Swarm intelligence based optimization of energy consumption in cognitive radio network

Abstract

The cognitive radio network provides a pioneered solution to the spectrum scarcity problem and represents a new paradigm for designing intelligent wireless networks. Energy efficient cognitive radio system maintaining reliability holds great importance in the present scenario of wireless communications. In a cognitive radio network, relays are used to enhance energy efficiency as well as to maintain the sensing reliability. Most of the works in the area of cognitive radio network focused on optimization of energy consumed during data transmission only, while neglecting the energy consumed during spectrum sensing. In this paper, an energy efficient multi-relay cognitive radio network is designed, in which both sensing energy and data transmission energy are jointly optimized. Also, optimal values of system parameters like sensing time and amplifying gain of the relays are determined for the energy efficient system. The minimization of the energy consumed under constraints of target throughput and sensing requirements of cognitive radio network is considered as an optimization problem. Swarm intelligence based optimization techniques like particle swarm optimization (PSO), Particle Swarm Optimization with Aging Leader and Challengers (ALCPSO), Human behavior based Particle Swarm Optimization (HPSO) and Whale Optimization Algorithm (WOA) are used to optimize energy consumption in the network. The analysis reveals that the proposed scheme makes the cognitive radio network more energy efficient than conventional schemes.

Keywords

Cognitive radio networks energy efficiency particle swarm optimization spectrum efficiency dynamic spectrum assignment

1. Introduction

In the present times, the numbers of wireless devices are increased tremendously, which causes scarcity in radio frequency spectrum. In a survey conducted by Federal Communication Commission (FCC), the underutilization of spectrum in static allocation policies is revealed [1]. To solve the problem, dynamic spectrum assignment (DSA) schemes are proposed. DSA can be done by implementing policy based intelligent networks known as cognitive radios [2]. These types of networks have been drawing great attention in the field of wireless communications, because of their great spectrum efficiency. In such networks, a secondary user can access the spectrum, if it is unoccupied by the primary user. The network performs spectrum sensing operation to sense the radio environment in order to find the available spectrum band that can be accessed by secondary users. Different types of spectrum sensing techniques have been proposed so far. Some of these techniques are [3]: Energy detector based sensing, Cyclostationarity-based sensing, Radio identification based sensing, Waveform-based sensing, etc. A secondary user can only transmit data when it is found that there is no active primary user present in the sensed spectrum. Liang et al. [4] derived the theoretical formula for the achievable throughput of secondary user and they also determine the optimal sensing time to maximize it under the constraint of minimum detection probability. Rashid et al. [5] presented a particle swarm optimization (PSO)-based scheme to address the tradeoff between sensing time and throughput of secondary users.

Design of energy efficient cognitive radio system maintaining the reliability of the network is an emerging area of research [6]. Relays are used in the network to enhance the spectrum sensing reliability and to increase the throughput of secondary users [7 –12]. Huang et al. [8] proposed an optimal power allocation strategy for maximization of the throughput of secondary users under both sensing reliability and power constraints in cognitive relay networks. Song et al. [9] designed a cognitive relay system that jointly optimizes sensing time and signal-to-noise ratio to maximize energy efficiency. Huang et al. [10] presented an energy-efficient cooperative spectrum sensing with amplify and forward relaying scheme. Minimizations of energy consumption in the data transmission process of multi-relay cognitive radio network were done by Chatterjee et al. [11]. Huang et al. [12] analyzed the tradeoff between sensing performance and data transmission energy of amplify and forward relay scheme considering both the case of fixed and variable amplifying gain. Optimization of both sensing and data transmission times was done to minimize the average energy consumed in cognitive radio network without using relay by Wu et al. [13]. To consider the spectrum-efficiency and the energy efficiency jointly, a utility function was presented by Hu et al. [14]. It is shown in the paper that to maximize the utility function, there exist an optimal number of cooperating secondary users.

