OPEN RADIO ACCESS NETWORK (O-RAN)

Radio Access Network

A radio access network (RAN) is a critical component of a wireless telecommunications
system that uses a radio link to connect individual devices to the rest of the network. The RAN
connects user equipment, such as a cell phone, computer, or other remotely operated machine,
via a fiber or wireless backhaul connection. That link connects to the core network, which
handles subscriber information, location, and other functions. The RAN, also known as the
access network, is the radio component of a cellular network. A cellular network is made up of
land areas called cells. A cell is supplied by at least one radio transceiver, while three is the
common number for cell sites.

RAN components include base stations and antennas that cover a specific region, depending
on their capacity. Silicon chips in both the core network and the user equipment provide RAN
functionality. A RAN is made up of three essential elements:

• Antennas, which convert electrical signals into radio waves.

• Radios, which transform digital information into signals that can be sent wirelessly and
ensure that transmissions are in the correct frequency bands with the right power levels.
• Baseband units (BBUs) provide a set of signal processing functions that make wireless
communication possible. Traditional baseband uses custom electronics combined with
multiple lines of code to enable wireless communication, typically using the licensed
radio spectrum. BBU processing detects errors, secures the wireless signal and ensures
that wireless resources are used effectively.

A RAN provides access to and coordinates the management of resources across the radio
sites. A handset or other device is wirelessly connected to the backbone, or core network, and
the RAN sends its signal to various wireless endpoints so it can travel with traffic from other
networks. A single handset or phone could be connected at the same time to multiple RANs,
sometimes called dual-mode handsets. RAN trends include the following:

• Open RAN is the hot topic in the access network world. It involves developing
interoperable open hardware, software and interfaces for cellular wireless networks that
use white box servers and other standard equipment, rather than the custom-made
hardware typically used in base stations.
 • C-RAN separates the radio elements in a base station into remote radio heads (RRHs).
These can be used atop the cell towers for the most efficient radio coverage. RRHs must
be connected to centralized baseband controllers via fiber or microwave radio links.
Most baseband processing uses standard white box servers.
• Global System for Mobile communications (GSM) RAN, or GRAN, was developed for
2G.
• GSM EDGE RAN, or GERAN is like GRAN, but it specifies the inclusion of Enhanced
Data GSM Environment packet radio services.
• Universal Mobile Telecommunications System (UMTS) Terrestrial RAN, or UTRAN,
came about with 3G.
• Evolved Universal Terrestrial RAN, or E-UTRAN, is part of LTE.

Figure 1 Basic RAN architecture

Evolution of Radio Access Network

In the 1G cellular network, RAN was not considered an independent unit. The RAN in 1st
generation comprises of Base Stations and antennas. 1st generation RAN comprising BS and
antennas.
 Figure 2 RAN in 1G

The RAN in the 2G GSM cellular network is called GRAN and GERAN. 2G cellular
network is fully distributed, meaning everything is at the cell site. 2G cellular network defines
the term BSS, consisting of BTS and BSC. RAN in 2G cellular networks comprises a radio BS
known as BTS. The BTS is controlled by an entity known as BSC. BSC performs various
functions, such as RRM, mobility management, and data encryptions. 2G GSM networks were
initially designed to perform voice calls (by circuit-switching (CS) means); even the SMS part
came later. GRAN is the RAN that allows voice calls and SMS services. The data core network
in the 2G GSM network is MSC, that is, CS core. In 2.5G GPRS enhancement, the data part
was added to the voice part. The RAN that allows performing mobile data services is called
2.5G GERAN via GPRS and EDGE (by packet-switching (PS) means). The PS core was added
in 2.5G GPRS, comprising Gateway GPRS Support Node (GGSN) and Serving GPRS Support
Node (SSGN).

The RAN in the 3G UMTS cellular network is called UTRAN. UTRAN consists of two key
components: NodeB and RNC. The RNC is located in between the NodeB and core network.
The BSs in 3G cellular network are termed as NodeB (3GPP term for BS). NodeB comprises
Tx and Rx to communicate with the UEs within the cell. NodeB communicates with the UE
just like BTS in a GSM cellular network. NodeB is controlled by an entity known as RNC.
RNC in 3G cellular networks performs similar functions just like functions performed by BSC
in a GSM cellular network. In UTRAN, some centralized functions exist in the RNC or
centralized controller.
 Figure 3 RAN in 2G (Left) and RAN in 2.5G (Right)

