A novel method for the optimization of Spectral-Energy efficiency tradeoff in 5 G heterogeneous Cognitive Radio Network. Computer Networks, 180, 107389.
Mirjalili, S., Jangir, P., & Saremi, S. (2017). Multi-objective ant lion optimizer: a multi-objective optimization algorithm for solving engineering problems. Applied Intelligence, 46, 79-95.
Ahmadi, B., Ceylan, O., & Ozdemir, A. (2021). Distributed energy resource allocation using multi-objective grasshopper optimization algorithm. Electric Power Systems Research, 201, 107564.
Chuang, C. L., Chiu, W. Y., & Chuang, Y. C. (2021). Dynamic multiobjective approach for power and spectrum allocation in cognitive radio networks. IEEE Systems Journal, 15(4), 5417-5428.
Anumandla, K. K., Sabat, S. L., Peesapati, R., AV, P., Dabbakuti, J. K., & Rout, R. (2021). Optimal spectrum and power allocation using evolutionary algorithms for cognitive radio networks. Internet Technology Letters, 4(4), e207.
Padmanaban, B., & Sathiyamoorthy, S. (2020). A metaheuristic optimization model for spectral allocation in cognitive networks based on ant colony algorithm (M-ACO). Soft Computing, 24, 15551-15560.
Han, R., Gao, Y., Wu, C., & Lu, D. (2018). An effective multi-objective optimization algorithm for spectrum allocations in the cognitive-radio-based Internet of Things. IEEE Access, 6, 12858-12867
Bhardwaj, P., Panwar, A., Ozdemir, O., Masazade, E., Kasperovich, I., Drozd, A. L., ... & Varshney, P. K. (2016). Enhanced dynamic spectrum access in multiband cognitive radio networks via optimized resource allocation. IEEE Transactions on Wireless Communications, 15(12), 8093-8106.
Vimal, S., Khari, M., Crespo, R. G., Kalaivani, L., Dey, N., & Kaliappan, M. (2020). Energy enhancement using Multiobjective Ant colony optimization with Double Q learning algorithm for IoT based cognitive radio networks. Computer Communications, 154, 481-490.
Zhang, X., Zhang, X., & Wu, Z. (2019). Spectrum allocation by wave based adaptive differential evolution algorithm. Ad Hoc Networks, 94, 101969.
Tang, M., & Xin, Y. (2016). Energy efficient power allocation in cognitive radio network using coevolution chaotic particle swarm optimization. Computer Networks, 100, 1-11.
Elhachmi, J., & Guennoun, Z. (2016). Cognitive radio spectrum allocation using genetic algorithm. EURASIP Journal on Wireless Communications and Networking, 2016, 1-11.
Gao, H. Y., & Cao, J. L. (2013). Non-dominated sorting quantum particle swarm optimization and its application in cognitive radio spectrum allocation. Journal of Central South University, 20(7), 1878-1888.
Hoang, A. T., Liang, Y. C., & Islam, M. H. (2009). Power control and channel allocation in cognitive radio networks with primary users' cooperation. IEEE Transactions on Mobile Computing, 9(3), 348-360.
Hu, H., Zhang, H., & Liang, Y. C. (2015). On the spectrum-and energy-efficiency tradeoff in cognitive radio networks. IEEE Transactions on Communications, 64(2), 490-501.
Hei, Y., Zhang, C., Song, W., & Kou, Y. (2019). Energy and spectral efficiency tradeoff in massive MIMO systems with multi-objective adaptive genetic algorithm. Soft Computing, 23, 7163-7179.
Brahma, D., Swayamsiddha, S., & Panda, G. (2022). Efficient Spectrum Allocation in Cognitive Radio using Yellow Saddle Goatfish Algorithm. In 2022 IEEE International Conference on Artificial Intelligence in Engineering and Technology (IICAIET) (pp. 1-6). IEEE.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Development and Performance Evaluation of Multi-Objective Approach-Based Spectrum Allocation in Cognitive Radio Network

Status:

Version 1

Abstract

Figures

1. Introduction

2. Related works

3. System Model

3.1 MOALO Algorithm Employed for Spectrum Allocation In CRN

3.2 SPECTRUM ALLOCATION USING THE PROPOSED MOALO ALGORITHM

4. Simulation Results And Discussion

Conclusion

Declarations

References

Additional Declarations

Status:

Version 1