Most of the aforementioned works focus on energy consumed in cognitive radio networks during data transmission time only, while neglecting the energy consumed during sensing time. In this paper, an energy efficient multi-relay cognitive radio network is designed, in which both sensing energy and data transmission energy are jointly optimized. The main contributions of the paper are as follows. Firstly, an expression for energy consumption in the network during sensing and data transmission time is presented. Secondly, the existence of signals of primary users both in sensing and data transmission process is considered, which makes the system more relevant to practical scenarios. Thirdly, optimal values of system parameters like sensing time and amplifying gain of the relays are determined for the energy efficient system. The minimization of the energy under the constraint of target throughput and sensing requirements of the cognitive radio network is considered as an optimization problem. Fourthly, swarm intelligence based optimization techniques like PSO [15, 16], ALCPSO [17], HPSO [18] and WOA [19] are used to solve the optimization problem. These optimization techniques are used in this work, because they rely on rather simple concepts and are easy to implement; do not require gradient information and can bypass local optima [16 –20].

The remainder of the paper is organized as follows. In section 2, the system model is proposed. The expression for energy consumption in the cognitive radio network in sensing and data transmission process is presented in section 3. In section 4, specifications of the system parameters for efficient transmission are given. In section 5, problem formulation is done. Results and discussions are presented in section 6 and the paper is concluded in section 7.

2. System model

The proposed system model is shown in Fig. 1, which consists of a primary network and an infrastructure less cognitive radio network. The primary network has a primary source (PU_s), a primary base station (PBS) and a primary destination (PU_d). The cognitive radio network has three components: a secondary source (SU_s), multiple amplify-and-forward (AF) relay nodes and a secondary destination (SU_d).

Fig. 1

System model.

In the proposed system model, the existence of signals of PU_s during data transmission is also considered. This assumption helps to make cognitive radio system aware about the probability of primary sources being active not only during sensing but also during the data transmission time and helps to adjust the values of system parameter of the network to increase the reliability of the system. Also, the transmission process of the cognitive radio network causes interference to PU_d. To mitigate the problem, the amplification gains of the AF relays are kept within suitable ranges. In the Fig. 1, the continuous directed lines signify the data links and the dashed directed lines signify the interference links. The cognitive radio network performs on a frame-by-frame basis i.e., it occupies a channel for frame duration of T_f. The time frame is shown in Fig. 2, which is divided into sensing time (T_s) and data transmission time (T_d). The data transmission time is again divided into two equal time slot T₁ and T₂. The sensing operation is executed during T_s to find whether any signal from the primary source is present at the particular channel. The energy detection based spectrum sensing scheme [4] is used in this work. If there is no signal from primary sources present in the channel then SU_s will transmit signals to SU_d via AF relays. During T₁, SU_s transmits to all AF relays and during T₂, the relays amplify and forward the received signals to SU_d.

Fig. 2

Time frame of the system.

3. Energy consumption in sensing and data transmission

In this section, the expression for energy consumed in the cognitive radio network during sensing and data transmission process is presented.

3.1. Energy consumption during data transmission time

Considering the time frame model as shown in Fig. 2, data transmission time is divided into two time slots: T₁ and T₂, the energy consumed in the cognitive radio network at each time of the slots is determined.

3.1.1. Energy consumed during T₁

As discussed in section 2, SU_s transmits signals to the relays during T₁. But, there is a chance of arrival of the signals of PU_s during time T₁, which carries a chance of occurrence of misdetection. The misdetection causes interference from PU_s to the cognitive radio network. Two scenarios can be assumed for the transmission process of SU_s during T₁ as follows.

(i) Non-occurrence of false alarm: The primary source PU_s doesn’t transmit any signal and SU_s doesn’t generate false alarm. Then the transmission power per unit time is given by $P_{1} = P (H_{0}) (1 - P_{f}) (P_{c} + P_{n})$ (1) where P (H₀) is the probability of the absence of the signals of the PU_s in the sensed channel. P_f is the probability of false alarm, P_c is the transmission power of SU_s and P_n is noise power.

