The article delves into the challenges and solutions for enhancing indoor path loss and received signal power in 5G mmWave communication systems. It highlights the use of a multi-objective genetic algorithm (MOGA) to optimize path loss models, focusing on indoor scenarios such as offices and shopping malls. The paper compares various path loss models and optimization techniques, showcasing the superior performance of MOGA in maximizing received signal power and minimizing path loss. The research also discusses the impact of antenna location, type, and operating frequency on path loss, providing a detailed analysis of indoor LOS path loss models. The findings are supported by numerical and simulation results, making the article a valuable resource for professionals seeking to improve indoor wireless communication performance.
AI Generated
This summary of the content was generated with the help of AI.
Abstract
The signal strength in 5G mobile communication systems is significantly influenced by the surroundings, with key factors including the path difference, operating frequency, and obstructions at specific locations. Consequently, planning a communication system that can deliver improved signal strength becomes highly challenging. To address this issue, indoor path loss models are employed to estimate signal loss in different environments, frequencies, and distances. This paper introduces an intelligent multi-objective genetic algorithm aimed at enhancing path loss and received signal power. A comparative analysis is conducted to evaluate the performance of the proposed intelligent optimization algorithm against the traditional approach. The path loss and received power of various scenarios are estimated using various path loss models. The 5GCM indoor officce, 5GCM InH shopping mall, 3GPP TR 38.91 InH office, mmMAGIC InH office, METIS InH shopping mall, and IEEE 802.11 ad InH office indoor path loss models estimates the path loss of 62.37 dB, 62.15 dB, 63.12 dB, 50 dB, 55.18 dB, and 52.89 dB in traditional approach and 36.87 dB, 35.86 dB, 36.84 dB, 68.80 dB, 36.23 dB and 33.94 dB using GA approach and received powers of \(-12.17~dBm, -11.37~dBm, -12.17~dBm, -5.80~dBm,\)\(-12.24~dBm\) and \(-8.68~dBm\) in traditional approach and 26.13 dBm, 27.14 dBm, 26.15 dBm, \(-5.80~dBm\), 26.75 dBm and 29.05 dBm using GA approach repectively. The 5GCM and 3GPP models produces the path loss difference above 25 dB and mmMAGIC, METIS and IEEE models produces a path loss below 19 dB. Except mmMAGIC model, all models produces the recceiver power difference above 37 dBm. Therefore, the highest path loss difference of 26 dB is observed in 5GCM InH shopping mall model and the highest reccieved power difference of 39.01 dBm is observed in METIS InH shopping mall model. The results clearly demonstrate that the proposed intelligent optimization approach outperforms the traditional approach across various indoor scenarios.
Notes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
1 Introduction
Together with data consumption, user demands, and user expectations, the use of wireless devices today is expanding quickly. New services and use cases to satisfy these demands and expectations were made possible by advancements in wireless evaluations [1, 2]. It has been suggested that the fifth generation (5G) of wireless technology will provide vital services including maximum throughput, high bandwidth, increased capacity, energy efficiency, and low latency [3‐7]. New technologies have been proposed to meet these essential requirements for 5G and beyond, such as millimetre wave (mmWave) spectrum for wide bandwidths, small cells for improved coverage, extreme mobile broadband (eMBB) and multi-input multi-output (MIMO) for high data rates and low latency, respectively [8‐11].
The millimeter wave and terahertz (THz) frequency ranges offer vast amounts of available bandwidth, making them highly attractive for providing ultra-fast data speeds of multi-gigabits per second (Gbps) and alleviating congestion in lower frequency mobile data traffic, specifically below 6 GHz [12, 13]. However, high-frequency communications present more challenging propagation limits compared to the below-6 GHz band utilized by previous generations of mobile networks [14]. Beyond 6 GHz, mmWaves are highly susceptible to shadowing, atmospheric influences, and exhibit poor propagation characteristics in most solid materials [15]. Heavy rain, hail, and foliage are also other limiting factors at high frequencies, and can directly affect network quality of service [16]. As a result of their ease of access and high propagation rates, high-frequency bands have gained popularity in recent years [17].
