12.1. Deployment options

Like in previous generations, the 3GPP standard organization defines a 5G core network (5GC) and a 5G radio access network (5G NR).

Unlike previous generations which simultaneously deployed the core network and the radio access network, the fifth generation allows the integration of fourth generation (4G) elements in different configurations:

• – standalone (SA) configuration: the connection of the radio access network to the core network is set up for the user plane and the control plane;
• – non-standalone (NSA) configuration: the connection of the radio access network to the core network is set up only for the user plane.

Option 2 corresponds to an SA configuration, for which the 5G NR connects to the 5GC.

Option 5 corresponds to an SA configuration, for which the evolved universal terrestrial radio access network (E-UTRAN) connects to the 5GC.

The NSA configuration implements the dual connectivity (DC) architecture, which has the following characteristics:

• – the master radio access network connects to the core network for the control plane and the user plane;
• – the secondary radio access network connects to the core network only for the user plane;
• – the master radio access network controls the secondary radio access network.

Option 3 corresponds to the NSA configuration EN-DC (E-UTRA-NR DC) for which the E-UTRAN is the master radio access network and the 5G NR is the secondary radio access network. The connection is done on the evolved packet core (EPC).

Option 4 corresponds to the NSA configuration NE-DC (NR-E-UTRA DC) for which the 5G NR is the master radio access network and the E-UTRAN is the secondary radio access network. The connection is done on the 5GC.

Option 7 corresponds to the NSA configuration NGEN-DC (NG-RAN E-UTRA-NR DC) for which the E-UTRAN is the master radio access network and the 5G NR is the secondary radio access network. The connection is done on the 5GC.

Figure 12.1 describes the five deployment options that implement both types of configurations.

There are several ways to start the deployment of the 5G network. The main advantage of option 3 is that it only requires the deployment of the 5G NR.

Since the 5G NR will increase the existing capacity of the E-UTRAN, option 3 allows for flexible deployment when this capacity increase is needed.

More specifically, the possibility of deploying the 5G NR offers the possibility of using the spectrum above 6 GHz, whose wide frequency bands make it possible to deliver large data rates.

12.2. Functional architecture

The 5G NR consists of a single type of entity, the en-gNB (E-UTRA-NR DC next generation Node B) station (Figure 12.2).

For the control plane, the en-gNB entity connects to the evolved node base station (eNB) via the X2-C interface.

For the user plane, the en-gNB entity can connect to the core network via the S1-U interface or to the eNB entity via the X2-U interface:

• – for option 3, the eNB entity connects to the core network via the S1-U interface, and the en-gNB entity connects to the eNB entity via the X2-U interface;
• – for option 3a, the eNB and en-gNB entities connect to the core network via the S1-U interface;
• – for option 3x, the en-gNB entity connects to the backbone via the S1-U interface, and the eNB entity connects to the en-gNB entity via the X2-U interface.

The en-gNB entity is divided into two types of modules (Figure 12.3):

• – a centralized unit (en-gNB-CU);
• – one or more distributed units (en-gNB-DU).

The en-gNB-CU is connected to the eNB entity via the X2-C and X2-U interfaces and to the EPC via the S1-U interface.

Each en-gNB-DU is connected to the en-gNB-CU via the F1 interface, including the F1-C interface for the control plane and the F1-U interface for the user plane.

12.3. Protocol architecture

12.3.1. Radio interface

The protocol architecture of the new radio (NR) interface has similarities to the LTE (Long-Term Evolution) radio interface described in Figure 1.4 of Chapter 1. This interface supports RRC (Radio Resource Control) signaling transmitted in the signaling radio bearer (SRB) and IP (Internet Protocol) packets transmitted in the data radio bearer (DRB) (Figure 12.4).

The NR interface introduces a new service data adaptation protocol (SDAP) that encapsulates IP packets. This protocol contains the QoS flow identifier (QFI), which indicates the level of quality of service applied to the flow.

For the LTE interface, the physical control format indicator channel (PCFICH) contains the number of OFDM (Orthogonal Frequency-Division Multiplexing) symbols used for the physical downlink control channel (PDCCH).

For the NR interface, this physical channel is deleted, the indication of the size of the PDCCH being provided by an RRC message.

For the LTE interface, the physical HARQ indicator channel (PHICH) contains the indication of a positive or negative acknowledgment (ACK/NACK) of the signal received in the physical uplink shared channel (PUSCH).