(ii) Occurrence of misdetection: The primary source PU_s transmits signals during T_s and SU_s doesn’t detect the signal. As a result, SU_s transmits signals to the relays during T₁ although the signal from PU_s is present. In this case, transmission power per unit time is given by $P_{2} = P (H_{1}) (1 - P_{d}) (G_{pc} P_{p} + P_{c} + P_{n})$ (2) where P (H₁) is the probability that primary sources being active, P_p is the transmission power of PU_s and G_pc is channel gain between the links PU_s to SU_s.

Considering above scenarios the average power transmitted during T₁ is given by

$\begin{matrix} P_{T_{1}} & = P_{1} + P_{2} = P (H_{0}) (1 - P_{f}) (P_{c} + P_{n}) \\ + P (H_{1}) (1 - P_{d}) (G_{pc} P_{p} + P_{c} + P_{n}) \end{matrix}$ (3)

Average energy consumed during T₁ is given by $E_{T_{1}} = P_{T_{1}} T_{1} = P_{T_{1}} (T_{f} - T_{s} - T_{2})$ (4) where T₁ = T_f - T_s - T₂, which can be obtained from Fig. 2. From (3) and (4), the value of E_{T
₁} is given as,

$\begin{matrix} E_{T_{1}} = [P (H_{0}) (1 - P_{f}) (P_{c} + P_{n}) + P (H_{1}) \\ (1 - P_{d}) (G_{pc} P_{p} + P_{c} + P_{n})] (T_{f} - T_{s} - T_{2}) \end{matrix}$ (5)

3.1.2. Energy consumption during T₂

As discussed in section 4.2, all relays amplify and forward the signals received in T₁ to SU_d. If the primary source PU_s is active during T₂ then it causes interference to the relays. In this case, the interference signals from PU_s is also amplified and forwarded by the relays. It is considered that all relays have equal gain i.e., β₁ = . . . = β_i = . . . = β_M = β. Two scenarios can be assumed for the transmission process of ith relay (CR_i) during T₂ as follows.

(i) Non-occurrence of interference from PU_s: PU_s does not transmit any signal to CR_i. Therefore, CR_i receives signal from SU_s only. In this case transmission power of CR_i per unit time is given as

$\begin{matrix} P_{3} & = β_{i} P (H_{0}) (G_{{cr}_{i}} P_{T_{1}} + P_{n}) \\ = β_{i} [1 - P (H_{1})] (G_{{cr}_{i}} P_{T_{1}} + P_{n}) \end{matrix}$ (6) where β_i is the amplification gain of ith relay, G_{cr
_i} is channel gain between the link SU_s and ith relay.

(ii) Occurrence of interference from PU_s: PU_s transmits signals to CR_i. Therefore, CR_ireceives signals from both SU_s and PU_s. In this case transmission power of CR_i per unit time is given as $P_{4} = β_{i} P (H_{1}) (G_{{cr}_{i}} P_{T_{1}} + G_{{pr}_{i}} P_{p} + P_{n})$ (7) where G_{pr
_i} is channel gain between the link PU_s and ith relay.

Considering the above scenarios the average transmission power of SU_s during T₂ is given by

$\begin{matrix} P_{T_{2}} = P_{3} + P_{4} = β_{i} [1 - P (H_{1})] (G_{{cr}_{i}} P_{T_{1}} + P_{n}) \\ + β_{i} P (H_{1}) (G_{{cr}_{i}} P_{T_{1}} + G_{{pr}_{i}} P_{p} + P_{n}) \\ = β_{i} [P (H_{1}) G_{{pr}_{i}} P_{p} + G_{{cr}_{i}} P_{T_{1}} + P_{n}] \end{matrix}$ (8)

Therefore, the average energy consumed by all relays during T₂ is given by $E_{T_{2}} = \sum_{i = 1}^{M} P_{T_{2}} T_{2} = \sum_{i = 1}^{M} P_{T_{2}} (T_{f} - T_{s} - T_{1})$ (9) where T₂ = T_f - T_s - T₁, that can be observed in Fig. 2. From (8) and (9), the value of E_{T
₂} can be obtained as

