Perspective Chapter: Integration of Fiberoptic and Radio Frequency Channels for Hybrid 5G MIMO Network

1.1 Integration of optical fiber networks and wireless RF networks

The integration of fiberoptic networks and radio access networks (RANs) is a hybrid system that is called FiWi (Fiber Wireless). Such a system is considered to be a practical solution for broadband, wireless, and cellular applications. According to this principle, the transmission and reception of the signals is implemented by the antennas and by the base station, but the digital signal processing (DSP) occurs inside the optical fibers. These advanced communication networks execute data transmission over the 5G network, which supports large bandwidths in the order of GHz.

1.2 Base stations and antennas in modern 5G networks

Communication networks require the use of base stations in order to transmit data over a long range and to provide optimal coverage for edge units. Every radio frequency (RF) signal is transmitted and received by the antennas of the relevant base station. Some of these antennas are located on a mast, while others are located on buildings.

5G networks make use of antennas that support the Ka domain (27–40 [GHz]); hence, the transmit/receive rates are quite high [1].

In addition, we can classify the base stations according to the following types [2]:

• MBS-Macro Base Station: The MBS has a wide area of coverage—from 1 [km] and up to 20 [km]—and is widely used in current wireless communication environments.

• MiBS-Micro Base Station: The MiBS has an area of coverage of 400 [m]–2 [Km]. Usually, the MiBS is deployed at hotspots in order to compensate for any missed-out points and for signals which cannot be transmitted adequately from the MBS.

• PRBS-Pico Base Station: PBSs are deployed for covering areas of 4–200 [m]. This type of base station is used in shopping malls, office buildings, etc. to provide dedicated capacity to large-sized individual groups.

• FBS-Femto Base Station: FBSs are mainly designed for the home or small business places. Its coverage area is 10–20 [m].

1.3 5G communication networks

The 5G network provides down-link maximum throughput of up to 20 Gbps, improved quality of service (QoS), low latency, high coverage, high reliability, and economically affordable services compared to the previous network (4G LTE), which delivers speeds of between 42 Mbps and 90 Mbps [4].

5G is separated into two classes: 6 GHz 5G and Millimeter wave (mm Wave) 5G.

6 GHz is situated in the intermediate frequency band that operates as a mid-point between high volume and wide coverage to enable a suitable 5G-ready infrastructure. The 6 GHz spectrum provides large bandwidth with enhanced links efficiency. It provides continual channels that will decrease the necessity for links densification when intermediate band spectrum is unavailable, enabling ubiquitous and cost-effective 5G connectivity for all.

On the other hand, the mmWave 5G part offers diverse services and requires less spectrum compared to 6 GHz 5G. In addition, it provides very high-speed wireless communication, and offers ultra-wide bandwidth for next generation mobile networks.

1.4 Optical fiber

An optical fiber is a transmission medium used for the transmission of information as light pulses through a glass (silica) or plastic fiber [5, 6]. It was first established in the mid-1980s and is used in fiberoptic communications. One major benefit is transportation of data over longer distances and with much higher bandwidth with less loss, compared to coaxial cables.

The optical fiber includes four layers:

The central part of the structure demonstrates a cylindrical shape and is composed of dielectric material, typically possessing a unique refractive index. Encompassing this core section is the cladding, which comprises glass or plastic with a refractive index lower than that of the core section. The cladding section is surrounded by an additional elastic layer as a buffer, made of plastic, which protects the optical fiber from physical damage and scattering losses caused by the micro bending, and a jacket.

Figure 1 presents the structure of the four layers of an optical fiber.

1.5 The problems of modern 5G communication networks and their influence over the edge units

Today there are many different consumers and systems that use communication networks. These networks are linked through transmission lines (TLs) that constitute the wireless network. The main problem lies in the fact that the number of users in the same network has greatly increased, which causes high data loads and reduces the quality of communication. Therefore, it is necessary to transmit at high rates in order to deal with this challenge. One of the best options for achieving this goal is the use of hybrid communication. This option makes it possible to regulate loads generated in the TLs and transfer them to fiberoptic infrastructures that work at GHz rates.

2.1 5G-XHaul system

The 5G-XHaul infrastructure adopts a common fronthaul/backhaul (FH/BH) network solution, deploying a wealth of wireless technologies and hybrid active/passive optical transport, supporting flexible fronthaul split options [8].