Advertisement
Several measurement campaigns have been carried out throughout the past thirty years with the goal of learning in-depth information on the spatial and temporal channel properties and, as a result, creating novel techniques to utilize the mmWave frequency bands. The development of channel models at mmWave and THz frequencies has been a current research priority for many research groups and industrial organizations working on 5G and beyond 5G networks that use mmWaves and THz frequency bands [18].
Large-scale propagation models are used to estimate the path loss and receiver signal strength at mmwave frequency bands, forming the basis for radio planning tools used in the network design and optimization process. These models can be divided into three categories: empirical, semi-deterministic, and deterministic, and are established using different approaches [19]. Based on where the antennas are placed, these models are further divided into interior and outdoor models. Similar mechanisms, like reflection, diffraction, and scattering, govern both indoor and outdoor propagation.But, because of the disparity in propagation conditions, their radio channels are purely different. For instance, in outdoor environments, obstructions like trees, buildings, advertisement boards, etc., and in indoor settings, the presence of walls, floors, furniture, humans, etc., may have a significant effect on the propagation of radio waves. Hence, antenna location, its tilt angle and type plays a very important role in estimation of path loss and energy efficiency [20‐22]. For indoor and outdoor applications, appropriate path loss models are necessary to accurately estimate the path loss (PL) and received signal power. In this paper, we considered the indoor scenario and its path loss models.
The literature has extensively explored indoor propagation models. However, there are currently very few tutorials and research papers that address the mmWave and THz spectrums in indoor scenarios. Indoor path loss was estimated using machine learning (ML) algorithms by optimizing the number of access points [23]. This approach provides better coverage and covers up to 99.4% of the receiver locations. The range estimation using the single-stage path loss model causes more errors. So, a multi-stage path loss estimation model was considered in the indoor environment to provide better accuracy in terms of distance and location [24]. A genetic algorithm (GA) is proposed to reduce the complexity due to the multi-stage model and provide a better system performance of up to 31.9%.
The path loss in a high-speed rail system is estimated using the GA and generalized reduced gradient approaches. These two approaches provide better signal strength even though there are many unknown parameters that cause path loss [25]. The Cost-231 Hata model was optimized in [26] using the weighted least square approach, genetic algorithm (GA), and hybrid GA. Among these methods, the hybrid model yields the lowest values for mean square error (MSE), root mean square error (RMSE), and PL. Cuckoo search optimization algorithm was used to optimize the path loss, which minimized the mean square error and mean absolute percentage error [27]. The GWO approach generated the lowest error statistics when compared to the GA and PSO algorithms. Using the least square optimization method, the path loss of empirical models, including Cost-231, ECC-33, and SUI models, was estimated. The optimized path loss model was identified as the Cost-231 model [28]. In comparison to the other models, the Cost-231 model produced the lowest path loss and error statistics.
Advertisement
Indoor localization techniques for IoT and real time tracking are examined in [29, 30]. An intelligent indoor localization system was considered using the Internet of Things to maximize the system’s performance [31]. GA was used to estimate the exact location of the 5G base station to enhance the coverage, throughput, and transmission power of the network [32]. Fuzzy logic based spectrum hand-off is considered to improve spectrum utilization and enhance throughput [33]. It provides an overview of indoor localization techniques for various applications. In an indoor environment, the path loss depends on the obstructions present in the surroundings. To maximize the received signal strength, the loss should be minimized. Therefore, intelligent optimization algorithms are to be used to minimize the path loss and maximize the receiver’s power.
Comparison analysis of path loss was conducted using wavelet-GA, GA and COST-231 models [34]. From the results it was observed that the wavelet-GA produces 99.2% enhanced path loss performance. In [35], the GA, PSO and improved cost-hatta models were used to estimate the path loss. The obtained path loss values are compared with the values of machine learning algorithms and observed that the improved cost-hatta model produces least RMSE of 1.52. In wireless ad hoc networks, an optimized stable clustering algorithm was proposed to reduce the clustering overhead due to changes in the cluster head. An additional backup node was introduced to serve as a cluster head in the event of failure of the actual cluster node [36]. A priority based clustering algorithm was proposed to reduce the clustering overhead in the selection of suitable cluster head [37]. A weighted optimization algorithm, priority based weighted optimization algorithm, and stable clustering optimization algorithms are analysed and compared [38]. The summary of the surveyed path loss models and optimization approaches are listed in Table 1.