$\begin{matrix} E_{T_{2}} = \sum_{i = 1}^{M} β_{i} [P (H_{1}) G_{{pr}_{i}} P_{p} + G_{{cr}_{i}} P_{T_{1}} + P_{n}] \\ (T_{f} - T_{s} - T_{1}) \end{matrix}$ (10)

3.2. Total energy consumption

Total energy consumed during sensing and data transmission process is the sum of energy consumed during T_s, T₁ and T₂ respectively. $E_{T} = E_{T_{s}} + E_{T_{1}} + E_{T_{2}}$

$\begin{matrix} or, E_{T} = P_{sn} T_{s} + \sum_{i = 1}^{M} β_{i} [P (H_{1}) G_{{pr}_{i}} P_{p} \\ + G_{{cr}_{i}} P_{T_{1}} + P_{n}] (T_{f} - T_{s} - T_{1}) \\ - 20 pt + [P (H_{0}) (1 - P_{f}) (P_{c} + P_{n}) + P (H_{1}) (1 - P_{d}) \\ (G_{pc} P_{p} + P_{c} + P_{n})] (T_{f} - T_{s} - T_{2}) \end{matrix}$ (11)

The objective is to minimize the average energy consumed in the cognitive radio network for sensing and data transmission process satisfying the constraints of sensing and throughput. Also, maintaining the interference to the PU_d below a predefined threshold value.

4. Specifications of system parameters for efficient transmission

The objective of the paper is to optimize the average energy consumption in the cognitive radio network during sensing and data transmission process. While doing the optimization, the system parameters like the probability of detection, the probability of false alarm, sensing time and throughput are needed to be kept within acceptable ranges so that reliability of the system can be maintained. In this section, the acceptable ranges of some of the system parameter are presented.

4.1. Sensing time

The minimum number of sensing samples (N_min) required to meet the target probability of detection and the target probability of false alarm is given as [5] $N_{min} = \frac{1}{γ^{2}} {[Q^{- 1} (\bar{P_{f}}) - Q^{- 1} (\bar{P_{d}}) \sqrt{2 γ + 1}]}^{2}$ (12) where $\bar{P_{d}}$ is the target probability of detection, $\bar{P_{f}}$ is the target probability of false alarm and γ is the sensed signal-to-noise ratio (SNR) by SU_s. Considering τ as sampling time, the minimum sensing time T_{s
_min} required to sense N_min number of samples is given as

$\begin{matrix} T_{s_{min}} & = τ N_{min} = τ \frac{1}{γ^{2}} \\ {[Q^{- 1} (\bar{P_{f}}) - Q^{- 1} (\bar{P_{d}}) \sqrt{2 γ + 1}]}^{2} \end{matrix}$ (13)

4.2. Amplifying gain

The average achievable throughput of the cognitive radio network can be expressed as [4]

$\begin{matrix} R = \frac{(T_{f} - T_{s})}{T_{f}} P (H_{0}) (1 - P_{f}) {log}_{2} \\ (1 + \frac{β \sum_{i = 1}^{k} G_{r_{i} d} G_{{cr}_{i}} P_{c}}{β \sum_{i = 1}^{k} G_{r_{i} d} P_{n} + P_{n}}) \end{matrix}$ (14) where P (H₀) is the probability that primary sources being inactive. In the proposed system model, average achievable throughput is maintained above target throughput $\bar{R}$ for $P_{d} \geq {\bar{P}}_{d}, P_{f} \leq {\bar{P}}_{f}$ and T_s ≥ T_{s
_min} i.e., $\begin{matrix} \frac{T_{f} - T_{s}}{T_{f}} P (H_{0}) (1 - P_{f}) \\ {log}_{2} (1 + \frac{β \sum_{i = 1}^{M} G_{r_{i} d} G_{{cr}_{i}} P_{c}}{β \sum_{i = 1}^{M} G_{r_{i} d} P_{n} + P_{n}}) \geq \bar{R} \end{matrix}$ $or, β \geq \frac{q}{\sum_{i = 1}^{M} G_{r_{i} d} G_{{cr}_{i}} P_{c} - q \sum_{i = 1}^{M} G_{r_{i} d} P_{n}} = β_{min}$ (15) where $q = 2^{\frac{\bar{R}}{(\frac{T_{f} - T_{s}}{T_{f}}) P (H_{0}) (1 - P_{f})}} - 1$ . During the transmission process, there may be interference to the PU_d from the cognitive radio network. The interference power can be estimated as