A fronthaul network is defined as optical fibers through which the data is transmitted, and a backhaul network connects the access network to the core network.

The passive solution employs Wavelength Division Multiplexing-Passive Optical Networks (WDM-PONs), while the active solution adopts the highly versatile Time-Shared Optical Network (TSON) that enables the use of wider bandwidth. In addition, this system supports split options for fronthaul-based networks. The split principle is based on semi-transmission of DSP from the RU to the CU.

Cloud Radio Access Networks (C-RANs) aim to surmount these constraints by enabling the integration of Access Points (APs) or remote units (RUs) with a Base Band Units (BBU) pool hosted in a Central Unit (CU) through a series of transport links known as fronthaul. The RU wireless signals are commonly transported over an optical fronthaul network, using either digital transmission (e.g., Common Public Radio Interface (CPRI)), or analog transmission (radio-over-fiber – (RoF)).

However, C-RANs require very high transport bandwidth due to the traffic volume created by the sampled radio signals transported to the CU and the very tight delay and synchronization specifications.

To harness the advantages of C-RAN and overcome its associated challenges, equipment vendors are enhancing their fronthaul solutions and embracing cutting-edge wireless technologies like Sub-6GHz and 60GHz bands, which include advanced beam tracking and MIMO techniques. Additionally, they are adopting new flexible and dynamic Wavelength Division Multiplexing (WDM) optical networks.

The concept of flexible functional splits is addressed by appropriately combining servers with low processing power (cloudlets) and relatively large-scale datacenters (DCs) placed in the access and metro domains, respectively.

Figure 2 illustrates the 5G-XHaul physical infrastructure, where both fronthaul and backhaul services are delivered over a shared wired/wireless network infrastructure. In the fronthaul scenario, certain portions of the BBU processing can be carried out locally, while other parts are performed remotely at the Data Centers (DCs), allowing for a flexible split paradigm in C-RAN deployment. The BBUs are executed as virtual entities on general-purpose servers. As for the backhaul services, they establish connections between end-users and virtual machines hosted in the DCs.

There are several units responsible for the integrity of the 5G-XHaul system:

• Interfaces play a crucial role in managing protocol adaptation, as well as the mapping and aggregation or de-aggregation of traffic across various domains. These domains, such as wireless and optical, may have distinct protocol implementations and offer varying levels of capacity, ranging from Mbps in the wireless domain to tens of Gbps in TSON (Time-Sensitive Optical Networks). Additionally, granularity can differ, with the wireless domain supporting Kbps and TSON supporting up to 100 Mbps. At the optical network ingress point, like the TSON edge node, the interfaces are responsible for receiving traffic frames originating from both fixed and mobile users. They organize these frames into different buffers. Subsequently, the incoming traffic is aggregated into optical frames and allocated to appropriate timeslots and wavelengths based on the adopted queuing policy before being transmitted within the TSON domain.

• The Infrastructure Management Layer (IML) assumes the role of overseeing distinct technology domains. It is tasked with enabling multi-tenant operation by implementing cross-domain slicing and virtualization, which, in turn, allows for joint fronthaul and backhaul services across the shared infrastructure. Network and compute controllers situated within this layer are responsible for information retrieval and communication between domains, effectively enabling resource abstraction and virtualization.

• The Control Layer (CL) is responsible for cross-domain orchestration of virtual and real programming interfaces (PIs), created and exposed by the IML, having an overall view of all network domains. The CL provides end-to-end connectivity services deploying converged control and management procedures with guaranteed QoS.

• The Management and Service Orchestration Layer (MSOL) handles orchestration requirements for the delivery of network and compute services in the system.

The main advantage of the system is the fact that transport through optical fibers only requires low-energy consumption. The impact of centralization is expressed through transport network overloading introduced by fronthaul services, leaving limited resources for backhaul services.

Therefore, the main disadvantage of 5G-XHaul system is the backhaul end-to-end delay, which can be estimated using the following formula:

where uBH, e, πBH, s represent the backhaul-related network and server capacity, respectively, Ue, is the total capacity of e, and Πs, Πc are the total processing capacity of the DCs and the cloudlet, c, respectively.

2.2 RoF

Integration of RF wireless networks with optical networks enables the transmission of services over a high bandwidth for both mobile users and fixed terminals (broadband services) [6, 9].

