Exploring the Potential of Wi-Fi in Industrial Environments: A Comparative Performance Analysis of IEEE 802.11 Standards

Abstract

The advent of Industry 4.0 brought about digitalisation and the integration of advanced technologies into industrial processes, with wireless networks emerging as a key enabler in the interconnection of smart devices, cyber–physical systems, and data analytics platforms. With the development of Industry 5.0 and its emphasis on human–machine collaboration, Wi-Fi has positioned itself as a viable alternative for industrial wireless connectivity, supporting seamless communication between robots, automation systems, and human operators. However, its adoption in critical applications remains limited due to persistent concerns over latency, reliability, and interference in shared-spectrum environments. This study evaluates the practical performance of Wi-Fi standards from 802.11n (Wi-Fi 4) to 802.11be (Wi-Fi 7) across three representative environments: residential, laboratory, and industrial. Six configurations were tested under consistent conditions, covering various frequency bands, channel widths, and traffic types. Results prove that Wi-Fi 6/6E delivers the best overall performance, particularly in low-interference 6 GHz scenarios. Wi-Fi 5 performs well in medium-range settings but is more sensitive to congestion, while Wi-Fi 4 consistently underperforms. Early Wi-Fi 7 hardware does not yet surpass Wi-Fi 6/6E consistently, reflecting its ongoing development. Despite these variations, the progression observed across generations clearly demonstrates incremental gains in throughput stability and latency control. While these improvements already provide tangible benefits for many industrial communication scenarios, the most significant leap in industrial applicability is expected to come from the effective implementation of high-efficiency mechanisms. These include OFDMA, TWT, scheduled uplink access, and enhanced QoS features. These capabilities, already embedded in the Wi-Fi 6 and 7 standards, represent the necessary foundation to move beyond conventional best-effort connectivity and toward supporting critical, latency-sensitive industrial applications.

1. Introduction

The advancement of digitalisation and automation in the industrial and manufacturing environments has led to the need for innovative communication systems capable of supporting applications with ultra-low latency requirements and high data rates. From the perspective of Industry 5.0 [1], the variety of emerging wireless technologies that enable converged networks represents an evolution in collaboration and coexistence between humans and machines in flexible and shared work environments. The advent of Industry 4.0 has radically transformed the automation process in the industrial sector by decreasing human intervention in manufacturing processes, thus increasing productivity and reducing workplace accidents [1,2]. However, the implementation of convergent networks in Industry 5.0 underscores the importance of shared human–machine work environments to highlight unique human skills such as creativity and innovation, problem-solving, and decision-making as key elements for success in manufacturing. In this way, the goal is to reintegrate the human factor into manufacturing production, promoting human–machine collaboration by prioritising customisation, flexibility, and innovation in manufacturing.

The emergence of these collaborative industrial environments has underscored the need for wireless technologies that can adhere to the stringent specifications of emerging industrial applications, including collaborative robotics, predictive maintenance, real-time monitoring, and teleoperation. Thanks to the increased performance of the latest Wi-Fi standards and their reduced installation costs, next-generation wireless technologies are emerging as leaders in manufacturing environments over traditional wired technologies. Currently, wireless alternatives such as the latest variants of the IEEE 802.11 standard [3], commercially known as Wi-Fi 6 [4,5] and Wi-Fi 7 [6], offer features similar to 5G cellular technology [7,8] in its different network architecture modalities. These wireless technologies are ideal candidates for the Industry 4.0 revolution, which is transitioning to the Industry 5.0 paradigm [9]. In this new era, the industrial and manufacturing sectors are transformed through the integration of the human dimension and the adaptation of automation.

Due to this ecosystem, where mobility and agility in a collaborative human–machine environment are fundamental characteristics of the industry of the future, the aim is to create convergent networks [10,11] that allow the coexistence of humans and machines to enhance human capabilities and skills in favour of innovation and the customisation of production. This is why convergent networks play a key role in the advance towards Industry 5.0. Another benefit of this type of network is the optimisation of radio resources, reducing the complexity of network management by integrating heterogeneous devices and systems into a unified network. This interconnection of emerging technologies under the umbrella of the same technology favours synergy between them, facilitating the creation of innovative industrial solutions and generating added value in manufacturing processes.

However, most experimental Wi-Fi studies still focus on a narrow set of deployment scenarios and interference patterns. Investigations such as [12,13] are conducted in tightly controlled laboratory environments, which limit the external validity of their conclusions. Regarding spectrum, the Wi-Fi 6E test at 6 GHz was confined to a single urban campus and mainly focused on downlink traffic [14], while extended-spectrum tests relied on prototype hardware and were only compared with Wi-Fi 6 [15]. Research on Wi-Fi 7 is even more preliminary, usually limited to laboratory experiments or simulations, and lacks direct empirical comparisons with previous generations [16]. A representative industrial study [17] benchmarks several 802.11 generations at different line-of-sight distances but does not consider the richer interference patterns present in real factory settings. These gaps underscore the need for broader, multi-scenario measurement campaigns that evaluate modern Wi-Fi technologies across varied bands, channel widths, and interference conditions.

To overcome these gaps and enhance the diversity of scenarios tested for Wi-Fi technologies, this article examines the performance of the latest versions of the IEEE 802.11 standard in different relevant environments, showcasing the technology’s evolutionary advancements and adaptability within the context of Industry 5.0. To evaluate the capabilities of these Wi-Fi technologies, typical network parameters such as Round-Trip Time (RTT), peak throughput (THR), packet loss (PL), and jitter are measured. Furthermore, the evaluation of these parameters is extended to two types of traffic with different characteristics.

The paper is organised as follows. Section 2 outlines the technical advances and theoretical performance of recent IEEE 802.11 standards. Section 3 and Section 4 describe the experimental setup, the evaluated configurations, and the three deployment environments. Section 5 and Section 6 present and analyse the results following the same structure. Finally, Section 7 concludes the paper and outlines directions for future work.

2. Related Work

The constant evolution of the IEEE 802.11 standard has transformed the way we access the Internet since its inception, enabling mobility and eliminating reliance on traditional wired connections. The exponential progression of the latest Wi-Fi standards reflects improvements in terms of performance, capacity, and efficiency, addressing the technological challenges that arise with the increasing number of ultra-low-latency industrial services with stringent communication requirements. Due to recent improvements, Wi-Fi networks have emerged as a promising candidate to digitise and automate the industrial sector [5,17]. Therefore, a detailed review of Wi-Fi standards is presented, with a focus on key advancements in IEEE 802.11n (Wi-Fi 4), IEEE 802.11ac (Wi-Fi 5), IEEE 802.11ax (Wi-Fi 6/6E) [4,13,18], and the future standard yet to be finalised, IEEE 802.11be (Wi-Fi 7) [5,19].

In order to contextualise these developments, it is essential to understand the physical-layer fundamentals that enable such performance improvements, especially the role of channel bandwidth and its relationship with spectrum allocation. All modern Wi-Fi standards are fundamentally shaped by how they utilise and aggregate frequency resources across the 2.4 GHz, 5 GHz, and 6 GHz unlicensed bands. Through the concept of channel bonding, IEEE 802.11 systems have moved from narrow 20 MHz channels to broader bandwidths of up to 320 MHz, allowing for increases in peak throughput. This expansion is not arbitrary but rather follows precise spectral structures regulated and engineered to ensure coexistence, efficiency, and backward compatibility.