A multi-stage path loss model was use dto enhance the accuracy interms of location and distance. The complexity of this model was optimized with the use of GA and enhanced the overall system performance
The path loss, RSME, MSE, SE are estimated using GA and hybrid GA. It was observed that the hybrid GA provides better path loss and least RSME compared to GA
Least mean square algorithm was used to optimize the path loss of Cost-231, ECC-33, and SUI models and observed that Cost-231 model produces lowest path loss
Path loss was estimated and compared using the GA, wavelet-GA and cost-231 models and observed that the cost-231 model provides better path loss compared to remaining two models
Path loss was estimated and compared using the GA, PSO and cost hatta models and observed that the cost hatta model provides better path loss compared to remaining two models
In these papers, the authors enhanced the path loss by optimizing the accesspoints, reducing
the complexity and using an optimization algorithms like GA, Wavelet-GA, hybrid GA, least square,
etc. Hence, our aim is to find the optimized path loss model in the indoor environment to enhane the
path loss and RSRP using MOGA.
Our paper
Indoor
In this paper, we considered the traditional simulation based approach and GA for the estimation of path loss and receiver power. The optimization of path loss model to enhance the receiver power and reduction of path loss is considered as a major objective
Multi-objective optimization algorithm (MOOA) was used to minimize the path loss and maximize the power received. In [39], authors considered MOOA to optimize the path difference between transmitter and receiver in an underwater communication system. The path loss and power density are balanced at an optimal distance of 36 m. In [40], authors considered MOGA along with weighted sum optimization techniques to optimize the node location for a given topology in wireless sensor networks (WSN). In [41], MOOA is used to optimize power consumption, spectrum usage, and exposure. The proposed MOOA improved all parameters by more than 27% compared to the conventional method. In [42], MOOA is considered to reduce the energy consumption and improve the energy efficiency of an air-to-air communication system. Using non-dominated GA along with MOOA improved system performance compared to the conventional GA approach.
Optimization of Energy consumption, coverage, power loss, and packet loss rate are considered the major parameters, and conducted a survey of MOOA in WSN [43]. 5G network topology is designed using the MOOA approach, to maximize coverage capacity, by minimizing up-link and down-link exposure [44]. In [45], authors considered MOOA to enhance the performance of MIMO systems in 5G networks. With the use of multi-objective algorithms with GA and PSO, network performance is optimized for multiple outputs. The performance of power amplifiers and antennas in a MIMO 5G system are optimized using MOOA [46]. In the literature authors used ant colony optimization (ACO), genetic algorithm, particle swarm optimization (PSO), and grey wolf optimization (GWO) to enhance path loss in various environments [19]. These algorithms are used to optimize more than one objective function, whereas the multi objective genetic algorithm is used to optimize more than one objective function. It provides a powerful and flexible approach for solving a multi objective optimization. Therefore, MOGA is used to improve the overall system performance by optimizing the number of antennas, power amplifiers, energy consumption and spectrum usage.
The path loss and receiver power are affeccted by many parameters like transmission power, antenna type and location of the antenna, path difference, operating frequency and the optimization parameters like population size, selection rate and mutation rate. Hence, we will employ the multi-objective genetic algorithm (MOGA) to maximize the received signal power and minimize the path loss in a 5G small cell network operating in an indoor environment by optimizing the path difference, operating frequency and path loss model. The remaining sections of the paper are organized as follows: Section II presents the system model, and the proposed genetic algorithm is discussed. Section III presents the numerical and simulation results of path loss and received signal power using the MOGA scheme and a traditional scheme. Finally, in Section IV, the conclusions are made.