For the NR interface, this physical channel is deleted, the ACK/NACK being provided either semi-statically in an RRC message or dynamically in the downlink control information (DCI) carried in the PDCCH.

12.3.1.1. Control plane

The RRC protocol exists independently at the level of the two entities, the eNB master station with the 4G RRC protocol and the en-gNB secondary station with the 5G RRC 5G. An RRC connection is established independently between the mobile and the master and secondary stations.

When establishing the initial connection, the SRB1 of the eNB entity uses the 4G packet data convergence protocol (PDCP).

After establishing the initial connection, SRB1 and SRB2 may be configured to use either the 4G or 5G PDCP.

5G RRC messages generated by the en-gNB entity can then be transmitted via the eNB entity. During the initial configuration, the eNB entity sends the 5G RRC message of the en-gNB entity via SRB1 using the 5G PDCP. The following reconfigurations can be transmitted via the entity eNB or directly to the mobile via SRB3 on the NR interface.

The shared SRB is supported, which allows the duplication of 4G RRC messages generated by the master station, via the LTE-Uu interface and via the secondary station. The shared SRB support uses the 5G PDCP.

12.3.1.2. User plane

There are three types of DRB, corresponding to the master cell group (MCG), the secondary cell group (SCG) and the split bearer:

• – MCG bearers are transmitted on the LTE interface;
• – SCG bearers are transmitted on the NR interface;
• – split bearers are transmitted on both the LTE and NR interfaces.

From the mobile point of view, the bearers use the following protocols (see Figure 12.5):

• – 4G or 5G PDCP, 4G RLC and 4G MAC for MCG bearers;
• – 5G PDCP, 5G RLC and 5G MAC for SCG bearers;
• – 5G PDCP, 4G RLC and 4G MAC or 5G PDCP, 5G RLC and 5G MAC for split bearers.

From the point of view of eNB and en-gNB entities, each bearer type, MCG, SCG or split bearers, can be terminated at either the eNB entity or the en-gNB entity (Figure 12.6).

MCG bearers use 4G or 5G PDCP. Other bearers consistently use 5G PDCP.

MCG bearers use 4G RLC. SCG bearers use 5G RLC. Split bearers use 4G and 5G RLC.

MCG bearers use 4G MAC. SCG bearers use the 5G MAC 5G protocol. Split bearers use 4G and 5G MAC.

12.3.2. F1 interface

Several options define the functional split between the en-gNB-CU and en-gNB-DU modules (Figure 12.7). The two main criteria that determine the choice of option are interface throughput and latency.

For option 1, the RRC and SDAP layers are hosted in the en-gNB-CU while the PDCP, RLC (Radio Link Control) and MAC (Medium Access Control) layers, as well as the physical layer (PHY) and the RF (Radio Frequency) module, are located in the en-gNB-DU.

Option 2 has similarities with the X2 interface for the user plane, and differences for the control plane since new procedures may be required.

For option 3, the split of the RLC layer is based on either the ARQ function or the direction of transmission:

• – option 3.1: the lower part contains the segmentation functions, while the higher part contains the ARQ function as well as other functions of the RLC layer;
• – option 3.2: the lower part is composed of the transmission functions, while the reception functions are assigned to the higher part.

For option 4, the RLC and higher layers are located in the en-gNB-CU while the MAC layer as well as the physical layer and the RF module are hosted in the en-gNB-DU.

For option 5, the higher MAC sub-layer hosted in the en-gNB-CU includes centralized scheduling and is responsible for controlling several lower MAC sub-layers, while the lower MAC sub-layer hosted in the en-gNB-DU includes functions with delay requirements such as the HARQ mechanism or random access control.

For option 6, the MAC and higher layers are hosted in the en-gNB-CU, while the physical layer and the RF module are hosted in the en-gNB-DU. The F1 interface must convey the configuration functions of the physical layer, defined by the MAC layer.

Several split configurations of the physical layer are defined. The description of the physical layer is given in Chapter 1, in Figure 1.5 for the downstream, and in Figure 1.6 for the upstream.

For option 7.1, the lower part hosted in the en-gNB-DU includes the inverse fast Fourier transform (IFFT) for both downlink and uplink, while the higher part hosted in the en-gNB-CU contains other functions.

For option 7.2, the lower part hosted in the en-gNB-DU adds the mapping function to the resource elements for both downlink and uplink.