In IEEE 802.11 wireless communication systems, the channel bandwidth is one of the most critical physical-layer parameters that define system throughput and spectral efficiency. With the evolution of user demand and latency-sensitive applications, modern Wi-Fi systems have adopted wider spectral allocations through channel bonding techniques, leading to significant gains in peak data rates [12,20]. Wi-Fi primarily operates in three unlicensed frequency bands: 2.4 GHz, 5 GHz, and 6 GHz. Within these bands, the fundamental spectral unit is the 20 MHz channel, shaped by regulatory guidelines and radio hardware constraints. A 20 MHz channel was originally divided into 64 OFDM subcarriers, spaced at 312.5 kHz [3]. To improve capacity, adjacent 20 MHz channels can be aggregated, forming wider bandwidths according to Equation (1), where B represents the total channel bandwidth in MHz, and N is the number of 20 MHz channels being bonded [21].

$$B=20·N\left(\mathrm{MHz}\right),N\in \{1,2,4,8,16\}$$

(1)

This hierarchical aggregation of adjacent 20 MHz channels results in wider operational bandwidths: 40 MHz (2 × 20 MHz), 80 MHz (4 × 20 MHz), 160 MHz (8 × 20 MHz, either as a contiguous block or split as 80 + 80 MHz), and up to 320 MHz (16 × 20 MHz), the latter introduced with Wi-Fi 7 in the 6 GHz band. This channel bonding approach enables significant increases in data transmission capacity; however, it also requires larger contiguous portions of the spectrum and careful coordination with legacy systems to avoid interference and ensure interoperability [12].

Figure 1 provides a visual reference for how adjacent 20 MHz channels are combined into broader bandwidths to meet increasing data rate demands, illustrating the core principle of channel bonding in modern Wi-Fi systems. The figure covers the 2.4 GHz, 5 GHz and 6 GHz bands, showing the structured allocation of 20, 40, 80, 160, and 320 MHz channels.

In the 2.4 GHz band, only three non-overlapping 20 MHz channels are available, which limits the aggregation potential. The 5 GHz band comprises several segments of unlicensed spectrum defined by regulatory entities, ETSI in Europe, and FCC in the United States, with each region imposing specific constraints on transmission power and indoor/and outdoor use. Depending on these constraints, up to 25 channels of 20 MHz are available in the US and up to 19 in Europe, enabling configurations such as six non-overlapping 80 MHz channels or two 160 MHz channels. The 6 GHz band, newly allocated to Wi-Fi under Wi-Fi 6E and further expanded in Wi-Fi 7, offers a significantly broader spectral window, supporting up to 59 channels of 20 MHz and up to three non-overlapping 320 MHz channels. The figure also highlights regional differences in spectrum allocation, marked with European and US flags and corresponding frequency boundaries. This expanded availability of unlicensed spectrum underpins the deployment of high-throughput, low-latency wireless links for applications in industrial automation, multimedia streaming, and mobile connectivity.

From a theoretical point of view, the capacity gains associated with wider channels are governed by the Shannon–Hartley theorem, which defines the upper bound of the achievable data rate as a function of channel bandwidth B and signal-to-noise ratio ($SNR$). The channel capacity C, in bits per second, is given by Equation (2).

$$C=B·{log}_2(1+\mathrm{SNR})\left[\mathrm{bps}\right]$$

(2)

This relationship shows that, under constant SNR, increasing the bandwidth directly increases the theoretical capacity. However, it also results in a proportional rise in noise power, which can degrade SNR if the received signal power is not adjusted accordingly [23]. This highlights a key trade-off between throughput and link robustness in practical wireless systems. This relationship is reflected by Equation (3).

$${\mathrm{SNR}}_{\mathrm{dB}}=P_{\mathrm{rx}}-N_0-10{log}_{10}\left(B\right)$$

(3)

where $P_{\mathrm{rx}}$ is the received power in dBm, $N_0$ is the noise power spectral density, and B is the bandwidth in hertz. As bandwidth increases, the term $10{log}_{10}\left(B\right)$ grows, effectively reducing the SNR unless the transmission power or antenna gain is increased accordingly. Moreover, the received power $P_{\mathrm{rx}}$ itself decreases with distance, which further impacts the SNR in real deployments. This effect is commonly described using the log-distance path loss model in Equation (4).

$$\mathrm{PL}\left(d\right)=\mathrm{PL}\left(d_0\right)+10n{log}_{10}\left({\displaystyle \frac{d}{d_0}}\right)+X_{\sigma }$$

(4)

In Equation (4), $\mathrm{PL}\left(d\right)$ represents the path loss at a given distance d, $\mathrm{PL}\left(d_0\right)$ is the path loss at a reference distance $d_0$, n is the path loss exponent that depends on the environment (e.g., indoor, urban, and open space), and $X_{\sigma }$ accounts for random shadowing effects. As the transmitter–receiver distance increases, the path loss increases logarithmically, leading to a lower received power and therefore to a lower SNR. These effects illustrate some key challenges in Wi-Fi systems. Although increasing bandwidth improves capacity, it also makes the link more sensitive to distance, obstacles, and interference. Maintaining reliable high-throughput connections over wider channels therefore requires the careful tuning of transmission parameters, the use of robust modulation schemes, and, in some cases, additional infrastructure support [24].

2.1. IEEE 802.11n (Wi-Fi 4)

Wi-Fi 4 marked a substantial improvement over its predecessors, increasing the maximum network performance up to 600 Mbps and allowing the use of both the 2.4 GHz and 5 GHz frequency bands. This standard introduced several innovative technologies that transformed the user experience in terms of performance, reliability, and coverage range. Networks based on this technology support bandwidths of up to 40 MHz, doubling the available bandwidth compared to previous standards and enabling higher network performance, resulting in faster connection speeds and better performance for high-demand services. One of the key advances in Wi-Fi 4 is the use of Orthogonal Frequency Division Multiplexing (OFDM), which improves spectral efficiency and enhances data transmission by splitting the signal into multiple subcarriers. This method reduces interference and improves signal robustness, especially in environments with a high density of devices.

This standard introduced modulation techniques, such as 64-QAM (Quadrature Amplitude Modulation), which improve spectral efficiency. It also incorporated frame aggregation, a technique that combines multiple data frames into a single transmission, reducing network overhead and latency while increasing network performance. Additionally, the Multiple-Input–Multiple-Output (MIMO) technology was added, which uses multiple antennas at both the transmitter and receiver to send and receive multiple data streams simultaneously. This improvement enhances network capacity and performance, providing a more robust connection. MIMO technology reduces interference, allowing the transmission of stronger and cleaner signals.

2.2. IEEE 802.11ac (Wi-Fi 5)

This standard has been a key factor in the adoption of high-speed wireless networks, providing the foundation for performance-demanding industrial applications such as the remote telecontrol of machinery. Wi-Fi 5 operates exclusively in the 5 GHz band, eliminating the congestion commonly experienced in the 2.4 GHz band and taking advantage of channel capacities of up to 160 MHz, four times wider than its predecessor.

The introduction of 256-QAM modulation improved spectral efficiency by encoding more bits per symbol, and the use of new coding schemes enhanced the error resistance and reliability of the wireless network. Additionally, the incorporation of Multiple-User–Multiple-Input–Multiple-Output (MU-MIMO) technology allows multiple devices to communicate with an access point simultaneously rather than taking turns, improving network capacity and reducing latency in high-density device environments. Another technology incorporated is the beamforming technique, which directs Wi-Fi signals towards specific devices instead of scattering them in all directions, resulting in more robust and stable connections.

2.3. IEEE 802.11ax (Wi-Fi 6 / Wi-Fi 6E)

The IEEE 802.11ax standard represented a significant advance over its predecessors, offering improvements in both network performance and efficiency. These capabilities were later extended to the 6 GHz band, commercially known as Wi-Fi 6E, providing even more of the spectrum and reducing congestion.