2 Methods
2.1 Indoor Path Loss Models
Nowadays providing signal coverage in indoor office and shopping malls is a major issue because of many obstracles like walls, ceiling, curtains, name boards, furnitur, etc. Therefore, we considered path loss estimation in indoor office or shopping malls in this paper. The indoor path loss models are essential in assessing the overall system’s performance since they are used to precisely build and compare radio systems. Recently, a lot of groups have been working on understanding the propagation mechanisms at frequencies beyond 6 GHz and creating path loss models that can anticipate channel impairments in a consistent and accurate manner. The five organizations created the most relevant path loss models to estimate overall propagation loss in indoor open and closed offices, as well as shopping malls [47]. They are 3GPP, 5GCM, mmMAGIC, METIS, and IEEE [15]. The summary of the indoor LOS path loss models is given in Table 2. It shows the path model, formula for the estimation of path loss, shadow fading parameter, path difference and operating frequencies. The path loss depends mainly on the path difference, operating frequency, antenna location and type of the antenna. The antenna location gives LOS and NLOS path between both transmitting antenna and receiving antennas and it gives shadowing factor which is listed in Table 2. In the indoor environment, we accounted for a maximum path difference of 150 m to accommodate the diverse obstacles within the indoor area, along with an operating frequency of 100 GHz.
In Table. 1, \(d_{3D}\) is the distance between BS and UE in meters, which is derived as \(d_{3D} = \sqrt{(d_{2D} )^2+(h_{BS}-h_{UE})^2}\) [15]. In this equation \(d_{2D}\) is the actual distance between the BS and UE, \(h_{BS}\) and \(h_{UE}\) are the BS and UE antenna heights in meters respectively and f is the operating frequency in GHz
The received signal power at the user equipment based on the path loss is given as [48]:
where \(P_r\) is the received signal power in dBm, \(P_t\) is the transmitted signal power in dBm, \(G_{BS}\) and \(G_{UE}\) are the gain of base station antenna and user equipment in dB respectively, \(A_t\) is the antenna loss and PL is the path loss estimated in the previous section.
2.2 Genetic Algorithm
A natural selection procedure serves as the basis for the optimization method known as a genetic algorithm. It uses the "survival of the fittest" concept and is a population-based search method [49]. The genetic operators are applied repeatedly to existing individual populations, and they create new populations to enhance the system’s performance. The main components of GA are the representation of the chromosomes, selection, recombination, mutation, and evaluation of fitness functions, which are shown in the GA flowchart in Fig. 1. Simple GA has been transformed into multi-objective GA. Assigning fitness functions is where MOGA and GA diverge. Multi-objective GA’s main objective is to create the best Pareto front possible in the objective space, which ensures that no fitness function can be further enhanced without having an impact on the other fitness functions [50, 51].
where \(fitness_{PathLoss}\): represents the path loss estimated using indoor path loss equations listed in Table 1. \(fitness_{ReceivedPower}\): represents the power estimated using the Friss transmission equation shown in Eq. (1). The optimization parameters to maximize the path loss and recceived power are path difference, oprtaing frequency, and path loss model.
Calculates the fitness value for each individual within the current population and then converts the raw fitness scores into a range of usable values.
Individuals with lower fitness values (path loss) and higher fitness values (receiver power) in the current population are transferred to the next population. These values give rise to optimal solution.
The procedure is repeated for many generations and the best solution is identified as the individual with the best fitness value.
Finally, the objective function is optimized by finding the best fitness value of individuals within the population.
Therefore, the genetic algorithm efficiently searches for and converges to a global minimum and maximum without the requirement of complex derivative computations.
The optimization process utilized a GA package in MATLAB to achieve its goals. The parameters considered for optimization are the path difference, ranging from 1 to 150 m, and the operating frequency, ranging from 6 to 100 GHz. Therefore, we considered path difference and operating frequency as decision variables in MOGA. The objective functions are determined as the minimization of path loss and maximization of received power using this MOGA.
