What does a 5G base station look like? What is the difference between it and 4G?

This is what ordinary people think of 4G and 5G base stations...

This is the 4G and 5G base stations in the eyes of communications professionals...

So what is the difference between 4G and 5G base stations?

Let’s first understand the composition of a base station site.

A base station site includes base station equipment and supporting equipment. The base station equipment includes baseband unit, wireless radio unit and antenna; the supporting equipment includes transmission equipment, power supply, backup battery, air conditioner, monitoring system and tower (pole).

Base station equipment is responsible for connecting mobile phones via radio waves and connecting to the core network and the Internet through transmission equipment, while power supplies, backup batteries, air conditioning and monitoring systems are responsible for ensuring the stable operation of base station equipment.

Like 4G sites, 5G sites also require supporting equipment. Usually, 4G and 5G base stations are co-located, that is, 5G equipment is superimposed on the original 4G site. Since the power consumption and transmission capacity of the base station equipment increase after the superposition of 5G equipment, the site supporting equipment needs to be upgraded and expanded accordingly.

But there are differences between 4G base station equipment and 5G base station equipment.

As shown in the figure above, 4G base station equipment consists of BBU (baseband unit) and RRU (radio remote unit). RRU is usually moved closer to the antenna. BBU and RRU are connected by optical fiber, and RRU and antenna are connected by feeder.

The 5G base station equipment divides the BBU into CU (central unit) and DU (distributed unit), and connects it to the AAU (active antenna unit) through optical fiber. The AAU includes the RRU and antenna functions, that is, the active RF part and the passive antenna are integrated.

To understand the specific differences, we have to start with the RAN (radio access network) protocol stack.

Let's first take a brief look at the functions of each layer of the protocol stack:

• RRC, Radio Resource Control layer, is responsible for connection configuration, policy-related signaling or control plane, and is not responsible for processing data packets on the user plane.

• PDCP, Packet Data Convergence Protocol layer, is responsible for compressing and decompressing IP headers, encryption and integrity protection of data packets.

In the dual-connection mode of NSA networking, the PDCP layer is also responsible for data diversion and aggregation between 4G base stations and 5G base stations. At the same time, in the deployment of 5G private networks, in order to prevent data from leaving the campus to protect the security of local data, the PDCP layer is also a key node for isolating and forwarding public network data flows and private network data flows, and realizing local data flow unloading.

• RLC, Radio Link Control layer, is responsible for segmentation/reassembly of data packets, ARQ error correction, duplicate packet detection, etc.

• MAC, media access control layer, is responsible for real-time resource scheduling decisions, multiplexing/demultiplexing, buffering and other functions.

The MAC layer is also responsible for carrier aggregation scheduling. Since wireless resources need to be scheduled in real time, the MAC layer has extremely high latency requirements.

• PHY, physical layer, is responsible for coding, modulation, FEC, etc.

After the above processing, the data is transmitted to the radio frequency unit and converted into an analog high-frequency signal, and then transmitted to the mobile phone via a wireless carrier.

As shown in the figure above, the 4G base station consists of a BBU and an RRU, in which the functions of the RRC, PDCP, RLC, MAC and PHY layers are integrated in the BBU.

However, in the 5G era, considering the trend of RAN virtualization, cloudification and centralization, and in order to reduce fronthaul capacity and latency, 5G base stations have been reconstructed and mainly split into three parts:

• CU, central unit, mainly includes RRC, SDAP and PDCP protocol layers, and is mainly responsible for non-real-time RRC and PDCP protocol stack functions.

CU can be deployed in a cloud-based manner, supporting the converged deployment of core network UPF sinking and edge computing. CU and DU are connected through the F1 interface. One CU can manage one or more DUs.

• DU, distributed unit, mainly includes nodes of RLC, MAC and PHY layers, and is mainly responsible for processing MAC layer functions and some physical layer functions with real-time requirements.

One DU can support one or more cells. Since the MAC layer is responsible for real-time scheduling of wireless resources and has extremely high latency requirements, the DU needs to be deployed close to the AAU (within 1ms). A typical deployment method is to deploy DU and AAU at the same site. It can also be used in scenarios such as campuses, factories, and shopping malls. One DU can connect to multiple distributed AAUs.

After this split, the fronthaul and backhaul of the 4G wireless access network are also split into three parts: fronthaul, midhaul and backhaul. The fronthaul is between the AAU and DU, the midhaul is between the DU and CU, and the backhaul is from the CU to the core network.

After talking about the baseband part of the base station, let’s talk about the RF part of the base station.

Why in the 5G era should we evolve from RRU+antenna to AAU with active RF part and antenna integrated?

The main reason is that 5G uses Massive MIMO technology.

Massive MIMO has two main technical advantages:

1) Improve coverage and reduce interference through beamforming

Beamforming is the process of adjusting the amplitude and phase of multiple antennas to give the antenna radiation pattern a specific shape and direction, so that the wireless signal energy is concentrated on a narrower beam, thereby enhancing coverage and reducing interference.

With beamforming, an accurate user-level ultra-narrow beam can be formed and move with the user's location, directing energy to the user's location. Compared with traditional wide-beam antennas, it can improve signal coverage while reducing user interference between cells.

At the same time, 3D beamforming can also add a usable dimension in the vertical dimension, so that the vertical coverage range of the cell can be adjusted more flexibly, changing the traditional two-dimensional wireless design method.

2) Improving cell capacity through spatial multiplexing

Massive MIMO can send multiple spatially multiplexed data streams to multiple users simultaneously through MU-MIMO, thereby doubling the cell capacity.

If wireless networks are likened to highways, this is equivalent to adding more roads without increasing the spectrum bandwidth.

But the problem is that to achieve Massive MIMO, multiple antennas are a prerequisite. The performance potential of beamforming technology will increase with the increase in the number of antennas. For this reason, 5G Massive MIMO uses dozens or even hundreds of antenna units.

It is precisely because Massive MIMO uses so many antennas that it requires an AAU device based on an integrated RF unit and antenna unit.

To understand this problem, we must start with the basic composition principles of base station radio frequency units and antennas.

Let’s first look at the RRU+antenna mode in the 4G era (as shown below). The RRU is mainly responsible for the transmit/receive signal processing from baseband to air interface, completing the conversion of digital signals and RF signals. It mainly includes digital systems, RF transceiver systems (TRX), power amplifiers, filters, etc., and then connects to the antenna through a feeder.

The 5G AAU integrates multiple antenna units and radio frequency units into one, and its general structure is as follows...

Imagine what would happen if the AAU did not integrate multiple antenna units and RF units into one unit?

This brings up the following questions:

1) Connecting so many feed lines between multiple RF and antenna units is an impossible task.

Looking back at the 4G era, as MIMO continues to upgrade, the number of feeder channels connecting RRUs to antennas has increased, and a large number of beards have grown on the towers. How many feeder channels are needed to connect more than a hundred antenna units in Massive MIMO?

2) There is not enough space on the tower to support so many feeder connections, and RRUs also take up space.

3) Engineering installation and maintenance will become increasingly complex.

4) Feeder lines will increase RF loss and affect signal coverage.

Obviously, supporting Massive MIMO requires the RF unit and antenna unit to be based on an integrated AAU.

Another key reason is that only when the RF unit and the antenna unit are integrated can more fine-grained digital control of multiple antenna units be performed to achieve beamforming.

These are the differences between 4G base stations and 5G base stations.