The incorporation of Orthogonal Frequency Division Multiple Access (OFDMA) technology allows a channel to be split into smaller sub-channels, each assigned to different devices, improving spectral efficiency and reducing latency in dense environments. This feature is particularly relevant for real-time industrial services and applications requiring ultra-low latency. Additionally, MU-MIMO is enhanced, allowing simultaneous communications with up to eight devices for both uplinks and downlinks. Another technique added to improve spectral efficiency is the use of Basic Service Set (BSS) colouring. This technique assigns different colours to different BSSs, enabling neighbouring networks operating on the same channels to reduce their interference. This standard also introduced new mechanisms such as Target Wake Time (TWT) to improve the energy efficiency of devices by allowing them to negotiate when and how often they wake up to send or receive data.

With the advent of Wi-Fi 6E and the expansion to the 6 GHz spectrum, congestion in the 2.4 GHz and 5 GHz bands is reduced, providing more non-overlapping channels and improving overall Wi-Fi network performance and capabilities. By expanding the available spectrum, Wi-Fi 6E supports more devices connected simultaneously without performance degradation, ideal for industrial environments demanding high device densities and performance. By incorporating these techniques and technologies, Wi-Fi 6E paves the way for future innovations in wireless networking.

2.4. IEEE 802.11be (Wi-Fi 7)

IEEE 802.11be, also known as Wi-Fi 7, is the next generation of Wi-Fi technology designed to meet the growing demands for high-capacity and ultra-low-latency wireless networks. Although the latest version of the standard is yet to be released, it is scheduled to be introduced in 2024, offering substantial improvements in terms of performance, efficiency, and network capacity.

One of the most significant enhancements is the expansion of the channel bandwidth. Wi-Fi 7 supports channels up to 320 MHz, effectively doubling the maximum channel width available in Wi-Fi 6/6E (160 MHz). This wider channel allows for a substantial increase in data throughput, leading to significantly higher connection speeds and improved performance for bandwidth-intensive applications. Furthermore, Wi-Fi 7 introduces 4096-QAM, also known as 4K-QAM. This modulation scheme significantly enhances spectral efficiency by encoding four times more bits per symbol compared to the 1024-QAM used in Wi-Fi 6 [25].

It supports new technologies such as Multi-Link Operation (MLO), Restricted Target Wake Time (r-TWT), and Multi-Resource Unit (MRU). MLO allows devices to simultaneously utilise multiple frequency bands and channels (e.g., 2.4 GHz, 5 GHz, and 6 GHz). This simultaneous transmission and reception across multiple links distinguishes Wi-Fi 7 from previous standards. By aggregating these links, MLO enables devices to achieve several key benefits: higher throughput by combining the bandwidth of multiple channels, lower latency by selecting the optimal link for each data packet and improved reliability by providing redundancy in case one link experiences interference or congestion [26]. A particularly important enhancement for latency-sensitive applications is r-TWT. Building on the Target Wake Time (TWT) mechanism introduced in Wi-Fi 6, r-TWT further reduces the time devices spend in a low-power sleep state between data transmissions. By enabling more frequent and shorter wake-up intervals, r-TWT significantly reduces latency and improves responsiveness, making Wi-Fi 7 more suitable for real-time applications such as online gaming, video conferencing, and industrial control systems.

This is achieved through more flexible scheduling and negotiation between the access point and the client devices [27]. Furthermore, MRU enables a more flexible and efficient mechanism for allocating radio resources. Unlike previous standards that relied on fixed resource unit sizes, MRU allows for more granular and dynamic allocation of resources to different users and applications. This improves spectral efficiency by minimising wasted resources and optimising data transmission based on the specific needs of each device, especially in scenarios with varying traffic demands [28].

Wi-Fi 7 is designed to support the demands of the next generation of services and industrial applications with higher data rates, lower latencies, and higher spectral efficiency that transform the user experience by significantly increasing the number of simultaneously connected devices without substantial performance degradation. Wi-Fi 7 is a strong candidate to enable the connectivity demands of Industry 5.0, facilitating seamless communication between humans, machines, and intelligent systems in increasingly complex and interconnected environments [29]. To highlight the evolution of Wi-Fi technology, the

Table 1. Summary of technical capabilities of the latest Wi-Fi standards.

Feature	Wi-Fi 4	Wi-Fi 5	Wi-Fi 6	Wi-Fi 6E	Wi-Fi 7
IEEE Standard	802.11n	802.11ac	802.11ax	802.11ax	802.11be
Theoretical Maximum Throughput	600 Mbps	3.5 Gbps	9.6 Gbps	9.6 Gbps	46 Gbps
Frequency Bands	2.4 GHz, 5 GHz	5 GHz	2.4 GHz, 5 GHz	2.4 GHz, 5 GHz, 6 GHz	2.4 GHz, 5 GHz, 6 GHz
Bandwidth	20 MHz, 40 MHz	20 MHz, 40 MHz, 80 MHz, 80 + 80 MHz, 160 MHz	20 MHz, 40 MHz, 80 MHz, 80 + 80 MHz, 160 MHz	20 MHz, 40 MHz, 80 MHz, 80 + 80 MHz, 160 MHz	20 MHz, 40 MHz, 80 MHz, 80 + 80 MHz, 160 MHz, 320 MHz
Frequency Multiplexing	OFDM	OFDM	OFDM and OFDMA	OFDM and OFDMA	OFDM and OFDMA
Maximum Modulation	64-QAM	256-QAM	1024-QAM	1024-QAM	4096-QAM
MU-MIMO	N/A	8 × 8 DL	8 × 8 UL/DL	8 × 8 UL/DL	16 × 16 UL/DL
Other advanced features	N/A	N/A	BSS Coloring, RU, TWT	BSS Coloring, RU, TWT	MLO, MRU, r-TWT

summarises the key technical aspects and advancements across the different standards.

3. Performance Evaluation

To evaluate the performance of the IEEE 802.11 standard, a series of tests were conducted in three relevant environments with different physical configurations. This section provides details on the Wi-Fi network architecture used, the various configurations assessed, the methods applied to evaluate network capabilities, and an explanation of the three environments selected, based on the network architecture depicted in Figure 2. Each of these tests incorporated three traffic patterns designed to emulate realistic data transmission behaviours commonly observed in industrial and urban deployments.

The first traffic pattern is based on the Internet Control Message Protocol (ICMP), which is widely implemented for network diagnostic and monitoring purposes, such as those executed via the ping utility. This pattern involves the transmission of Echo Request messages and the reception of corresponding Echo Reply messages from the destination node. While ICMP imposes minimal load on the network, it is instrumental in capturing key performance indicators, including round-trip latency, packet loss rates, and link stability, critical parameters in industrial use cases.

In parallel, Transmission Control Protocol (TCP) traffic was generated using iperf3, enabling the emulation of stateful, connection-oriented communication typical of services such as web browsing, cloud storage, and multimedia streaming. Owing to its inherent congestion control and error-correction features, TCP traffic is particularly suited for stress-testing network capacity and throughput stability.

Furthermore, traffic based on the User Datagram Protocol (UDP) was also generated via iperf3. Unlike TCP, UDP supports a connectionless paradigm, which makes it ideal for simulating delay-sensitive applications where timely delivery is prioritised over reliability. Such applications include Voice over IP (VoIP), online conferencing, and real-time video, core components of many industrial IoT systems. Since UDP does not implement retransmission mechanisms, it provides valuable insight into network behaviour under conditions of constrained bandwidth and variable interference, particularly in assessing packet loss and jitter performance.