The GA tuning is used to optimize the indoor path loss models and to achieve the desired path loss. In this paper, we considered MOGA for the optimization of path loss and receiver power because GA optimizes only one objective function whereas MOGA optimizes multiple objective functions. The two parameters which plays an improtan role in signal transmission and coverage enhancement are path loss and receiver power, which can be optimised using MOGA by adjusting the basic parameters of optimization algorithem i.e., population size, mutation rate and selection rate. We conducted a series of simulations to fine-tune the MOGA parameters with various combinations of population sizes, mutation rates, and selection rates. Based on the outcomes of these preliminary experiments, we observed that a population size of 12, a mutation rate of 2, and a selection rate of 1 yielded satisfactory results such as enhanced path loss and received power. The optimization process is carried out for a maximum of 50 iterations. In this paper, we compared the traditional MATLAB coding approach with the genetic algorithm approach for estimating path loss and received power in indoor environments.
3 Results and Discussions
In this paper, we considered indoor propagation models with a directional antenna to estimate path loss and received signal power within office and shopping mall environments. To perform these estimations, we referred to Table 2 and Eq. (1). Our analysis encompassed a range of path differences, from 1 to 150 ms, and operating frequencies spanning from 6 to 100 GHz.
To obtain simulation results, we utilized the Genetic Algorithm as an optimization tool and traditional MATLAB programming for analytical measurements. In the traditional method, path loss and received power are estimated using path loss equations in Table 2 by substituting various frequencies and path differences. From the path loss models shown in Table 2, it is observed that the path loss depends on operating frequency and path difference. There is a linear relationship between path loss and path difference and operating frequency i.e., the path loss increases with the increase of path difference, because the increase in path difference may cause NLOS path. It also increases with the increase in the operating frequency. Similarly, the receiver signal power reduces as the path loss increases. In GA approach, GA is configured with specific tuning parameters, such as population size, mutation rate, selection rate, and number of iterations. These parameter values were carefully chosen to facilitate an effective optimization process and ensure accurate evaluation of system performance.
The traditional and GA based results pertaining to path loss and receiver power for all six indoor propagation models are shown in Figs. 2, 3, 4, 5, 6, 7. These figures provide a comprehensive overview of the outcomes achieved through our simulation analysis. Figures 2 and 5 depict the path loss and received signal power obtained from the analytical approach using the mentioned indoor propagation models. The results reveal a clear trend where path loss increases and receiver power decreases as the distance between the transmitter and receiver increases. It is important to note that path loss and receiver power are influenced by various factors such as the path difference, presence of obstacles, antenna location, tilt angle, and polarization.
The analytical values of path loss and receiver power are listed in Tables 3 and 4. From Tables 3 and 4, it is observed that the mmMAGIC InH models provide better path loss and receiver power compared to other indor office models such as 5GCM, 3GPP, and IEEE 802.11 ad models. In the indoor shopping mall, METIS provides reduced path loss and enhanced receiver power compared to the 5GCM model. Therefore, mmMAGIC InH office model and METIS InH shopping mall models are considered as optimal path loss models in indoor office and shopping mall respectively in the traditional approach.
The minimum and maximum values of path loss and receiver power exhibit variations that are dependent on the specific propagation model employed, the path difference between transmitter and receiver, as well as the operating frequency. These factors contribute to the diversity of path loss and receiver power observed in different scenarios. Understanding these variations is crucial for accurate performance evaluation and optimizing the design of wireless communication systems. The path loss and receiver power can be enhanced by using an optimization approach. Hence, the optimization using MOGA is preferred in this paper. We considered the population size of 12, mutation rate of 2, selection rate of 1, and total number of iterations as 50 to achieve better path loss and receiver power. Indoor office path loss and received power is estimated using 5GCM, 3GPP, mmMAGIC and IEEE models and indoor shopping mall using 5GCM and METIS models. Figs 3, 4, 6, and 7 provide a comparative analysis of the path loss and receiver power in both indoor office and shopping mall environments using both the Multi-Objective Genetic Algorithm (MOGA) method and the traditional method. Notably, these figures demonstrate that the GA method consistently leads to a reduction of over 28 dB in the minimum path loss and an enhancement of 38 dBm in the maximum receiver power, compared to the traditional approach. The receiver power enhancement of 38 dBm provides 30% expansion in coverage area.