For option 7.3, the en-gNB-CU hosts, only for the downstream, the error detection and correction code as well as the rate adaptation.

Option 8 separates the physical layer and the RF module. This split makes it possible to centralize all the processes in the en-gNB-CU entity.

In addition, the enhanced common public radio interface (eCPRI) is the subject of industrial cooperation aimed at defining the specifications available for the interface between the CU and DU modules of a radio station.

Figure 12.7 provides the correspondence between the options of the F1 interface and CPRI. Options 7.2 and 7.3 of the F1 interface also have their concordance with CPRI.

12.4. Procedures

12.4.1. Adding a secondary node

The procedure for adding a secondary node (en-gNB entity) is initialized by the eNB entity (master node). It is used to establish a context at the en-gNB entity to provide radio resources to the mobile.

For SCG bearers requiring radio resources, this procedure adds at least the first primary cell (PCell) (Figure 12.8).

1) The eNB entity sends to the en-gNB entity the message X2-AP SgNB Addition Request in order to allocate radio resources for the establishment of an S1-U bearer with the serving gateway (SGW).

2) If the request is accepted, the en-gNB entity determines the primary cell and possibly the secondary cells (SCell) and provides the new radio resource configuration to the eNB entity in an NR RRC configuration contained in the message X2-AP SgNB Addition Request Acknowledge.

3) The eNB entity transmits to the mobile the message RRC ConnectionReconfiguration including an NR RRC configuration message.

4) The mobile applies the new configuration and responds to the eNB entity with the message RRC ConnectionReconfigurationComplete, including, if necessary, an NR RRC response message.

5) The eNB entity informs the en-gNB entity of whether the mobile has successfully completed the reconfiguration procedure or not, via the message X2-AP SgNB Reconfiguration Complete, including an NR RRC response message, if it has been received from the mobile.

6) The mobile synchronizes on the PCell of the en-gNB entity and performs the random access procedure.

7) In the case of DC option 3a, to establish the SI-U bearer between en-gNB and SGW entities, the eNB entity transmits the message S1-AP E-RAB Modification Indication message to the mobility management entity (MME).

8) and 9) The establishment of the SI-U bearer continues with the exchange of GTPv2-C messages between the MME and SGW entities.

10) The MME entity acknowledges the request of the eNB entity by sending the message S1-AP E-RAB Change Confirmation.

11) The S1-U bearer between the SGW and en-gNB entities is established.

12.4.2. Changing a secondary node

The procedure for changing a secondary node is initialized either by the eNB entity (Figure 12.9) or by the source en-gNB entity. It is used to transfer the mobile context of the source en-gNB entity to the target en-gNB entity.

1) and 2) The procedure for changing the source en-gNB entity starts with exchanges of the messages X2-AP SgNB Addition Request and SgNB Addition Request Acknowledge between the eNB and the target en-gNB entities.

3) and 4) If the resource allocation from the target en-gNB entity has been successful, the eNB entity triggers the resource release to the source en-gNB entity with the exchange of the messages X2-AP SgNB Release Request and SgNB Release Request Acknowledge.

5) and 6) The eNB entity triggers the new configuration at the mobile level in the message RRC ConnectionReconfiguration, containing an NR RRC configuration message generated by the target en-gNB entity. The mobile applies the new configuration and sends the message RRC ConnectionReconfigurationComplete.

7) The eNB entity informs the target en-gNB entity via the message X2-AP SgNB Reconfiguration Complete that the reconfiguration procedure is successful.

8) The mobile synchronizes on the PCell of the target en-gNB entity and performs the random access procedure.

9) and 10) If the bearer uses the RLC with the acknowledgment mode, the source en-gNB entity sends the PDCP sequence number in the message X2-AP SN Status Transfer, which the eNB entity transfers to the target en-gNB entity.

11) If applicable, the data transfer for the downstream takes place between the en-gNB and SGW entities. It can be initialized as soon as the source en-gNB entity receives the message X2-AP SgNB Release Request from the eNB entity.

12) In the case of DC option 3a, to start the establishment of the S1-U bearer between the target en-gNB and SGW entities, the eNB entity transmits the message S1-AP E-RAB Modification Indication to the MME entity.

13) and 14) The S1-U bearer establishment continues by the exchange of GTPv2-C messages between the MME and SGW entities.

15) The MME entity acknowledges the request of the eNB entity by sending the message S1-AP E-RAB Change Confirmation.