A RoF system modulates radio signals over optical fibers enabling the transmission of modulated radio signals through optical fibers.

This technology involves low losses, making transmission possible over long distances, which is not possible for the direct transmission of radio signals. The RF signal is created at the base station (radio station). In order to transport the data to the base station, optical fiber backhaul is employed. RoF systems use fibers to transport a single or a number of different types of analogous carriers through them as follows:

The modulated laser transmits the RF signal created at the photodiode. Then, the photodiode is used to retrieve the signal at the receiver side.

Figure 3 illustrates how a signal reaches the edge equipment (e.g., a communication cable), and after transport through the edge equipment it moves through the fiber feed network and is converted to an analog signal that reaches the remote base unit (antenna), which supplies the wireless communication network for the edge units (consumers). Thus, data is sent and received in a RoF system.

2.2.1 Integration of hybrid PONs with RoF

Hybrid passive optical networks (PON) combine both WDM and time division multiplexing (TDM) into a single PON, offering reduced cost, high scalability, and increased data rates [10]. The system may have potential for the integration of in-built wireless networks with fiber access network RoF technology, which enables it to be used for the direct transmission of RF through the fiber without the need for frequency conversion at the receiver. In addition, the system uses digital modulation techniques like QPSK, M-PSK, M-QAM that provide high spectral efficiency, better utilization of bandwidth, and increased bitrate, while preserving minimal signal bandwidth [11, 12, 13, 14].

We now present a network composed of both RoF and PON. PON uses only fiber and passive components like splitters and combiners. The PON supports a maximum data rate of 100 Gbps by using the Orthogonal Frequency Division Multiplexing (OFDM) technique in an optical access network. The objective of PON is to get the fiberoptics as close as possible to the end-user—ideally right into the subscribers’ homes/offices.

The PON architecture consists of three essential devices: the optical line terminals (OLTs), the optical network units/optical network terminals (ONUs/ONTs), and the optical distribution network (ODN). On the OLT side, the transmitters produce a singular wavelength that carries data intended for a specific ONU. This wavelength is generated by modulating the RF signal using a DPSK sequence generator. Subsequently, the modulated RF signal is combined with a laser operating at wavelengths ranging from 193.1 to 193.8 THz.

There are two multiplexing techniques which are applied here: WDM-TDM and WDM-subcarrier multiplexing (SCM).

2.2.1.1 WDM- SCM

SCM is multiple RF carrier signals transmitted through optical fiber using single wavelength [11]. The most significant advantage of SCM in optical communications is its ability to place different optical carriers closely together. SCM must be used in conjunction with WDM to exploit a significant fraction of the fiber bandwidth. Using SCM/WDM for RoF results in high bandwidth over long distances. Therefore, the bandwidth utilization efficiency of SCM is much better than conventional optical WDM. The operation of SCM is similar to that of TDM, such that TDM is commonly used in digital transmission systems.

2.2.1.2 WDM-TDM

This method consists of two sub-methods, WDM and TDM [12]. WDM gives better utilization of the large bandwidth of optical fiber and can increase the capacity of the cable network. Through WDM, signals from two- or more-line systems are transmitted over the same fiber. TDM provides moderate bandwidth (compared to WDM), but more channels that can be used during optical fiber transmission.

Figure 4 presents the PON, cellular mobile base station, and wireless station that are integrated over the infrastructure of signal fiber in a Metropolitan Area Network (MAN), which is a computer network that connects edge units within a single town (or several towns/settlements).

This system displays several advantages [14]:

• High bandwidth available for internet usage - this benefit boosts the transmission speed and increases the rate of DSP.

• Low attenuation loss of optical fiber - due to the non-existence of the skin effect, which is strongly dependent on frequency, rising rapidly with increased frequency. Due to the fact that optical fiber is immune to external RF propagation, the optical fiber is also immune to high frequency transmission, and therefore the skin effect does not degrade the quality of transmission over the optical fiber.

• Immunity to RF interference - this advantage stems from the fact that optic fibers transmit signals as light waves rather than electrical impulses. Consequently, they remain impervious to EMI (electromagnetic interference) and do not generate any EMI themselves. Additionally, fiberoptic systems are immune to RFI (radio frequency interference) and do not emit any radiation, as they do not produce RF signals.