Figure 4 RAN in 3G

The RAN in 4G LTE is called E-UTRAN. Like 2G, in 4G, everything is distributed at the
cell site. In contrast to GRAN and UTRAN, which uses a combination of BS and a controller
for RAN functions, E-UTRAN uses eNodeB (eNB) for all radio communication (nearly all
parts of RNC are incorporated into eNB). Such architecture is called D-RAN (as defined
earlier) architecture (or flat architecture). 4G LTE network only maintains the PS part (doesn’t
carry the CS part). In E-UTRAN, eNodeB is directly associated with 4G CN (MME and S-
GW). MME handles the signaling part, while S-GW and P-GW take the data part. The interface
between eNBs is the x2 interface, while the interface used between eNBs and MME/SGW is
the S1 interface.
 Figure 5 RAN in 4G

Open Radio Access Network (O-RAN)

An Open Radio Access Network (O-RAN) is a non-proprietary variant of the Radio Access
Network (RAN) architecture that enables interoperability between cellular network equipment
from different vendors. The future goal of O-RAN is for any hardware and software in the
cellular network to interoperate effortlessly and securely, regardless of vendor. Today, O-RAN
may have several vendors, however the radio unit (RU) software is provided by the same
business that makes the antennas, whereas the Distribution Unit (DU) and Centralized Unit
(CU) may come from a different company. In addition, as previously stated, RAN has
historically been a proprietary (closed) network section. Currently, it is the only remaining
section of the network that is primarily proprietary.

O-RAN is based on interoperable lower layer splits, but where best to perform this split is a
complex question requiring a compromise between RU simplification, support for advanced
multipoint radio frequency capabilities, and requirements of the front haul transport. Splits
have been assigned numbers to assist in comparing the many available options. The higher
number assignments represent splits that reside lower in the protocol stack, with less
functionality deployed below the RU split. Lower layer splits are centered in the protocol stack
below the Medium Access Control (MAC) layer. Options for this include a location between
the MAC and Physical (PHY) layers (Split 6), within the PHY layer (Split 7), and between the
PHY layer and RF functionality (Split 8). Still, there are currently multiple challenges in multi-
vendor deployment model such as: (1) added complexity to identify and isolate unwanted
performance issues in the network, herein, the role of the system integrator becomes vital for
managing and controlling the added troubleshooting complexity, and (2) security is a particular
area of concern as the infrastructures deployed by different vendors’ equipment could increase
threat surface areas. Therefore, there has been a coordinated global effort to improve O-RAN
architecture performance through virtualized network elements and open interfaces that
incorporate intelligence over RAN and to leverage emerging learning methods to employ
intelligence in every layer of the RAN architecture.

Gambar 6 Architecture O-RAN Vs Traditional RAN

The difference between O-RAN and Traditional RAN is that O-RAN is not tied to one
vendor but can use several vendors because it is an open interface so that the costs used are
cheaper compared to Traditional RAN (O-RAN Alliance, 2020). With the implementation
of O-RAN on 2G and 4G networks, a technical comparison was made between Open RAN
and Existing RAN. The comparison was carried out in several stages, namely calculating
the link budget, capacity dimensioning, and e-Node B throughput.

O-RAN in 5G
The fifth generation (5G) and beyond cellular networks are expected to support multiple
quality of service (QoS) classes that require a diverse and broad set of radio performance
targets, including broadband data rates, stringent radio latency, strict link reliability, advanced
processing, and computing power. However, existing single-vendor network deployments,
which use proprietary interfaces and equipment, severely limit cellular technology innovation,
development, and ability to support future key use cases. Thus, the O-RAN has offered new
interfaces and architectures based on openness and interoperability, allowing for programmable
data-driven control and intelligence in network deployments.
 O-RAN, as promoted by the O-RAN Alliance (one of the major open RAN standard
organizations), is built on the disaggregation of traditional RAN systems into radio unit (RU),
distributed unit (DU), and central unit (CU) components, as well as multiple hardware and
software platforms. O-RAN is a change of existing mobile networks; it involves a broad
ecosystem in the creation of RAN infrastructures, rather than typical vendor-proprietary
solutions. It enables operators to select diverse hardware and software components from
multiple sources without being limited to a single telecom vendor. We should emphasize that
O-RAN technology began as a movement that applies to all mobile technology generations (all
xGs), namely 2G, 3G, 4G, 5G, and all future Gs. On top of the 3GPP specifications, the O-
RAN Alliance specifies Open RAN internal interfaces between major RAN building pieces to
assure multi-vendor compatibility. While O-RAN technology is completely compliant with the
3GPP standards, it further expands the RAN capabilities into truly open and intelligent RAN
systems, offering the following important enablers: (Alavirad et al., 2023):

• Open standardized interfaces between RUs and baseband units (BUs), including the
element management system (EMS) of the radio/baseband, the network management
system, control and data planes, and the CU and the DU, respectively.
• Decoupling the deployed software from the hardware platforms executing it.
• Open hardware that offers platforms with general purpose processors and accelerators.
• Open software that is commercially viable to meet all the high performing critical
performance requirements to support real-time system specifications.