16) The SGW entity uses the control message GTP-U End Marker Packet to notify the target en-gNB entity that traffic for the downstream will be forwarded directly.

17) Upon receipt of the message X2-AP UE Context Release, the source en-gNB entity may release the resources allocated to the mobile.

12.4.3. Removing a secondary node

The procedure for removing a secondary node is initialized either by the eNB entity (Figure 12.10) or by the en-gNB entity. It is used to remove the mobile context at the en-gNB entity.

1) The eNB entity initiates the procedure for removing the secondary node by sending to the en-gNB entity the message X2-AP SgNB Release Request. Data transfer to the eNB entity may be requested.

2) The en-gNB entity confirms the receipt by sending the message X2-AP SgNB Release Request Acknowledge.

3) and 4) In the message RRC ConnectionReconfiguration, the eNB entity indicates to the mobile that it must release the entire SCG bearer configuration. The mobile acknowledges the received message in the response message RRC ConnectionReconfigurationComplete.

5) In the case of data transfer, if the released bearer uses the acknowledgment mode for RLC, the en-gNB entity sends the message X2-AP SN Status Transfer.

6) The downlink data transfer to the eNB entity is made via the en-gNB entity.

7) In the case of DC option 3a, the S1-U bearer establishment procedure between the eNB and SGW entities takes place.

8) Upon receipt of the message X2-AP UE Context Release, the en-gNB entity may release the resources allocated to the mobile.

12.5. Transmission chain

The transmission chain of the NR interface has an identical structure to that of the LTE interface, described in Chapter 1, in Figure 1.5 for the downlink and in Figure 1.6 for the uplink.

The following sections describe the major differences in the technical characteristics of LTE and NR interfaces.

12.5.1. Frequency bands

Two frequency ranges (FR) are defined:

• – the frequency range FR1 covers the frequency bands between 450 MHz and 6 GHz. The radio channel bandwidth is between 5 and 100 MHz;
• – the frequency range FR2 covers the frequency bands between 24.250 GHz and 52.600 GHz. The radio channel bandwidth is between 50 and 400 MHz.

Several transmission modes are supported for the frequency range FR1:

• – frequency-division duplex (FDD): uplink and downlink use a specific frequency band;

• – time-division duplex (TDD): uplink and downlink temporally share the same frequency band;
• – supplementary uplink (SUL): an additional frequency band is allocated only for uplink;
• – supplementary downlink (SDL): an additional frequency band is allocated only for downlink.

The frequency range FR2 only supports the TDD mode.

The bandwidth of receiving and transmitting a mobile should not necessarily be as wide as the radio channel bandwidth and can be adjusted.

For initial access and until the mobile configuration is received, the mobile uses a bandwidth part (BWP) of the radio channel. The mobile can be configured with multiple bandwidth portions, only one of which can be active for a radio channel.

12.5.2. Waveform

The OFDM (Orthogonal Frequency-Division Multiplexing) signal using the cyclic prefix is the downlink waveform. The DFT-S-OFDM (Discrete Fourier Transform Spread OFDM) signal using a cyclic prefix is the uplink waveform, but the DFT-S precoding function can be disabled.

For the LTE interface, the spacing between the sub-carriers has a single value equal to 15 kHz. For the NR interface, the spacing between the sub-carriers Δf takes several values according to the parameter μ (Table 12.1):

μ = {0, 1, 3, 4} for the primary synchronization signal (PSS) and the secondary synchronization signal (SSS), as well as for the physical broadcast channel (PBCH);

μ = {0, 1, 2, 3} for the physical downlink shared channel (PDSCH), the PUSCH, the PDCCH and the physical uplink control channel (PUCCH).

The spacing between the sub-carriers is correlated with the frequency range (Table 12.1):

• μ = {0, 1}: these values are reserved for the frequency range FR1;
• μ = {2}: this value is used for the frequency ranges FR1 and FR2;
• μ = {3, 4}: these values are reserved for the frequency range FR2.

The normal cyclic prefix is supported for all spacing values between sub-carriers. The extended cyclic prefix is supported only for the value of μ = 2 (Table 12.1).

The physical resource block (PRB) is made up of 12 consecutive sub-carriers. The minimum and maximum number of blocks in a radio channel depends on the spacing between the sub-carriers (Table 12.2), which determines the minimum and maximum bandwidth of the radio channel (Table 12.3).

Random access preamble sequences of two different lengths are supported by the physical random access channel (PRACH).