• Reduced power consumption - the simplification of each base station results in significantly lower costs for system installation and maintenance. Moreover, the reduced power level obviates the necessity for expensive frequency multiplexers and high-power amplifiers, which are currently utilized in base stations. As a result of implementing uncomplicated remote antenna units (RAUs) with reduced equipment, the overall power consumption is considerably reduced.

• They are difficult to tap into without being detected, which provides privacy and security.

• Dynamic capacity - for example, more capacity can be allocated to an area in accordance with the current needs, and then reallocated to other areas off-peak when allocating constant capacity would be a waste of resources.

• Centralized control for electronic circuits - this advantage allows the base stations to detect and transmit optical millimeter wave signals.

However, this system displays some drawbacks [15]:

• Due to dispersion effects in multi-mode fiber, every light beam travels a different distance as it propagates, and the ray arrives at different times at the fiber output so that the light pulse spreads in time, which can cause inter symbol interference (ISI), degrading system performance.

• Since nonlinear equipment is used, the likelihood of amortization increases, which decreases system performance.

• Noise from laser sources causes inaccuracies in the transmission of the analog signal (RF).

2.3 Hybrid 5G optical wireless SDN-based networks

The notion of intelligent heterogeneous networks (HetNets) has surfaced as a versatile and hierarchical solution for establishing cognitive fronthaul wireless architectures comprising multiple technologies. By combining three potent technologies, HetNets offer a compelling approach to address various communication challenges [16]: network function virtualization (NFV), software defined radio (SDR), and the software defined networks (SDNs).

3.1 Data stream parameters inside the SISO system

For a channel with AWGN, the Shannon-Harley formula is usually used to calculate the capacity of the channel [17]:

where BW is the bandwidth, described by the maximum channel capacity of communication channel at a certain time, and is measured in Hz.

The signal to noise ratio (SNR) is defined by the following formula:

where NR is the receiver noise and PR is the receiver power.

The power of the AWGN which defined as: NR≡Nadd=N0BW (where N0 defines AWGN) and additionally, there is multiplicative noise in the channel, then formula (4) can be rewritten as:

We now define the K parameter, which describes the fading rate of the signal (strong fading/weak fading), as [18]:

where Ico=LOScomponent power,Iinc=Multipath component power.

The LOS component describes signal loss along the path of direct visibility between both antennas. The multipath component is defined as interruptions that occur due to multiple reflections, diffraction, and scattering from various obstructions, which effect both the transmitter and receiver.

As a result, we acquire a channel capacity formula that depends on the K parameter:

For a single channel, we define the spectral efficiency (the rate of information that can be transmitted over a given bandwidth measured in [bps/Hz]) using the following formula:

In this equation, there are two extreme states of parameter K that we will examine:

• Given a channel with strong fading, for which K→0, we get log21=0. From this we conclude that Shannon’s theorem does not apply for strong fading.

• Given a channel with weak fading, for which K→∞, and AGWN, we will get spectral efficiency depending on the signal to noise ratio added SNRadd (Figure 5):

For this system, it was found that as the value of the fading coefficient K increases, a higher spectral efficiency is obtained, but at the same time, the influence of the fading coefficient is reduced if the system has a higher value of added SNR.

To calculate the BER, we define the noise probability density function (PDF) as the Rician PDF, written:

where σ is the standard deviation of Rician fading described as follows:

The I0 parameter described as Bessel function of the first kind and zero order.

Therefore, the formula for the BER calculation is as follows (Figure 6):

A Rician PDF indicates a change of the resulting BER in the system. It was found that for a larger fading coefficient, the BER is reduced.

3.2 MIMO channel with fading

First of all, we consider an AWGN MIMO system (i.e., only Gaussian or additive noise). In the first case, the parameters in the channel array can be considered to be statistically independent (therefore these variables are uncorrelated). We examine the combined optic-radio system shown in Figure 7.

Then, for M > N we acquire the following formula:

For a MIMO system, we define the term “interpretation”:

Interpretation: the power is divided equally between N channels, but each of them has a gain M, that is, the power for each channel is SN·M – the signal power divided by Nadd. The capacity or spectral efficiency is the same for each of the N input channels, and the total value is N times this amount.

For the MIMO system, there is a direct relationship between the number of channels and the channel gain and its spectral efficiency. This is reflected in the fact that as the number of channels N increases, the spectral efficiency increases correspondingly, along with the requirement for higher M channel gain (for some channels, the spectral efficiency is zero due to the lack of high enough channel gain).

