In the following chapter the telecommunication technologies WiFi, 4G and 5G are introduced and their main characteristics are explained. Further they are compared concerning their main differences, technical details and advantages. These details are then translated into a competitive rating which will serve as the basis for the use case matching.
WiFi refers to the different IEEE 802.11 networking standards. It is commonly used to create wireless local area networks (WLAN) and thus, connecting devices to the internet. However, it still faces challenges, as will be outlined in the following chapter. Standardizations for WiFi 4 and WiFi 5 were finalized in October 2009 and December 2013. Since WiFi 5 only employs the 5GHz spectrum, routers combine the use of WiFi 4 and 5, in order to ensure backward compatibility and supplementary utilization of the 2,4 GHz spectrum. For that reason, WiFi 4 and 5 will be combined for comparison. WiFi 6 (or 802.11ax) was standardized in September 2019. As of now, routers are available, but preceding standards are still used predominately because deployment just started. WiFi 6 primarily improves features that were introduced in WiFi 4 and 5. According to the WiFi Alliance, key benefits are higher data rates, increased capacity, better performance in device-teeming environments and improved power efficiency. Because of its availability and simplicity to deploy, WLAN has also become the de facto standard for office connectivity.
Long-Term-Evolution-Advanced (LTE-A) is the fourth-generation mobile communications standard. Being standardized since April 2011, it is used to create mobile networks. It builds upon LTE, which was standardized in 2008 and only requires a software update There is some debate whether the initial release of LTE is considered a 4G network, as it did not formally fulfill the requirements of the responsible certification authority, the International Telecommunication Union Radiocommunication Sector (ITU-R). The focus of improvement, compared to LTE, are higher data rates: LTE provides up to 300 Mbit/s in the downlink, but typically only realistically only reached 12,5 Mbit/s, while LTE-A provides up to 1200 Mbit/s with 100 Mbit/s being reached typically. Other areas of improvement are faster switching between power states to save energy and improved performance at the cell edge. Base stations only require a software upgrade. In manufacturing, LTE-A is primarily used to connect networks, rather than single devices, to the internet, often only as a fail-safe solution.
5G is the fifth mobile communications standard. Besides being the next generation of mobile networks, 5G was developed with the intention to make it the next standard for connectivity within the industry. To that end, ITU-R identified three usage scenarios for 5G networks, based on ongoing changes and trends, such as the need for low latency communications and greater device density:
1. Enhanced Mobile Broadband (eMBB) use cases focus on human-centric access to multi-media content, services and data. Therefore, higher data rates are the main requirement. Included use cases are edge- or cloud-computing as well as virtual and augmented reality.
2. Ultra-Reliable Low-Latency Communications (URLLC) use cases entail time- or otherwise critical data, such as control of industrial manufacturing or human-machine interaction. Thus, strict requirements regarding latency and availability must be met.
3. Massive Machine-Type Communications (mMTC) use cases are centered around massive numbers of devices that typically require only low data rates but high energy efficiency.
Based on those application scenarios, the requirements of the network of the future were derived. Standardization of the initial deployment, phase 1, was finished by June 2019. Phase 2, which will bring further improvements to latency and IoT support, is expected to be finished by June 2020. Deployments will widely be in the non-standalone (NSA) mode, i.e., a 5G radio access network (RAN) in combination with a 4G core network because investment costs are significantly lower. The RAN establishes the connection to the client, while the core network handles network functions such as user authentication and connection to the internet or different networks. As a result, some features of 5G will not be possible with existing infrastructure, rather, additional upgrades to base stations would be necessary.
5G is not a single technology but a close integration of multiple technologies and therefore, not all use cases may be realized by a single deployment. Instead, a network will be comprised of different deployment options, i.e., base stations with different cell sizes. Smaller cells use higher frequencies. In consequence, their range is lower, but allow higher data rates and device density. Therefore, a network will not offer the same capabilities ubiquitously. Instead, capabilities will be realized locally to meet the requirements of the usage. Thus, the full potential of 5G can only be realized by a combination of those deployment options. .
As aforementioned, 5G, 4G and WiFi employ different frequencies to transmit data. WiFi utilizes the unlicensed frequencies at 2,4 GHz and 5 GHz. These frequencies are part of the industrial-, medical- and scientific (ISM) bands, resulting in two main obstacles.
Firstly, interference management is challenging. As the ISM-Bands do not require any license, they are used quite heavily by a variety of network standards, e.g., Bluetooth, ZigBee and various other proprietary standards operate at 2,4 GHz. ZigBee is a so-called low-power wide-area network (LPWAN) that allows large coverage and little energy consumption, but only small data rates. Proprietary standards range from garage door openers to input devices for computers. Interference management is not only challenging because of the noise level and heavy usage, but because of the different standards operating parallelly. The 5 GHz ISM-Band is used less but suffers from bad propagation properties, especially for non-line-of-sight propagation, i.e., when obstacles are between the sender and receiver. 5G and 4G on the hand use licensed frequency bands. As a result, the noise level is lower and interference management is simple.