The long sequence length is applied with sub-carrier spacing of 1.25 kHz and 5 kHz. The short sequence length is applied with sub-carrier spacing of 15, 30, 60 and 120 kHz.

12.5.3. Time frame

Downlink and uplink are organized in frames which duration is equal to 10 ms duration, each frame consisting of ten sub-frames of 1 msec. Each frame is divided into two half-frames of size equal to five sub-frames:

• – half-frame 0 is composed of sub-frames 0 to 4;
• – half-frame 1 is composed of sub-frames 5 to 9.

For the LTE interface, the sub-frame is composed of two time slots. For the NR interface, the number of time slots in the sub-frame depends on the spacing between the sub-carriers (Table 12.4).

For the LTE interface, the value of the transmission time interval (TTI) corresponds to the duration of the sub-frame and has a fixed value equal to 1 ms. For the NR interface, the value of the TTI corresponds to the duration of the time slot and has a value that depends on the spacing between the sub-carriers (Table 12.4).

For the LTE interface, the time slot includes seven OFDM symbols for the normal cyclic prefix and six OFDM symbols for the extended cyclic prefix.

For the NR interface, the time slot comprises 14 OFDM symbols for the normal cyclic prefix and 12 OFDM symbols for the extended cyclic prefix.

For the TDD mode, the elementary resource assigned to the downlink or uplink is different for the LTE and NR interfaces:

• – for the LTE interface, the elementary resource corresponds to a sub-frame and seven configurations are defined;
• – for the NR interface, the elementary resource corresponds to an OFDM symbol and 62 configurations are defined.

12.5.4. Error correction codes

The PDSCH and PUSCH use turbo code for the LTE interface and low-density parity check (LDPC) for the NR interface.

The PDCCH and PUCCH use tail-biting convolutional coding (TBCC) for the LTE interface and polar code for the NR interface.

12.5.5. Reference signals

For the LTE interface, the cell-specific reference signal (CRS) is used to demodulate the PBCH, PDSCH and PDCCH.

For the NR interface, the CRS is suppressed and the demodulation uses the demodulation reference signal (DMRS) dedicated to the physical channel. The number of DMRS symbols and the mapping to the resource elements are configured by the en-gNB entity.

The phase tracking reference signal (PTRS) may be transmitted over additional symbols to reduce the phase noise of the receiver oscillator when the frequency range FR2 is used.

12.5.6. PSS, SSS and PBCH

For the LTE interface and FDD mode, the PSS occupies the last symbol of time slots 0 and 10 on the time domain, and 62 sub-carriers on the frequency domain. The PSS is transmitted with a periodicity of 5 ms (Figure 12.11).

The SSS occupies the penultimate symbol of time slots 0 and 10 on the time domain, and 62 sub-carriers on the frequency domain. The PSS is transmitted with a periodicity of 5 ms (Figure 12.11).

The PBCH occupies the first four symbols of time slot 1 on the time domain, and 72 sub-carriers on the frequency domain. The PBCH is transmitted with a periodicity of 10 ms (Figure 12.11).

For the TDD mode, the PSS occupies the third symbol of time slots 2 and 12 on the time domain, and 62 sub-carriers on the frequency domain. The SSS occupies the last symbol of time slots 1 and 11 on the time domain, and 62 sub-carriers on the frequency domain.

For the NR interface, the PSS, SSS and PBCH form a block of four symbols. The PSS and SSS respectively occupy the first and third symbols of the block on the time domain and 127 sub-carriers on the frequency domain (Figure 12.12).

The PBCH occupies the last three symbols of the block on the time domain and, on the frequency domain, 240 sub-carriers for the second and the fourth symbol and 96 sub-carriers for the third symbol (Figure 12.12).

The blocks are transmitted in the first half-frame with a period of 20 ms. The maximum number of blocks in a half-frame is determined from the radio channel frequency:

• – four blocks for frequencies less than or equal to 3GHz;
• – eight blocks for frequencies between 3GHz and 6 GHz;
• – 64 blocks for frequencies in FR2.

Each block is transmitted by a specific beam, radiated in a certain direction.

By way of example, Figure 12.13 describes the position of the four blocks for a spacing of 15 kHz and a frequency of less than or equal to 3GHz, the symbols being numbered from 0 to 69 for the first half-frame.

Table 12.5 shows the number of the first symbol of the block for the different values relating to the spacing between the sub-carriers and the radio channel frequency.