$\begin{matrix} I = \frac{T_{f} - T_{s}}{T_{f}} P (H_{1}) (1 - P_{d}) β \sum_{i = 1}^{M} \\ (G_{pc} G_{{cr}_{i}} P_{p} + G_{{pr}_{i}} P_{p} + G_{{cr}_{i}} P_{c} + P_{n}) G_{r_{i} Pd} \end{matrix}$ (16)

The interference to the PU_d is maintained below the threshold value $\bar{I}$ i.e., $\begin{matrix} \frac{T_{f} - T_{s}}{T_{f}} P (H_{1}) (1 - P_{d}) β \sum_{i = 1}^{M} (G_{pc} G_{{cr}_{i}} P_{p} \\ + G_{{pr}_{i}} P_{p} + G_{{cr}_{i}} P_{c} + P_{n}) G_{r_{i} Pd} \leq \bar{I} \end{matrix}$

$\begin{matrix} or, β \leq \frac{\bar{I}}{(\frac{T_{f} - T_{s}}{T_{f}}) P (H_{1}) (1 - P_{d}) \sum_{i = 1}^{M} (G_{pc} G_{{cr}_{i}} P_{p} + G_{{pr}_{i}} P_{p} + G_{{cr}_{i}} P_{c} + P_{n}) G_{r_{i} Pd}} = β_{max} \end{matrix}$ (17)

The minimum and maximum allowable limits for amplifying gain of the relays are given in (15) and (17) respectively.

5. Problem formulation

The objective is to minimize the average energy consumption in cognitive radio network for sensing and data transmission process satisfying the constraints of sensing and throughput. Also, maintaining the interference to the PU_d below a threshold value. Mathematically, the optimization problem can be formulated as

$\begin{matrix} min E_{T} \\ s . t . P_{d} \geq {\bar{P}}_{d}, P_{f} \leq {\bar{P}}_{f}, T_{s} \geq T_{s_{min}}, \\ β_{min} \leq β \leq β_{max} \end{matrix}$ (18)

In Table 1, the list of variable parameters and their specifications considered in the optimization process are given. The minimum and maximum ranges of the amplifying gain (β) given in Table 1 are determined with the help of (15) and (17), respectively. Similarly, the minimum limit of sensing time T_s is determined with the help of (13). The objective of the optimization process is to minimize the sensing and data transmission energy of cognitive radio network in such a way that system parameters like probability of detection, probability of false alarm, amplifying gain of the relays, sensing time and transmission power of secondary users lie within acceptable limits.

Table 1

List of parameters for optimization

Parameter	${\bar{P}}_{d}$	${\bar{P}}_{f}$	T_{s _min} (ms)	β _min	β _max	$\bar{R}$ (bps/Hz)	$\bar{I}$ (dBW)
Values	90%	5%	5.4	2.12	16.26	0.25	–20

6. Results and discussions

In this paper, analysis of energy consumption in cognitive radio network under different channel gain conditions is done. The transmission power of the primary sources (P_p) and the cognitive source (P_c) are considered to be 0 dBW. The power consumption during the sensing process is taken as –15 dBW. The frame duration and sampling time are considered as 224.25 ms and 1 ms respectively.

6.1. Variations of energy consumed with different parameters

In Fig. 3, it is shown that energy consumption in cognitive radio network decreases with the increase of the transmission power of the primary source. This is due to the fact that the increase in transmission power of primary sources lowers the amplifying gain of the relays, which in turn decrease the energy consumption in the system. Also, it is noticed from Fig. 3, that the increase in number of relays makes the system more energy efficient.

Fig. 3

Energy consumed versus transmission power of PU_s.