3.1. Network Architecture

The experimental network deployment used to measure the capabilities and performance of the different IEEE 802.11 standards is described in detail in Figure 2. The components of the Wi-Fi network are as follows:

• Netgear AXE7800 is a tri-band access point that supports Wi-Fi 6 and 6E (IEEE 802.11ax) and operates in the 2.4 GHz, 5 GHz, and 6 GHz bands. It ensures backward compatibility with older standards (Wi-Fi 4 and 5). Internally, it integrates the Broadcom BCM6715 + BCM6756 chipset combination.
• Asus RT-BE88U is an advanced access point that supports the Wi-Fi 7 standard (IEEE 802.11be) and operates in the 2.4 and 5 GHz bands. Although the 6 GHz band is not available in this model, it enables access to advanced features such as 4K QAM and enhanced multi-link operations. It is powered by the Broadcom BCM4916 network processor along with the BCM6726 Wi-Fi 7 radio chipset.
• The Raspberry Pi 4 Model B is responsible for running the iperf3 server to evaluate the performance of the Wi-Fi network. It is connected to the network via an Ethernet cable to the access point, as shown in Figure 2.
• The desktop computer with an Intel Core i7-10700 processor with a Gigabyte AORUS GC-WIFI7 network card, connected via a PCI-E port. This card supports Wi-Fi 4 to Wi-Fi 7 and operates in three bands (2.4, 5, and 6 GHz). Internally, it uses the Qualcomm QCNCM825 chipset, designed to provide full IEEE 802.11be functionality.

The use of COTS hardware in this study aims to evaluate the performance of Wi-Fi in representative and reproducible scenarios. Although these devices are not designed for extreme industrial environments, they incorporate the same chipsets as many industrial-grade solutions, enabling comparable results at the MAC and physical layers.

3.2. Wi-Fi Configuration Parameters and Methodology Evaluation

The testbed is designed to evaluate different wireless technologies defined by the IEEE 802.11 standard, including both earlier versions and the most recent advancements. Each technology is analysed within its characteristic frequency bands and bandwidths to assess its performance under various operational conditions, as detailed in

Table 2. Summary of technical capabilities of the latest Wi-Fi standards.

IEEE Standard	Wi-Fi Name	Frequency Band	Bandwidth	Theoretical Maximum Bitrate	Acronym
802.11n	Wi-Fi 4	2.4 GHz	20 MHz	144 Mbps	WiFi${}^4$_2.4${}^{20}$
802.11n	Wi-Fi 4		40 MHz	300 Mbps	WiFi${}^4$_2.4${}^{40}$
802.11ax	Wi-Fi 6		20 MHz	287 Mbps	WiFi${}^6$_2.4${}^{20}$
802.11ax	Wi-Fi 6		40 MHz	574 Mbps	WiFi${}^6$_2.4${}^{40}$
802.11be	Wi-Fi 7		20 MHz	344 Mbps	WiFi${}^7$_2.4${}^{20}$
802.11be	Wi-Fi 7		40 MHz	689 Mbps	WiFi${}^7$_2.4${}^{40}$
802.11n	Wi-Fi 4	5 GHz	20 MHz	150 Mbps	WiFi${}^4$_5${}^{20}$
802.11n	Wi-Fi 4		40 MHz	300 Mbps	WiFi${}^4$_5${}^{40}$
802.11ac	Wi-Fi 5		20 MHz	173 Mbps	WiFi${}^5$_5${}^{20}$
802.11ac	Wi-Fi 5		40 MHz	400 Mbps	WiFi${}^5$_5${}^{40}$
802.11ac	Wi-Fi 5		80 MHz	866 Mbps	WiFi${}^5$_5${}^{80}$
802.11ac	Wi-Fi 5		160 MHz	1733 Mbps	WiFi${}^5$_5${}^{160}$
802.11ax	Wi-Fi 6		20 MHz	287 Mbps	WiFi${}^6$_5${}^{20}$
802.11ax	Wi-Fi 6		40 MHz	573 Mbps	WiFi${}^6$_5${}^{40}$
802.11ax	Wi-Fi 6		80 MHz	1201 Mbps	WiFi${}^6$_5${}^{80}$
802.11ax	Wi-Fi 6		160 MHz	2402 Mbps	WiFi${}^6$_5${}^{160}$
802.11be	Wi-Fi 7		20 MHz	344 Mbps	WiFi${}^7$_5${}^{20}$
802.11be	Wi-Fi 7		40 MHz	689 Mbps	WiFi${}^7$_5${}^{40}$
802.11be	Wi-Fi 7		80 MHz	1441 Mbps	WiFi${}^7$_5${}^{80}$
802.11be	Wi-Fi 7		160 MHz	2882 Mbps	WiFi${}^7$_5${}^{160}$
802.11ax	Wi-Fi 6E	6 GHz	20 MHz	287 Mbps	WiFi${}^{6E}$_6${}^{20}$
802.11ax	Wi-Fi 6E		40 MHz	573 Mbps	WiFi${}^{6E}$_6${}^{40}$
802.11ax	Wi-Fi 6E		80 MHz	1201 Mbps	WiFi${}^{6E}$_6${}^{80}$
802.11ax	Wi-Fi 6E		160 MHz	2402 Mbps	WiFi${}^{6E}$_6${}^{160}$

. For a comprehensive evaluation of the technologies detailed above, the testbed is divided into two iterations.

• The first iteration encompasses the study of Round-Trip Time (RTT). The “ping” command was used in three consecutive tests, sending one thousand ICMP packets in each test, with a time between consecutive packets of 5 ms. Results were measured in three environments tested at 15 m.
• The second iteration consists of the evaluation of operational thresholds. In this iteration, the performance for TCP and UDP traffic in the uplink direction was assessed. The iperf3 service was used for this purpose, running in client mode on the desktop computer and in server mode on the Raspberry Pi connected to the AP. Additionally, these operational thresholds were evaluated at three communication distances (CDs) in residential and laboratory environments, while in the industrial environment, because of limitations, the performance was evaluated at only two distances. A summary of this second test is shown in

  Table 3. Configurations parameters used during the second round of experimental evaluation.

  Environment	Communication Distance				Line of Sight (LoS)	Indoor/Outdoor
  	5 m	15 m	45 m	60 m
  Laboratory Area	✓	✓	✓	✓	✓	Indoor
  Residential Area	✓	✓		✓	✓	Outdoor
  Factory		✓	✓			Indoor

  .

4. Test Environments

The selected environments correspond to three typical Wi-Fi network deployment zones, each chosen to represent varying levels of spectrum saturation. These differences arise due to the distinct characteristics of each environment, such as the density of wireless devices, the number of overlapping networks, and the presence of external interference sources. Evaluating these diverse conditions enables a comprehensive analysis of the performance of different IEEE 802.11 technologies under real-world interference and congestion scenarios. To quantify these differences and better understand the spectrum conditions in each environment, a spectrum analysis was conducted.

This analysis consists of capturing and visualising the frequency activity within the selected bandwidth over time. In these captures, the horizontal axis represents the frequency range, while the vertical axis corresponds to time. The resulting spectrograms provide a colour-coded representation of spectrum occupancy, where, in general, blue indicates an unoccupied or low-activity spectrum, while other colours signify increasing levels of signal presence and interference. However, the shades of these heats vary depending on the background noise level of the environments selected for testing.

For example, the spectrogram shown in Figure 3 captures the activity of the 2.4 GHz band, displaying a 100 MHz bandwidth analysis, spanning from 2.4 to 2.5 GHz, divided into 10 MHz slots for enhanced visualisation. The temporal representation covers 6 s, with millisecond resolution, allowing observation of the spectrum’s dynamics. A high concentration of activity, characterised by a transition from blue to red (low to high saturation), is evident in the first 30 MHz of the spectrum, indicating the intensive use of Wi-Fi channels 1–3 and consequent congestion. Moving toward higher frequencies, the activity decreases, showing weaker and more intermittent signals, suggesting the presence of other networks or interferences with less impact. This information is crucial for optimising the Wi-Fi network, enabling the selection of channels that are less congested and minimising interference. Consequently, spectrogram analysis was used to estimate the percentage of the spectrum occupied by interfering networks.