Fig. 4
Comparison of path loss of InH-shopping mall using MOGA and traditional method
To provide a comprehensive overview, Table 5 presents the path loss, receiver power, and the difference in path loss and difference in received power between the GA and traditional methods for the indoor models. It is worth noting that in all propagation models, except for the mmMAGIC InH-office model, the GA method effectively minimizes path loss and enhances received power. These results highlight the effectiveness and superiority of the optimization approach, specifically the GA method, in improving path loss and receiver power in various indoor scenarios, thereby demonstrating its potential for optimizing the performance of wireless communication systems.
Table 6 presents the minimum, maximum, and mean path loss values obtained from the indoor propagation models using the analytical approach. This table provides a comprehensive summary of the range and average path loss values, allowing for a clear comparison and assessment of the performance of the different propagation models. The 5GCM indoor officce, 5GCM InH shopping mall, 3GPP TR 38.91 InH office, mmMAGIC InH office, METIS InH shopping mall, and IEEE 802.11 ad InH office indoor path loss models estimates the path loss of 62.37 dB, 62.15 dB, 63.12 dB, 50 dB, 55.18 dB, and 52.89 dB in traditional approach and 36.87 dB, 35.86 dB, 36.84 dB, 68.80 dB, 36.23 dB and 33.94 dB using GA approach and received powers of \(-12.17~dBm, -11.37~dBm, -12.17~dBm, -5.80~dBm,\)\(-12.24~dBm\) and \(-8.68~dBm\) in traditional approach and 26.13 dBm, 27.14 dBm, \(26.15~dBm, -5.80~dBm, 26.75~dBm\) and 29.05 dBm using GA approach repectively. The 5GCM and 3GPP models produce the path loss difference above 25 dB and mmMAGIC, METIS and IEEE models produce a path loss below 19 dB. Except mmMAGIC model, all models produce the receiver power difference above 37 dBm. Therefore, the highest path loss difference of 26 dB is observed in the 5GCM InH shopping mall model and the highest recieved power difference of 39.01 dBm is observed in METIS InH shopping mall model.
Table 5
Comparison of path loss and receiver power
Model
Path loss (dB)
Receiver power (dBm)
Trad
MOGA
PL diff.
Trad
MOGA
RP diff.
5GCM Indoor office
62.37
36.87
25.50
\(-\)12.17
26.13
38.30
5GCM InH shopping mall
62.15
35.86
26.29
\(-\)11.37
27.14
38.51
3GPP TR 38.91 InH-office
63.12
36.84
26.28
\(-\)12.17
26.15
38.32
mmMAGIC InH-office
50
68.80
\(-\)18.80
\(-\)5.80
\(-\)5.80
0
METIS InH-shopping mall
55.18
36.23
18.95
\(-\)12.24
26.75
39.01
IEEE 8022.11 ad InH office
52.89
33.94
18.95
\(-\)8.68
29.05
37.37
To determine the optimal path loss model and accuracy of the considered models, we conducted a comparison between the path loss values obtained through the genetic algorithm and the traditionally measured path loss values. This analysis allowed us to calculate error statistics such as mean squared error (MSE), root mean squared error (RMSE), and the standard deviation of the error.
In Figs. 2, 3, 4, 5, 6, 7, we examined the minimum, maximum, and mean path loss and received power for each indoor scenario, enabling us to identify the model that exhibited the lowest path loss and highest received power i.e., an optimal path loss model. All the error statistics are summarized in Table 6. From our observations in Table 6, we found that in the indoor office environment, the METIS model displayed lowest MSE and RMSE values compared to the other models. Conversely, in the indoor shopping mall environment, the mmMAGIC model demonstrated lower MSE and RMSE values when compared to the 5GCM model. Consequently, we designate the METIS model as the optimized path loss model for the indoor office environment, while considering the mmMAGIC model as the optimized path loss model for the indoor shopping mall environment. These findings highlight the superior performance and accuracy of these respective models in their respective indoor scenarios. Thus, compared to the traditional approach, the utilization of intelligent optimization algorithms can improve the path loss and received power in indoor environments.
