The spectrum sensing efficiency and energy efficiency of the CRN in the 5G heterogeneous network are trade-offs. The spectrum sensing and spectrum sharing problems are solved by focusing on energy economy when developing the battery-powered CRN network[2, 3]. The existing solutions are primarily focused on using convex optimization to solve the energy efficiency optimization problem in spectrum sensing.

Real-time spectrum sensing, on the other hand, is a non-convex optimization issue. As a result, (ARODN) is created in this study to improve energy efficiency for various spectrum sensing scenarios. The (ARODN) method is used to achieve spectrum sensing and sharing in the CRN network. Figure 1 depicts the proposed algorithm block diagram.

## 3.1 Multiple Attribute Decision Making

MADM approaches are scientific decision-making procedures for choice issues involving a variety of decision qualities [3]. MADM methods are used to choose the best option out of a limited number of options. The option is chosen based on the information provided by each attribute in relation to each alternative. Several factors are known to influence the outcome. An MADM method specifies how attribute values are processed to arrive at a decision. SAW, TOPSIS, GRA, cost function-based, and ETOPSIS are some of the MADM approaches available.

## 3.1.1 Simple Additive Weighting Method

The weighted average is used to calculate the weighted linear combination or a scoring method [2, 3]. The normalized value of each attribute is multiplied by the weight of the linked network, then the weighted sum is added for each network to obtain an assessment score [2, 3]. It includes the following steps:

Step 1: Create a matrix with various attributes.

Step 2: Produce a normalized matrix with many attributes:

Step 3: Determine the weighted average

Step 4: Determine the network's ranking: In a cognitive radio network, the network with the highest value is chosen as the best network

## 3.1.2 Ideal Solution Technique For Order Preference By Similarity

When the characteristics have roughly identical values and their normalized values are relatively close to each other, the TOPSIS approach[3] is more precise than other methods. It entails the following procedures:

Step 1: Create a matrix with various attributes.

Step 2: Normalized values of numerous characteristics are computed.

Step 3: Formulation of a normalized multiple attribute matrix

Step 4: Finding the best positive and negative solution

Step 5: Using the Euclidean distance approach, calculate the distance from the ideal solution.

Step 6: Relative proximity

Step 7. Ranking the networks: In a cognitive radio network, the network with the greatest Hi value is chosen as the best network for spectrum handoff.

## 3.1.3 Grey Relational Analysis Method

The level of available information is employed in GRA for the system's analysis. A white system is one in which all of the information is known, while a black system is one in which all of the information is unknown[10]. The steps for GRA-based optimal network ranking are as follows:

Step 1. To classify the properties of the networks. Delay, data rate, PLR, and price per unit are all characterized as lower is better.

Step 2: Establish upper and lower bounds for attributes: X is a multi-attribute matrix made up of n networks ( X1, X2, ...., Xn )

Step 3: Normalize the attributes: The following equation is used to construct the normalisation matrix of several properties such as delay, PLR, Price, and crosstalk.

$${X}_{i}^{*}\left(j\right)=\frac{{Y}_{j}-{X}_{i}\left(j\right)}{{Y}_{j}-{Z}_{j}}$$

1

The Normalized matrix is represented as

$${A}_{norm}=\left[\begin{array}{ccccc}{X}_{1}^{*}\left(1\right)& {X}_{1}^{*}\left(2\right)& .....& ......& {X}_{1}^{*}\left(n\right)\\ X\left(1\right)& {X}_{2}^{*}\left(2\right)& ......& .......& {X}_{2}^{*}\left(n\right)\\ ....& ......& .....& ......& ....\\ ......& ......& ......& .......& .....\\ {X}_{m}^{*}\left(1\right)& {X}_{m}^{*}\left(1\right)& ......& .......& {X}_{m}^{*}\left(n\right)\end{array}\right]$$

2

Step 4: Calculate the Gray Relationship: The Gray Relationship is calculated as follows: \(G{Rc}_{i}=\frac{1}{1+{\sum }_{j=1}^{m}{A}_{norm}\left|{X}_{j}^{*}\left(n\right)-1\right|}\) (3)

Step 5. Ranking the networks: In a cognitive radio network, the network with the greatest GRci value is chosen as the best network for spectrum handoff.

## 3.1.4 Cost Function Based Method

The cost function-based method [10] picks the best network in CR networks from a pool of feasible networks.

Step 1: Identifying attributes that are reliant on the cost function: The cost function is influenced by a number of factors, which are presented as

$$C{FB}_{i}=f({X}_{iDR}{Y}_{iDR},{X}_{id}{Y}_{id},{X}_{iPLR}{Y}_{iPLR},{X}_{iCT}{Y}_{iCT},{X}_{iP}{Y}_{iP})$$

4

Where \(C{FB}_{i}\) is the cost function of multiple attributes such as \({X}_{iDR},{X}_{id},{X}_{iPLR},{X}_{iCT},{X}_{iP}\) are data rate, delay, PLR, Crosstalk, price. The weight are obtained by the entropy method.

Step 2: Calculate the cost: The cost is determined as follows:

(5)

Step 3: Determine the network's ranking: In a cognitive radio network, the network with the highest CFBi value is chosen as the best network for spectrum handoff.

## 3.1.5 Enhanced TOPSIS Algorithm

Consider the inclination estimations of voice, information, and video administrations while choosing the best framework for handoff range[2, 10]. Delay, Data Rate, PLR, Price, Jitter, Traffic Density, Direction, and Power Consumption are among the many attributes considered.

Step 1: Define the various properties grid

Step 2: Calculate standardised estimations for various traits:

Step 3: Estimation of entropy and deviance

Step 4: Calculate the normalised value of a number of properties.

Step 5: Determine the positive and negative optimum arrangements.

Step 6. The Euclidean distance associated to positive, negative solutions and relative closeness are represented as

Step 7.In an intellectual radio system, the system with the highest ETi estimation is chosen as the optimum system for range handoff.