To assess the spectrum conditions in each environment, spectrogram captures were taken using the Anritsu MS2090A spectrum analyser. Figure 4 shows each capture visualising seven seconds of the analysed spectrum, providing a snapshot of frequency activity immediately before conducting the performance tests. These spectrograms offer insight into the spectral occupancy within the three selected environments, highlighting variations in interference levels across the three usable frequency bands.

In the residential environment, the spectrograms reveal minimal spectral activity across all three frequency bands. In the 2.4 GHz band, the spectrum is mostly blue, indicating low levels of interference and occasional transmissions. The 5 GHz band also shows predominantly blue areas, reflecting sparse utilisation, which is typical in residential areas where fewer devices operate on this band. In the 6 GHz band, the spectrogram remains uniformly blue, highlighting negligible activity. This aligns with the fact that 6 GHz Wi-Fi technology is still emerging and not widely adopted in residential settings. Notably, this residential environment corresponds to a less densely populated area, more isolated than typical urban residential zones, which contributes to the overall lower spectral activity observed.

In contrast, the laboratory environment shows a substantial increase in spectral activity, especially in the 2.4 GHz and 5 GHz bands. The 2.4 GHz spectrogram displays green, yellow, and occasional red regions, signifying moderate to high levels of interference and active transmissions. This reflects the dense deployment of legacy Wi-Fi systems in laboratory environments, where multiple devices often operate simultaneously. The 5 GHz band also exhibits significant activity, with visible signal peaks indicating higher utilisation compared to the residential setting. This is likely due to the preference for 5 GHz Wi-Fi in laboratories to reduce interference and support higher data rates. However, the 6 GHz band shows minimal activity, with mostly blue regions and a few isolated peaks. This suggests limited use, likely for experimental purposes or controlled testing of next-generation Wi-Fi equipment.

The industrial environment exhibits the highest levels of spectrum saturation, particularly in the 2.4 GHz band. The spectrogram for this band is dominated by green, yellow, and red regions, indicating intense interference and congestion. This is consistent with the high usage of the 2.4 GHz band in industrial settings, where legacy Wi-Fi systems and other communication technologies coexist. The 5 GHz band shows moderate activity, with visible signal peaks that suggest active transmissions but less interference compared to the 2.4 GHz band. However, the 6 GHz band remains largely unused, with a predominantly blue spectrum and only minor peaks. This indicates that 6 GHz Wi-Fi technology has yet to see significant adoption in industrial environments, likely due to its recent introduction and the high cost of upgrading existing systems.

These observations highlight the clear differences in spectrum utilisation across the three environments. The 2.4 GHz band consistently exhibits the highest levels of activity, particularly in industrial and laboratory settings, due to its widespread adoption and compatibility with legacy systems. The 5 GHz band shows moderate to high utilisation in environments where reduced interference and higher data rates are critical, such as laboratories and industrial zones. In contrast, the 6 GHz band remains largely unused across all environments, reflecting its nascent stage of adoption.

This analysis underscores the impact of deployment scenarios on spectrum use and interference levels, providing valuable insights into the performance of IEEE 802.11 technologies in diverse real-world conditions. The numerical summary of this analysis is presented in

Table 4. Spectrum analysis in the relevant environments.

Frequency Bands	Spectrum Characteristics	Relevant Environments
		Residential	Laboratory	Factory
2.4 GHz	Free spectrum	91.33 %	50.89 %	62.13 %
	Peak signal interference	−73.7 dBm	−48.75 dBm	−73.19 dBm
5 GHz	Free spectrum	99.71 %	67.13 %	87.14 %
	Peak signal interference	−63.81 dBm	−40.63 dBm	−64.71 dBm
6 GHz	Free spectrum	99.97 %	99.81 %	99.97 %
	Peak signal interference	−60.17 dBm	−62.81 dBm	−60.72 dBm

, illustrating the spectrum occupancy for each environment. Nevertheless, it is crucial to acknowledge that these findings are indicative of specific moments and may be influenced by dynamic factors, such as the movement of obstacles and interference from other networks, which continuously modify the spectrum in real time.

5. Results

This section presents the results obtained from two stages, conducted in three representative environments and using the previously specified Wi-Fi technologies. The structure of this section mirrors the design of the two iterations of the testing campaign. In the first stage, network latency, specifically the Round-Trip Time (RTT), is analysed through the evaluation of test packet propagation. Subsequently, the second stage focuses on determining the operational thresholds for uplink transmission, specifically evaluating TCP and UDP traffic performance to characterise network stability and throughput efficiency.

5.1. Round-Trip Time (RTT)

This first stage presents a quantitative analysis of the network latency observed across the three previously defined environments, measured at a distance of 5 m. The experiment consisted of transmitting 1000 ICMP echo request (ping) messages, repeated over three consecutive iterations, yielding a total of 3000 transmitted packets. A fixed inter-packet interval of 5 ms was maintained to ensure consistent temporal spacing between transmissions. Each “ping” message had a payload size of 64 bytes.

The performance assessment was carried out by measuring the RTT between a Wi-Fi client, operating on a desktop computer, and the associated access point, as depicted in Figure 2. The results were independently analysed for each of the three environments considered. Data visualisation was performed using box-and-whisker plots, where the results from all three test iterations were aggregated to enhance clarity. These plots illustrate the maximum ($RT{T}_{max}$), minimum ($RT{T}_{min}$), and mean ($RT{T}_{mean}$) RTT values, with the mean denoted by a red horizontal line.

To further quantify the consistency of network performance, the mean deviation ($mdev$) was computed. This metric represents the average absolute deviation from the mean RTT value across all test iterations. A lower mdev indicates a more stable environment, where RTT values exhibit minimal variation, suggesting lower levels of interference. Conversely, higher mdev values indicate increased performance fluctuations, likely due to environmental factors such as interference, congestion, or varying channel conditions. Additionally, the dispersion of RTT values, as captured by $mdev$, is proportionally reflected in the size of the box in the box-and-whisker plots, providing a visual representation of performance variability.

5.1.1. Laboratory

The results are arranged from left to right in ascending order of frequency. As shown in Figure 5, the highest RTT levels are observed in the 2.4 GHz band, which has the highest number of interfering signals. This saturation of the spectrum is observed both in high values of the mean deviation of the RTT in Figure 5 and in the spectrogram in Figure 4 for the laboratory environment.

It should be noted that as we increase the frequency band of the Wi-Fi network, we decrease $RT{T}_{mean}$ value. This is because most of the adjacent Wi-Fi networks are configured in 2.4 GHz and 5 GHz bands. However, as seen in Figure 5, the 6 GHz band experiences the least interference, reflected in RTT${}_{max}$ values below 5 ms. Additionally, the measured mdev remains below 0.5 ms, indicating a stable and consistent performance.

These practical measurements indicate that the RTT performance achieved by the Wi-Fi 7 standard not only fails to surpass previous versions but, in some cases, actually falls below them, contradicting theoretical expectations of improvement. Specifically, tests conducted using an access point fully compliant with the 802.11be (Wi-Fi 7) specification yielded mean $RT{T}_{mean}$ values below 15 ms in the 2.4 GHz band and 10 ms in the 5 GHz band, with a higher level of dispersion as well. These findings suggest that, although the latest Wi-Fi standards theoretically have the potential to meet the strict latency requirements of industrial applications, Wi-Fi 7 devices have yet to deliver the expected benefits. This is mainly due to incomplete software and driver-level integration, which currently limits the full exploitation of the advanced features and capabilities of Wi-Fi 7.