Table 6
Path loss and received power error statistics
Path loss error statistics
PL (dB)
5GCM InH-office
5GCM InH-SM
3GPP InH-office
METIS InH-SM
mmMAGIC InH-office
IEEE InH-office
Trad
GA
Trad
GA
Trad
GA
Trad
GA
Trad
GA
Trad
GA
Min
74.6
36.87
73.6
35.85
74.58
36.85
68.8
68.8
74.55
36.25
71.68
33.95
Max
112.15
111.95
111.14
110.94
112.13
111.93
108.73
108.75
104.50
104.33
115.09
114.37
Mean
104.62
95.1
103.62
94.1
104.61
95.1
100.73
98.7
98.5
89.24
106.39
96.57
MSE
4.517
4.517
4.517
1.892
3.9485
5.1124
RMSE
2.1253
2.1253
2.1253
1.3757
1.987
2.261
SE
1.9625
1.9625
1.9625
1.4090
1.7737
2.1293
Received power error statistics
RP (dBm)
5GCM InH-office
5GCM InH-SM
3GPP InH-office
METIS InH-SM
mmMAGIC InH-office
IEEE InH-office
Trad
GA
Trad
GA
Trad
GA
Trad
GA
Trad
GA
Trad
GA
Min
\(-\)49.14
\(-\)48.95
\(-\)48.13
\(-\)47.94
\(-\)49.12
\(-\)48.93
\(-\)45.73
\(-\)45.75
\(-\)41.49
\(-\)41.33
\(-\)52.08
\(-\)51.87
Max
\(-\)11.60
26.13
\(-\)10.59
27.14
\(-\)11.58
26.15
\(-\)5.8
\(-\)5.8
\(-\)11.54
26.75
\(-\)8.68
29.05
Mean
\(-\)41.62
\(-\)32.09
\(-\)40.61
\(-\)31.08
\(-\)41.60
\(-\)32.07
\(-\)37.73
\(-\)35.7
\(-\)35.49
\(-\)26.24
\(-\)43.39
\(-\)33.57
MSE
4.517
4.517
4.517
1.892
3.9485
5.1124
RMSE
2.1253
2.1253
2.1253
1.3757
1.987
2.261
SE
1.9625
1.9625
1.9625
1.4090
1.7737
2.1293
In this paper, we utilized the MOGA approach to optimize path loss and received power in an indoor environment. We estimated the error statistics based on the obtained results and observed that the minimum, maximum, and mean values of these statistics depend on the path loss model employed. By employing the MOGA approach, we were able to estimate the optimized path loss and received power. We compared the optimized values of path loss and received power across six different path loss models and identified the model that yielded the lowest path loss and highest received power. This optimized path loss model holds the potential to enhance the quality of service in indoor wireless networks and can be implemented by network providers.
This study specifically focused on LOS propagation models for indoor scenarios. However, our future work aims to expand this investigation to NLOS propagation models and estimate the path loss and received power in diverse indoor environments.
4 Conclusion
In this paper, indoor path loss models were considered to estimate the path loss and received signal power in indoor office and shopping mall scenarios for a path difference of 1–150 m and an operating frequency of 6–100 GHz. An optimization algorithm, MOGA is proposed to enhance the path loss and received signal power using the MATLAB GA tool. The simulation results using the GA tool and the traditional approach are compared. From the results obtained, it is apparent that the optimized MOGA predicts better path loss and receiver power compared to the traditional approach. The path loss is reduced by more than 26 dB, and the receiver power is enhanced by more than 38 dBm in all five indoor propagation models using the proposed optimization algorithm compared to the traditional approach. This will provide better coverage i.e., approximately 30% more in the indoor region at mmwave frequencies. METIS and mmMAGIC models are considered as optimized path loss models in the indoor office and shopping mall environments respectively.
Therefore, the genetic algorithm efficiently searches for and converges to a global minimum without the requirement of complex derivative computations.
Declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Data Availability
Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