The system was sampled discretely because the number of channels is an integer NM∈ℕ. On the other hand, in the second case the parameters are correlated, that is, coupled with each other, in a unified system on both sides (Figure 8).

As a result, we acquire the following formula:

Here, we have the same C or C∼ for all channels, but the total power is M·N·S, divided by Nadd.

The main difference obtained in the given case is the fact that the maximum spectral efficiency obtained on the scale tested was lower than that of the case in Eq. (14) Also, in this case the system was sampled discretely because the number of channels is an integer NM∈ℕ (Figure 9).

Next, we derived the final formulas for calculating the spectral efficiency of the system for the cases of only AWGN (additive) and of both AWGN and multiplicative noise.

• For the first case, in which the parameters are uncorrelated, and there is only AWGN we acquired Eq. (14).

• For the second case, in which the parameters are correlated on both sides, and there is only AWGN we acquired Eq. (15).

• For first case, in which the parameters are uncorrelated and there are both AWGN and multiplicative noise, we acquired the following formula:

Computational analysis of Eq. (16) for various M and N uncorrelated elements at the input multibeam antenna and output multibeam antenna of a MIMO system is shown in Figure 10.

For the MIMO system, there is a direct relationship between the number of channels and the channel gain and its spectral efficiency. This is reflected in the fact that as the number of channels N increases, the spectral efficiency increases correspondingly, along with the requirement for a higher M channel gain (for some channels, the spectral efficiency is zero due to the lack of high enough channel gain). It is important to note that in this case, the influence of the fading coefficient K was more significant in cases where the channel gain M was small, which degraded the value of the resulting spread.

• For the second case, in which the parameters are correlated on both sides, and there are both AWGN and multiplicative noise, we acquired the following formula:

Computational analysis of Eq. (17) for various M and N correlated elements at the input multibeam antenna and output multibeam antenna of a MIMO system is shown in Figure 11.

For the MIMO system, there is a direct relationship between the number of channels and the channel gain and its spectral efficiency. This is reflected in the fact that as the number of channels N increases, the spectral efficiency increases correspondingly, along with the requirement for a higher M channel gain (for some channels, the spectral efficiency is zero due to the lack of high enough channel gain). It is important to note that in this case, the influence of the fading coefficient K was more significant in cases where the channel gain M was small, which degraded the value of the resulting spectral efficiency. The main difference obtained in the given case is the fact that the maximum spectral efficiency obtained in the tested scale is higher than that of Eq. (16).

The BER calculation is performed by using the following formulas:

For the first case:

Computational analysis of Eq. (18) describing BER for various M and N uncorrelated elements at the input multibeam antenna and output multibeam antenna of a MIMO system is shown in Figure 12.

For the MIMO system, there is a direct relationship between the number of channels and the channel gain and its BER. This is reflected in the fact that as the number of channel N increases, along with the increment of channel gain M, the BER decreases, which improves system performance.

Computational analysis of the BER versus the standard deviation and the fading parameter K, using Eq. (18) describing the BER via these parameters for various M and N uncorrelated elements at the input and output multibeam uncorrelated antenna elements of MIMO system, is shown in Figure 13.

For this case, it can be seen that as the value of the fading coefficient K increases, the system’s BER also increases, reducing MIMO system performance. This is directly due to the fact that in the formula there is a division of channel M gain by the number of channels N; thus, a direct relationship was obtained between the fading coefficient and the BER of this system.

For the second case - correlated antenna elements of MIMO system - we get:

Computational analysis of the BER versus the standard deviation and the fading parameter K, using Eq. (19) describing BER via these parameters for various M and N correlated elements at the input and output multibeam correlated antenna elements of MIMO system, is shown in Figures 14 and 15.

For MIMO systems, it was found that the BER is infinitely high, which is due to the lack of multiplication in the number of channels N outside the logarithm in Eq. (15), so the graph cannot be displayed. However, we acquired a graph in which the displayed value of the BER was zero.

For this case, it can be seen that as the value of the fading coefficient K increases, the system’s BER also increases, reducing MIMO system performance. Compared to the first case (Figure 13), in the given formula, the number of channels N is multiplied by the channel gain M; thus, a direct relationship was obtained between the fading coefficient and the BER of this system.