5.1.2. Residential Area

In this environment, interfering signals are also present in the 2.4 GHz band, but to a lesser extent than in the laboratory environment. As can be seen from results in Figure 6, WiFi${}^6$_2.4${}^{20}$ technology demonstrates higher robustness than WiFi${}^4$_2.4${}^{20}$ technology, as it shows less variability in RTT results obtained. While WiFi${}^6$_2.4${}^{20}$ technology presents an $RT{T}_{max}<25$ ms and $mdev\approx 1$ ms, WiFi${}^4$_2.4${}^{20}$ technology offers an $RT{T}_{max}<30$ ms and $mdev\approx 1.5$ ms, demonstrating that WiFi${}^6$_2.4${}^{20}$ technology is more stable in a partially occupied spectrum. Although Wi-Fi 6 exhibits a slight performance improvement in the 2.4 GHz band, the results demonstrate that the performance of Wi-Fi 7 closely aligns with that of Wi-Fi 4. This observation suggests that Wi-Fi 7 is functioning in a legacy compatibility mode rather than utilising its inherent technological advancements, thereby constraining its anticipated performance enhancements.

It is important to highlight that the remaining evaluated technologies, operating in the 5 GHz band, exhibit significantly lower latency, with $RT{T}_{max}<5$ ms and $mdev<0.5$ ms. This improved performance is attributed to the lower spectrum congestion in the residential environment. Additionally, it is clearly observed that the advancements in Wi-Fi 6 enhance stability even at higher bandwidths. However, this is only achievable in an environment with reduced interference, as increasing the channel bandwidth inherently raises the probability of inter-symbol interference and signal degradation due to multipath effects and spectral overlap. Furthermore, the results once again demonstrate that Wi-Fi 7 lags, delivering performance comparable to that of Wi-Fi 5, confirming that it is operating in compatibility mode due to the lack of maturity in its drivers.

5.1.3. Factory

The evaluated factory environment presents a scenario with significant interference in the 2.4 GHz band, stemming from multiple industrial sources. This environment is classified as an indoor non-LoS setting, which naturally contributes to higher levels of interference compared to typical residential environments. However, the performance of the latest Wi-Fi standards remains robust in this more challenging setting, as shown in Figure 7. Specifically, WiFi${}^6$_2.4${}^{20}$ and WiFi${}^4$_2.4${}^{20}$ technologies achieve values of $RT{T}_{max}$ below 20 ms and mdev around 1.5 ms. Additionally, it is evident that in the 2.4 GHz band, increasing the channel bandwidth significantly elevates the RTT, as shown with WiFi${}^7$_2.4${}^{80}$ and WiFi${}^7$_2.4${}^{160}$. This effect is primarily due to higher susceptibility to interference, increased contention in the shared spectrum, and the inherent limitations of lower-frequency bands, which exacerbate congestion and retransmissions, ultimately impacting latency stability. In contrast, the other evaluated technologies reach $RT{T}_{max}$ values less than 5 ms and mdev values below 0.5 ms, providing highly satisfactory results despite additional interference typical of industrial environments.

The results of the analysis indicate that the WiFi${}^{6E}$_6${}^{80}$ and WiFi${}^{6E}$_6${}^{160}$ configurations provide the most optimal performance in terms of latency for short to medium distances, regardless of the environment evaluated and without taking LoS into account. This is primarily because they operate in the 6 GHz band, which remains largely unsaturated due to the limited number of devices currently capable of utilising this portion of the spectrum.

5.2. Peak Throughput

This section presents a comprehensive performance evaluation of different Wi-Fi technologies in various environments, considering three distinct distances: short, medium, and long. Analysing performance at varying distances allows for a detailed assessment of signal attenuation, multipath effects, and throughput degradation, providing insight into the robustness of each Wi-Fi standard under different propagation conditions. Both TCP and UDP traffic are measured to capture the distinct characteristics of connection-oriented and connection-less communication. TCP incorporates congestion control and retransmission mechanisms to ensure reliable data transfer, making it highly sensitive to network conditions and latency variations. In contrast, UDP, which lacks reliability mechanisms, prioritises low-latency transmission, making it more representative of real-time applications such as VoIP, video streaming, and online gaming. Evaluating both transport protocols provides a holistic view of Wi-Fi performance across different traffic profiles.

This study focuses on uplink traffic, or traffic from clients to AP, because unlike downlink traffic, uplink transmission is more susceptible to channel contention and medium access constraints, particularly in environments with high device density. The evaluation of uplink performance allows for a broader understanding of how different Wi-Fi generations handle resource allocation and interference mitigation.

Performance measurements are performed using iperf3, a widely adopted tool for precise network throughput, latency, and jitter analysis. A Raspberry Pi connected via Ethernet is designated as the server to provide a stable, low-latency endpoint, minimising external variability. The desktop computer, acting as the client, generates Wi-Fi traffic, ensuring that performance metrics accurately reflect wireless link behaviour without interference from wired network inconsistencies. This experimental setup ensures a controlled environment for evaluating the efficiency and stability of Wi-Fi technologies in real-world deployment scenarios.

5.2.1. Laboratory

For this environment, which presents higher spectrum saturation in the 2.4 GHz and 5 GHz bands, all technologies detailed in Table 3 have been evaluated. The results are presented separately in Figure 8 to account for the distinct performance characteristics of each frequency band. Specifically, on the left side of Figure 8, the results obtained for the 2.4 GHz band are shown, while on the right side, results for the 5 GHz and 6 GHz bands are shown. In the 2.4 GHz band, it is observed that the results for WiFi${}^7$_2.4${}^{20}$ technology are slightly lower than those for WiFi${}^6$_2.4${}^{20}$ technology. This result demonstrates that, even under optimal conditions, the achieved bitrates are at least 30% lower than the theoretical maximum bitrate. In the 5GHz band, better results are obtained with WiFi${}^7$_5${}^{80}$ technology than with WiFi${}^6$_5${}^{160}$. On the other hand, it is worth noting high peak throughput results obtained in both 5GHz and 6 GHz bands due to low or no spectrum saturation in this environment, reaching up to 940 Mbps with WiFi${}^{6E}$_6${}^{160}$ technology.

5.2.2. Residential Area

Experimental throughput results for TCP and UDP traffic were obtained in an outdoor residential environment with LoS, as detailed in Table 3. As shown in Figure 9, the highest throughput was recorded in the 5 GHz and 6 GHz bands. Specifically, WiFi${}^{6E}$_6${}^{160}$ and WiFi${}^7$_5${}^{160}$ delivered the best performance at short distances, achieving approximately 1720 Mbps and 1700 Mbps for TCP traffic. At longer distances, WiFi${}^6$_5${}^{160}$ outperformed other technologies, reaching up to 1450 Mbps. This superior performance can be attributed to minimal interference in the 5 GHz and 6 GHz bands, combined with the high bandwidth (BW = 160 MHz) enabled by these configurations, supported by Equations (2) and (4).

Additionally, it is observed that Wi-Fi 6 consistently outperforms WiFi${}^5$ and WiFi${}^4$ technologies, except at CD = 60 m, where a decrease in performance is evident. At this distance, the efficiency of Wi-Fi 5 and Wi-Fi 4 remains relatively stable, with low jitter and no packet loss, likely due to their narrower channel bandwidths and lower spectral demands, which reduce sensitivity to external factors. In contrast, Wi-Fi 6 configurations exhibit increased jitter and packet loss, suggesting that at longer distances, the use of wider channels (80 MHz and 160 MHz) may lead to greater susceptibility to channel fading, multipath effects, and overall spectral inefficiencies, ultimately impacting performance.