In Fig. 4, the variations of energy consumed by the network with transmission power of SU_s for different number of relays are shown. As transmission power of SU_s increases, the more energy is consumed by the system in the data transmission process which increased overall energy consumption. It is also observed that energy consumption decreases with an increase in the number of relays.

Fig. 4

Energy consumed versus transmission power of SU_s.

6.2. Optimization of the energy consumption using swarm intelligence based techniques

To fulfill the objective, swarm intelligence based algorithms like PSO, ALCPSO, HPSO and WOA are used. The implementation of swarm intelligence based algorithms in practical scenarios are involved the following steps.

Step 1: Dimension and population size of the swarm are initialized. Also, each particle’s position and velocity are initialized to independent random values.

Step 2: For each iteration k, the velocity and position of the ith particle in n dimension search space are updated as follows [10]

$\begin{matrix} v_{in}^{k + 1} & = ω . v_{in}^{k} + c_{1} . {rand}_{1}^{k} . ({pbest}_{in}^{k} - p_{in}^{k}) \\ + c_{2} . {rand}_{2}^{k} ({gbest}_{n}^{k} - p_{in}^{k}) \end{matrix}$ (19)

$p_{in}^{k + 1} = p_{in}^{k} + v_{in}^{k + 1}$ (20) where the indices i, d and k indices vary in the range {1….,Number of Particles (P)}, {1….,maximum dimension (n)} and { 1……,maximum iteration number (K)} respectively, w is inertia factor, c₁ and c₂ are the acceleration coefficients, rand₁ and rand₂ are two uniformly distributed random numbers between 0 and 1, ${pbest}_{in}^{k}$ is the local best position ith particle in kth iteration, ${gbest}_{n}^{k}$ is the global best position in the swarm till kth iteration.

Table 2

Computational time complexity of the algorithms

Algorithm	Complexity
PSO	O (n × P + P × (2 + log ₂P))
HPSO	O (n × P + P × (2 + 2 × log ₂P))
ALCPSO	O (n × P + P × (2 + 8 × n + log ₂P))
WOA	O (n × P + P × (1 + 2 × n + log ₂P))

Step 3: The current fitness value of each particle is compared with its previous best value. If the current fitness value is better than the previous best value, then the current value is set as pbest. The gbest is set as the best of all pbest values.

Step 4: Update the position and velocity of each particle using (19) and (20) respectively till termination condition is satisfied.

Step 5: Evaluate the optimal solution, which is the best gbest value of all iterations.

The time complexity [21, 22] of swarm intelligence based algorithms is O (n × P + P × cof), where n is the dimension of the problem and P is the population size and cof is the cost of the objective function. The term cof varies according to the optimization process of the algorithms. In the analysis of the complexity, n × P signifies the computation of solutions for a population of size P with n dimensions. P × log ₂P signifies the sorting of the best solution among P numbers of solutions. The computational time complexity of the algorithms to solve the optimization problem is given in Table 2, which reveals that PSO is better in terms of computational time complexity than the other algorithms.

The results of the optimization processes of energy consumption in the cognitive radio network are represented with the help of convergence plots of the aforementioned swarm intelligence based techniques. The comparison of convergence plots of all the algorithms at relatively low channel gain conditions, i.e., G_pc = G_{pr
_i} = G_{r
_i
pd} = -10 dB, G_{cr
_i} = G_{r
_i
d} = -4 dB is shown in Fig. 5 and the obtained values of system parameters are given in Table 3. Similarly the comparative analysis of the energy consumption at relatively high channel gain conditions i.e., G_pc = G_{pr
_i} = G_{r
_i
pd} = 3 dB, G_{cr
_i} = G_{r
_i
d} = 3 . 5dB is shown in Fig. 6 and the obtained values of system parameters are listed in Table 4. From the comparative analysis given in Figs. 5, 6, Tables 3 and 4, it can be concluded that the WOA has reported the best results for the concerned problem of minimum energy consumption in cognitive radio network satisfying the reliability constraints of the system both at low and high channel gain conditions.

Fig. 5

Convergence plots of the algorithms at low channel gains.

Fig. 6

Convergence plots of the algorithms at high channel gains.