Secondly, there is a legal limit to the strength of emitted radiation, i.e., signal strength, and to the radiation present in the surrounding environment. The equivalent isotropically radiated power (EIRP) in the 2,4 GHz bands is limited to 100 mW, in the 5 GHz bands to 0,2 W (5,15 GHz-5,35 GHz) and 1 W (5,47 GHz-5,725 GHz). The former is only permitted for indoor use. For that reason, the signal strength of WiFi networks is limited. As a result, range and penetration are low. In 4G, the base stations send out up to two magnitudes more depending on the frequency range. Furthermore, the spectral power density, i.e., the power of radiation present in each frequency range in the surrounding environment, is limited. If the limit is reached, the device is prohibited from emitting additional radiation, i.e., transmit data. That way, increased device density leads to increased latency and decreased reliability. Therefore, even mobile devices such as smartphones can lead to disturbances. Furthermore, WiFi networks are especially susceptible to disturbances due to older standards being present, because of the backward combability in WiFi. As a result, a significant loss in efficiency can occur. In contrast, 4G and 5G networks are not backward compatible and the network manages used frequencies centrally, i.e., devices are told which frequency they should use. Therefore, mobile networks are very resilient.
To increase reliability and scalability, WiFi 6 employs technologies known from mobile networks, namely multi-user multiple-input multiple-output (MU-MIMO) and orthogonal frequency-division multiple access (OFDMA). As a result, the throughput per user is quadrupled in crowded areas, although single-user speed has only increased by 37%. Also, up to 8 devices will be able to communicate with the base station simultaneously. Depending on the base station, LTE networks allow several hundred up to a few thousand devices to communicate simultaneously. The scheduling in WiFi 6 changed. While access to older WiFi standards was contention-based, i.e., clients competed for resources, uplinks in WiFi 6 are scheduled to minimize conflicts. Because the device can sleep until the next scheduled transmission, energy efficiency is increased. However, prioritization of data is still not possible. In contrast, 4G has multiple mechanisms to manage traffic. Therefore, prioritized traffic is deterministic, i.e., it can be processed within a constant time frame. Only non-prioritized traffic can suffer from high latencies and high loss probabilities in areas of high device density. That is because there is no constant connection in 4G networks and thus, accesses are random. In 5G, prioritization will be an inherent component. Additionally, in 5G stand-alone (SA) network slices will be available and capable of reserving resources. To increase connection stability within areas with multiple WiFi networks present, WiFi 6 introduces basic service set coloring (BSS), i.e., adding a locally unique identifier to each network’s transmissions. That way, devices can distinguish transmissions of their network from those of neighboring networks. As a result, the transmissions of neighboring networks can be ignored, making simultaneous transmissions in neighboring cells possible. Therefore, WiFi 6 networks are better expandable than WiFi 4/5 networks, although many devices still cause problems due to legal limits and interference. Modifications to WiFi networks are limited because of the narrow capability profile. Mobile networks, such as 4G and 5G, on the other hand, are designed to combine many different base stations into a single network. For that reason, expansion is straightforward. However, modifications to a 4G network are costly, as core functions are realized by specialized hardware. 5G introduces software-defined Networking (SDN) and network functions virtualization (NFV) to remedy that. SDN solves network tasks through software instead of specialized hardware, while NFV centralizes control over network functions, previously done by routers and switches. Therefore, modification of a 5G network is simple. For all those reasons, 5G is best suited for (time-)critical application scenarios and for application scenarios with incomplete specifications that might require adaptation later on, such as Industry 4.0 use cases. Despite that, WiFi 6 improves the capabilities of WiFi networks greatly compared to WiFi 4/5.
Another aspect is cyber-security. WiFi only requires the SSID (network name) and a password. As a result, WiFi networks are inherently vulnerable to man-in-the-middle (MITM) attacks. Unfortunately, whitelisting MAC addresses to increase security does not prevent MITM attacks, as MAC addresses can be spoofed, i.e., faked. As data security of WiFi networks cannot be ensured, drastic measures for critical data were enacted within the industry. Critical data was secured using air-gap-security, i.e., separating the secured computer or wired network by a physical or mechanical gap. To transfer data to or from these secured computers or networks, removable storage mediums, e.g., USB drives, are necessary. However, air-gap-security is not the optimal solution either, as it results in significant downsides in usability and has been compromised nonetheless, e.g., by Stuxnet. Although WiFi 6 introduces a new security protocol with WPA3 (as opposed to WPA2), MITM attacks are still possible. Furthermore, new exploits were already found. To remedy these exploits, a non-backward compatible upgrade could be necessary. As a result, already rolled out hardware would become obsolete. Despite that, WPA3 is generally regarded as safer than WPA2. In contrast, 4G and 5G networks require unique sim cards to access the network, making them inherently hard to hack. While designing 5G, a key consideration was safety. To achieve improved safety, 5G employs well-proven 4G security mechanisms as well as new enhancements, e.g., mutual authentication, which prevents MITM attacks, and encryption of end-user data within the mobile network. For that reason, 5G is believed to be safer than 4G, although Software-defined Networking (SDN) and Network Functions Virtualization (NFV) might pose new risks. As outlined above, SDN and NFV result in the omission of hardware and centralization of network functions. Therefore, there are no more physical limits, i.e., hardware bottlenecks, to limit the access of hackers, as all services are centrally provided by software.
The different approach to roaming, i.e., how a client leaving a network or the coverage of a single access point (AP) is handled, has major implications for a particular set of use cases. Although WiFi has a protocol for seamless roaming, many clients do not support them. As a result, clients must connect anew once they lost connection to the access point due to being out of reach. Up to that point, the client would have suffered from sub-optimal performance, as there is no automatic handover. For use cases that require mobility, such as AGVs, those reconnects mean breaks until a connection to the new access point has been established. That way, AGV-traffic jams might occur, as every cell edge is a bottleneck. In contrast, mobile networks, such as 4G and 5G, incorporated seamless handovers a long time ago. As a result, clients are automatically connected to the optimal available access point. Therefore, 4G and 5G are better suited for those use-cases.