In the case of UDP traffic, the WiFi${}^{6E}$_6${}^{80}$ configuration outperforms WiFi${}^{6E}$_6${}^{160}$, suggesting that the narrower bandwidth (80 MHz) may offer better performance in specific environments with less interference. However, none of the other technologies evaluated exceed the performance levels achieved by WiFi${}^{6E}$_6${}^{80}$ and WiFi${}^{6E}$_6${}^{160}$. This highlights that the selection of the frequency band and available bandwidth are the most influential factors in network performance across all Wi-Fi standards.

For Wi-Fi 7, the performance shows some inconsistency in comparison with Wi-Fi 6. At shorter distances (C.D. = 5 m and 15 m), Wi-Fi 7 achieves comparable performance, with slightly better throughput and lower jitter in some cases. However, at longer distances (CD = 60 m), the performance of Wi-Fi 7 fluctuates, with higher jitter and some packet loss observed, suggesting that current Wi-Fi 7 devices and drivers may still be undergoing optimisation. These results highlight that while Wi-Fi 7 offers theoretical advantages in terms of speed and capacity, its current performance might be hindered by early-stage driver and firmware implementations.

5.2.3. Factory

In this non-LoS indoor environment, tests could only be carried out at medium (CD = 15 m) and long (CD = 60 m) distances due to limitations in a factory undergoing digitalisation. Therefore, Figure 10 shows the experimental results obtained for TCP and UDP traffic.

The highest performance was observed with WiFi${}^7$_5${}^{160}$, exceeding 1500 Mbps for UDP traffic. However, this configuration also exhibited the lowest ratio between theoretical and actual bitrates, indicating significant room for optimisation in this technology. The next best-performing configuration was WiFi${}^6$_5${}^{160}$, which slightly outperformed WiFi${}^{6E}$_6${}^{160}$ for both TCP and UDP traffic. At a distance of 15 m, Wi-Fi 6 achieved approximately 1400 Mbps for TCP and 1300 Mbps for UDP, while Wi-Fi 6E displayed a comparable performance. At longer distances, the impact of signal attenuation at higher frequencies became evident, with throughput dropping to around 640 Mbps for both types of traffic.

This behaviour is consistent with the logarithmic signal attenuation over distance described by the path loss model in Equation (4), especially at higher frequencies. In particular, the experimental results obtained with WiFi${}^{6E}$_6${}^{160}$ for medium distance in indoor non-LoS scenarios are superior to those obtained in the previous two environments. This may be due to some signal reflection effect from active machinery and manufacturing processes during tests, as the rest of the results do not show such marked differences. These findings suggest that Wi-Fi 6E operation in the 6 GHz band may benefit from multipath propagation effects in complex indoor industrial settings, reinforcing its potential for enhanced performance in such environments.

As expected, Wi-Fi 6 outperformed previous generations across all conditions. However, the results for Wi-Fi 7 show some variability. At 15 m, they achieved slightly better throughput than Wi-Fi 6 in some configurations, but at 60 m, its performance fluctuated and was, in some cases, lower than that of Wi-Fi 6. This suggests that the current generation of Wi-Fi 7 devices and drivers may still be in an early stage of optimisation, which could lead to inconsistencies in real-world performance. Further refinements in firmware, driver implementation, and hardware optimisation may be necessary to fully exploit its potential.

At long distances (CD = 60 m), WiFi${}^6$_5${}^{40}$ exhibited slightly lower performance than WiFi${}^6$_5${}^{80}$, with differences of approximately 80 Mbps in TCP and 50 Mbps in UDP. This indicates that the additional bandwidth provided by 80 MHz channels may help compensate for propagation losses, particularly in the 6 GHz band. The results confirm the significant improvements introduced by Wi-Fi 6 in challenging environments with obstacles. Although Wi-Fi 7 appears promising, its current performance appears to be influenced by early-stage implementations rather than fundamental technological limitations. Future tests with more mature devices and software updates could provide a more accurate assessment of their real-world capabilities.

The results presented in

Table 5. Analysis of jitter and packet loss for Wi-Fi transmissions at two different distances.

Acronym	UDP Traffic
	C.D. = 15 m		C.D. = 60 m
	Jitter (ms)	Packet Loss (%)	Jitter (ms)	Packet Loss (%)
WiFi${}^6$_2.4${}^{20}$	0.171	0.14	0.749	0.16
WiFi${}^4$_2.4${}^{20}$	0.502	0.32	0.21	0.45
WiFi${}^5$_5${}^{40}$	0.048	1.5	0.067	0.32
WiFi${}^6$_5${}^{40}$	0.042	0.5	0.075	0.1
WiFi${}^7$_5${}^{40}$	0.045	1.2	0.12	0.5
WiFi${}^{6E}$_6${}^{80}$	0.022	0.44	0.085	1.6
WiFi${}^7$_6${}^{80}$	0.027	0.5	0.1	1.8
WiFi${}^{6E}$_6${}^{160}$	0.016	0.39	0.041	2.1
WiFi${}^7$_6${}^{160}$	0.02	0.4	0.065	2.4

show a general trend of improvement in the performance of the newer Wi-Fi technologies, although with some variations. In particular, WiFi${}^6$ exhibits the best performance in terms of jitter at both distances, along with relatively low packet loss. A comparison across frequency bands shows that Wi-Fi technologies operating in the 2.4 GHz band (WiFi${}^6$_2.4${}^{20}$ and WiFi${}^4$_2.4${}^{20}$) exhibit higher jitter values than their counterparts in 5 GHz and 6 GHz; they also maintain consistently lower packet loss, especially at longer distances. This behaviour can be attributed to the superior propagation characteristics of the 2.4 GHz band, which allows for better signal penetration and stability over distance, despite higher latency.

Furthermore, while WiFi${}^7$_5${}^{40}$ and WiFi${}^7$_6${}^{80}$ show a slight improvement in jitter at shorter distances, as the distance increases, the performance of Wi-Fi 7 appears to become less consistent, with an increase in both jitter and packet loss. This suggests that current implementations, including implementations of its devices and drivers, have not yet reached optimal maturity. Despite this, the evolution of Wi-Fi 7 is expected to improve these aspects as hardware and software optimisations are refined in future developments.

6. Discussion of Results

The experimental findings in residential, laboratory, and industrial environments demonstrate that WiFi${}^{6E}$_6${}^{160}$ can deliver peak throughputs exceeding 1.6 Gb/s under LoS conditions and low spectral occupancy. However, performance degrades significantly with distance or in the presence of obstructions, particularly in industrial scenarios, similar to the issues shown in [15]. In such cases, throughput reductions of up to 40% were observed beyond 60 m, whereas 80 MHz configurations exhibited more gradual and stable degradation. These results indicate that effective system performance is not determined only by physical-layer capabilities but rather by the interplay between SNR, channel width, and the spectral environment.

This behaviour aligns with previous findings in which WiFi${}^6$_5${}^{80}$ was shown to outperform WiFi${}^5$ only up to approximately 50 m in outdoor scenarios [30]. Similarly, indoor measurements indicate that WiFi${}^{6E}$_6${}^{160}$ throughput drops below 1 Gb/s beyond 14 m [15]. In industrial outdoor settings, broader configurations such as WiFi${}^{6E}$_6${}^{160}$ offer high performance under LoS, whereas narrower configurations like WiFi${}^6$_6${}^{80}$ or WiFi${}^6$_5${}^{20}$ exhibit superior reliability and latency in non-LoS conditions [17]. Moreover, 160 MHz configurations appear more sensitive to external interference, which limits their applicability in congested environments. In contrast, 80 MHz channels offer a favourable trade-off between throughput and robustness, especially in scenarios with multiple obstacles or overlapping networks. Empirical measurements confirm that such configurations can maintain average RTTs below 2.5 ms with low jitter, making them suitable for latency-sensitive industrial applications.