Table 3

Results of optimization process at low channel gains

Parameter	PSO based optimum values	PSO based optimum energy (mJ)	ALCPSO based optimum values	ALCPSO based energy (mJ)	HPSO based values	HPSO based optimum energy (mJ)	WOA based optimum values	WOA based optimum energy (mJ)
P _d	91.82%	1.604	90%	1.083	93.86%	0.916	95%	0.819
P _f	3.60%		4.87%		4.68%		5%
T _s	8.3 ms		9.8 ms		12 ms		12 ms
β	2.12		2.12		2.12		2.12

Table 4

Results of optimization process at high channel gains

Parameter	PSO based	PSO based	ALCPSO	ALCPSO	HPSO based	HPSO based	WOA based	WOA based
optimum	optimum	based	based optimum	optimum	optimum	optimum	optimum
values	energy (mJ)	optimum values	energy (mJ)	values	energy (mJ)	values	energy (mJ)
P _d	92.41%	4.094	95%	3.291	93.86%	3.450	95%	3.252
P _f	5%	4.52%	5%	5%
T _s	8 ms	12 ms	10 ms	12 ms
β	3.91	3.91	3.91	4.02

6.3. Performance comparison

In Fig. 7, a three-dimensional plot of interference to PU_d and system throughput under different values of the transmission power of the cognitive radio network is shown. From Fig. 7, it is revealed that with optimal values of parameters obtained from the optimization process of WOA, the values of interference and throughput lies within their acceptable limits as given in Table 1. Also, the analysis reveals that the WOA based optimization scheme makes the network more energy efficient than the schemes used in [10 –12] as shown in Fig. 8.

Fig. 7

Three dimensional plot interference, throughput and transmission power.

Fig. 8

Comparison of energy consumption.

The performance comparison in terms of achievable throughput in the network of the proposed WOA based optimization scheme with schemes used in [10 –12] is shown in Fig. 9. From Fig. 9, it is seen that the proposed WOA based scheme enables the cognitive radio network to achieve greater throughput than the other schemes.

Fig. 9

Average achievable throughput versus average transmitted power.

7. Conclusion

The cognitive radio network has capability to meet the demand of high spectrum utilization in 5 G communications. Because of the ability to adapt its parameters according to requirements of radio environment, the cognitive radio network can be used in military operations, health care services, disaster management services etc. In this paper, optimization of energy consumption in cognitive radio network in sensing and data transmission process is done using PSO, ALCPSO, HPSO and WOA. The optimization is done both in low and high channel gain conditions. Simulation results show that the system optimized with WOA is more efficient than that of the other algorithms in terms of energy consumption satisfying the system reliability. Also, it is shown that the WOA based optimization scheme makes the network more efficient in terms of energy consumption and throughput than the conventional schemes.

References

Marcus ,

Burtle ,

Franca ,

Lahjouji and

McNeil , “Report of the unlicensed devices and experimental licenses working group.” Federal Communications Commission Spectrum Policy Task Force (2002).

Simon

Haykin

, “Cognitive radio: brain-empowered wireless communications.” IEEE Journal on Selected Areas in Communications 23.2 (2005), 201–220.

Zeng ,

Y.C.

Liang ,

A.T.

Hoang and

Zhang , “A review on spectrum sensing for cognitive radio: challenges and solutions.” EURASIP Journal on Advances in Signal Processing 2010.1 (2010), 381–465.

Y.C.

Liang ,

Zeng ,

E.C.

Peh and

A.T.

Hoang , “Sensing-throughput tradeoff for cognitive radio networks.” IEEE transactions on Wireless Communications 7.4 (2008), 1326–1337.

R.A.

Rashid ,

A.H.F.B.A.

Hamid ,

Fisal ,

S.K.

Syed-Yusof ,

Hosseini ,

Lo and

Farzamnia , “Efficient in-band spectrum sensing using swarm intelligence for cognitive radio network.” Canadian Journal of Electrical and Computer Engineering 38.2 (2015), 106–115.

Xiong ,

Lu and

G.Y.