Advantages of O-RAN
ORAN develops a unified architecture by disaggregating hardware and software, which
brings various benefits (such as low latency and network slicing). In addition to allowing
network automation, O-RAN has various benefits, such as (Singh et al., 2020):

1) Agility: The unification of the software enabled architecture makes the network suitable for
existing/past and future generation.
2) Deployment Flexibility: Dis-aggregation and software association make the network
flexible for installation and extension.
3) Real time responsiveness: O-RAN is the software driven service specific network which
behaves based on intended service and thus prefers the real time services which require very
low latency over the less critical services.
4) Operating Cost Reduction: It is estimated that the plug and play feature of O-RAN and
modern learning methods may reduce the maintenance cost up to 80%. Putting the software
at the heart of the network, the operators can unify the connectivity gains of all the
generations under the same umbrella. Doing this, the operators can save millions of dollars.

Technical Advancements and Opportunities of O-RAN

Accelerating 5G growth and O-RAN adoption will lead to breakthroughs in modern
technology. Applications for IoT devices, machine learning, and mobile edge computing are
projected to reach their peak in the next years. Key technical breakthroughs and prospects are
outlined below (Singh et al., 2020):

1. IoT-enabled Network
The deployment of O-RAN will remove the limitations on IoT connectivity. It would offer
a flexible design ideal for current IoT networking. More things should be related to various
applications, such as health care, retail, security, and so on. New applications may now be
added (for example, digital locks, e-health). These apps would connect IoT devices via the O-
RAN architecture. Furthermore, disaggregation of RIC non-RT allows for enormous IoT
connectivity to devices with low throughput but high coverage and low battery consumption.
2. MEC Computing

MEC is a type of architecture that allows edge devices to perform computational activities.
Because of the significant increase in the number of connected devices, the next generation
RAN must be capable of intelligently managing traffic across long distances. MEC is thought
to be a major technique for this. Currently, most programs store their material and perform
online computations on distant sensors that are typically located far from the end user. MEC
will make those processes more accessible to the end user. This change will help to alleviate
congestion on the mobile network and cloud computer. In addition to minimizing congestion,
MEC will play an important role in lowering the latency of the 5G network. Bringing the data
closer to the end unit and streaming it more directly to the end device can achieve extremely
low latency, allowing it to handle applications that require high-speed data and compute.

3. Inclusion of Modern Learning Methods

Machine Learning allows a machine to learn without an explicit program. ML is

distinguished by its ability to learn relevant information from input data sets, making it ideal
for applications with dynamic processing environments. Specifically, ML can improve the
wireless framework in the following ways:

• ML based resource management, mobility and networking algorithms can significantly

adapt the dynamic environment.
• ML is considered as the key to realize the goals of self-realizable network.

According to a survey, several machine learning algorithms have already been developed for
future wireless networks, such as Q-learning for resource allocation and interference
coordination. In addition, Bayesian learning, or channel estimation are used in MIMO
networks.
 Daftar Pustaka
https://www.cisco.com/c/en/us/solutions/what-is-open-ran.html

https://www.techtarget.com/searchnetworking/definition/radio-access-network-RAN

https://www.lumenci.com/research-articles/evolution-of-ran-the-road-from-1g-to-5g

Alavirad, M., Hashmi, U. S., Mansour, M., Esswie, A., Atawia, R., Poitau, G., & Repeta, M.
(2023). O-RAN architecture, interfaces, and standardization: Study and application to
user intelligent admission control. Frontiers in Communications and Networks, 4.
https://doi.org/10.3389/frcmn.2023.1127039

Singh, S. K., Singh, R., & Kumbhani, B. (2020). The Evolution of Radio Access Network
Towards Open-RAN: Challenges and Opportunities. 2020 IEEE Wireless
Communications and Networking Conference Workshops, WCNCW 2020 - Proceedings.
https://doi.org/10.1109/WCNCW48565.2020.9124820