These results support a deployment model in which 80 MHz channels are used as the default configuration for robust performance, while 160 MHz channels are reserved for short-range links with low spectral congestion. This approach is well aligned with the requirements of Industry 5.0, optimising stability without compromising capacity. Although other wireless technologies such as 5G have been considered a promising candidate for industrial applications due to their theoretical advantages in latency and reliability [8], the deployment, configuration, and ongoing maintenance remain significantly more complex and costly compared to Wi-Fi-based solutions. In contrast, experimental evidence obtained using commercial hardware shows that WiFi${}^{6E}$ offers a competitive, cost-effective, and sufficiently robust alternative for real-world industrial environments. This is particularly true when its high-efficiency features are fully utilised. The IEEE 802.11ax standard introduces several enhancements aimed at improving spectral efficiency and throughput density, including OFDMA, MU-MIMO, spatial reuse, and dynamic sensitivity control [4]. These mechanisms offer substantial potential, although their effectiveness strongly depends on context-aware tuning, thus motivating future research into the dynamic and adaptive optimisation of industrial-grade WiFi networks.

A summary of the most suitable Wi-Fi configurations by performance metric is shown in

Table 6. Summary of the Wi-Fi technologies’ performance for each environment and communications distance.

Environment	C.D.	RTT Mean	Peak TCP Traffic	Peak UDP Traffic
Residential área	5 m	WiFi${}^5$_5${}^{40}$	WiFi${}^{6E}$_6${}^{160}$	WiFi${}^{6E}$_6${}^{80}$
Laboratory zone	5 m	WiFi${}^{6E}$_6${}^{80}$	WiFi${}^{6E}$_6${}^{160}$	WiFi${}^{6E}$_6${}^{80}$
	15 m	WiFi${}^{6E}$_6${}^{80}$	WiFi${}^{6E}$_6${}^{160}$	WiFi${}^{6E}$_6${}^{160}$
	60 m	WiFi${}^{6E}$_6${}^{80}$	WiFi${}^{6E}$_6${}^{160}$	WiFi${}^{6E}$_6${}^{160}$
Factory	15 m	WiFi${}^6$_5${}^{40}$	WiFi${}^{6E}$_6${}^{160}$	WiFi${}^{6E}$_6${}^{160}$
	60 m	WiFi${}^6$_5${}^{40}$	WiFi${}^{6E}$_6${}^{160}$	WiFi${}^{6E}$_6${}^{160}$

. As discussed in previous sections, for TCP throughput, WiFi${}^{6E}$_6${}^{160}$ consistently delivers the highest performance regardless of the communication distance. However, for UDP traffic at short distances, the best-performing technology is WiFi${}^{6E}$_6${}^{80}$. The technologies with the lowest $RT{T}_{mean}$ results for the various environments differ from each other, although all technologies have $RT{T}_{mean}$ results below 2.5 ms, as shown in Figure 5, for all evaluated environments. Nevertheless, WiFi${}^6$_5${}^{40}$ technology stands out as the most suitable option in environments with minimal interference, such as residential areas. In contrast, in environments with overlapping access points and higher interference levels, WiFi${}^{6E}$_6${}^{80}$ offers the best results among the Wi-Fi technologies evaluated.

Based on current findings, several future research directions are proposed to fully exploit the capabilities of WiFi${}^6$, WiFi${}^{6E}$, and the emerging WiFi${}^7$ standard, whose features are increasingly geared towards deterministic performance. Enhancements such as improved multi-link operation (MLO), time-sensitive capabilities, and more flexible channel access are designed to meet the stringent latency, reliability, and synchronisation requirements of Industry 5.0. These developments position the latest generations of Wi-Fi as strong candidates for time-critical and automation-driven industrial deployments.

One key area for future work involves mobility and seamless handover. The promising performance of the 6 GHz band in short-range communications motivates a deeper investigation into the challenges of deploying WiFi${}^{6E}$_6${}^{160}$ configurations in mobile scenarios, both indoors and outdoors. Particular attention should be given to optimising Fast Roaming procedures to minimise handover latency between access points. This is especially relevant for industrial environments involving autonomous robot fleets, where mobility and uninterrupted connectivity are essential for system coordination and safety.

Another critical direction is the development of centralised medium-access scheduling strategies. Future work should explore the coordinated use of OFDMA and Triggered Uplink Access (TUA), introduced in IEEE 802.11ax to improve channel utilisation and reduce contention in dense deployments with heterogeneous traffic patterns. Such approaches could improve determinism and enable more predictable network behaviour across coexisting devices and services.

Finally, the integration of artificial intelligence techniques presents a promising avenue to improve adaptability and resilience. Research should focus on the real-time generation of medium access schedules and dynamic QoS management using machine learning models. These would allow WiFi networks to self-optimise in response to fluctuations in load, topology, and interference, thereby supporting autonomous operation and adaptive service provisioning in complex industrial settings. These lines of investigation aim to consolidate high-efficiency WiFi as a core enabler of future industrial wireless infrastructures, capable of meeting the evolving demands of flexibility, determinism, and scalability inherent to Industry 5.0.

7. Conclusions

This paper has successfully achieved its objective of evaluating the performance and limitations of Wi-Fi, across multiple generations of IEEE 802.11 standards, from Wi-Fi 4 (IEEE 802.11n) to early adoptions of Wi-Fi 7 (IEEE 802.11be). To accomplish this, we deployed a comprehensive testbed in three distinct environments (residential, laboratory, and industrial) and tested six representative Wi-Fi configurations, each defined by a specific combination of frequency band, channel bandwidth, and IEEE 802.11 standard. All experiments were conducted under consistent conditions to simulate practical deployments, and we analysed the spectrum occupancy in each band to account for interference while measuring network latency and throughput. Experimental results show that Wi-Fi 6/6E (IEEE 802.11ax) consistently outperforms earlier standards across diverse environments. Configurations with wider bandwidths, such as Wi-Fi 6 with 160 MHz channels, deliver up to 20% higher throughput at short and medium ranges compared to Wi-Fi 4 and 5, maintaining performance even under moderate interference. Wi-Fi 7, despite its theoretical advantages, exhibited throughput up to 10% lower than Wi-Fi 6 in some cases, suggesting current implementations are not yet fully optimised.

Latency improvements are also evident: Wi-Fi 6 reduces delays by approximately 15% compared to Wi-Fi 5 and also offers more stable performance suitable for time-critical applications. However, in high-traffic environments, latency remains variable due to the limitations of contention-based access. Interference impacts remain significant, with latency and throughput fluctuating by up to 30% between congested and low-traffic scenarios. Even within a single band, channel performance may differ by up to 20%, and wider channels, while boosting throughput, show increased sensitivity to interference. This underscores the need for strategic spectrum and channel selection tailored to deployment conditions and reliability requirements. Despite these limitations, the performance levels achieved by Wi-Fi 6 and Wi-Fi 7 are acceptable for many industrial applications, particularly those related to digitalisation and non-critical monitoring. However, their suitability for latency-critical control tasks is still limited due to the inherent unpredictability of contention-based access mechanisms. One promising development in this context is the introduction of Resource Units (RUs), which are already part of the IEEE 802.11ax standard. These elements provide a finer granularity for spectrum allocation, enabling better scheduling and control of medium access. Although the standard does not yet define a complete framework for their deterministic use, RUs represent a key foundation for reducing internal contention and enhancing temporal predictability in future industrial Wi-Fi deployments.