Li , “Energy-Efficient Spectrum Access in Cognitive Radios.” IEEE Journal on Selected Areas in Communications 32(3) (2014), 550–562.

Zhang ,

Jia and

Zhang , “Cooperative relay to improve diversity in cognitive radio networks.” IEEE Communications Magazine 47.2 (2009), 111–117.

Huang ,

Chen and

Zhang , “Optimal power allocation for spectrum sensing and data transmission in cognitive relay networks.” IEEE Wireless Communications Letters 1.1 (2012), 26–29.

Song ,

Zhang and

Yubin , “Energy efficiency and throughput optimization of cognitive relay networks.” CIT. Journal of Computing and Information Technology 22.3 (2014), 151–158.

10.

Huang ,

Chen ,

Zhang and

Zhao , “Energy-efficient cooperative spectrum sensing with amplify-and-forward relaying.” IEEE Communications Letters 16.4 (2012), 450–453.

11.

Chatterjee ,

S.P.

Maity and

Acharya , “Energy efficient cognitive radio system for joint spectrum sensing and data transmission.” IEEE Journal on Emerging and Selected Topics in Circuits and Systems 4.3 (2014), 292–300.

12.

Huang ,

Chen ,

Zhang , and

H.H.

Chen , “Sensing-energy tradeoff in cognitive radio networks with relays.” IEEE systems journal 7.1 (2013), 68–76.

13.

Wu and

D.H.

Tsang , “Energy-efficient spectrum sensing and transmission for cognitive radio system.” IEEE Communications Letters 15.5 (2011), 545–547.

14.

Hu ,

Zhang ,

Yu and

Chen , “Spectrum-energy-efficient sensing with novel frame structure in cognitive radio networks.” AEU-International Journal of Electronics and Communications 68.11 (2014), 1065–1072.

15.

Kennedy and

Eberhart , “Particle swarm optimization.” In: Proceedings of IEEE International Conference on Neural Networks IV, pages (1995) Vol. 1000.

16.

Shi and

Eberhart , “A modified particle swarm optimizer.” Evolutionary Computation Proceedings, 1998. IEEE World Congress on Computational Intelligence., The 1998 IEEE International Conference on. IEEE, 1998.

17.

W.N.

Chen ,

Zhang ,

Lin ,

Chen ,

Z.H.

Zhan ,

H.S.H.

Chung and

Y.H.

Shi , “Particle swarm optimization with an aging leader and challengers.” IEEE Transactions on Evolutionary Computation 17.2 (2013), 241–258.

18.

Liu ,

Xu ,

G.Y.

Ding and

Y.B.

Sun , “Human behavior-based particle swarm optimization.” The Scientific World Journal 2014 (2014).

19.

Mirjalili and

Lewis , “The whale optimization algorithm.” Advances in Engineering Software 95 (2016), 51–67.

20.

A.J.

Gogoi ,

C.L.

Singh ,

Nath and

K.L.

Baishnab , “Optimization of sensing time in energy detector based sensing of cognitive radio network.” International Journal of Applied Engineering Research 11.6 (2016), 4563–4568.

21.

Sharma, Nitin,

Anand

Kamat Tarcar ,

Varghese

Antony Thomas and

K.R.

Anupama , “On the use of particle swarm optimization for adaptive resource allocation in orthogonal frequency division multiple access systems with proportional rate constraints.” Information Sciences 182(1) (2012), 115–124.

22.

Gogoi, Ashim Jyoti,

Naushad M.

Laskar ,

Ch Lison

Singh ,

Krishna Lal

Baishnab and

Koushik

Guha , “Throughput Optimization in Multi-user Cognitive Radio Network using Swarm Intelligence Techniques.” Journal of Information Science & Engineering 34(4) (2018).

Swarm intelligence based optimization of energy consumption in cognitive radio network

Abstract

Keywords

1. Introduction

2. System model

3.1. Energy consumption during data transmission time

3.1.1. Energy consumed during T1

4.1. Sensing time

6.1. Variations of energy consumed with different parameters

References

3.1.1. Energy consumed during T₁