5G Timing and Synchronization Handbook For TDD Deployment: VIAVI Solutions
5G Timing and Synchronization Handbook For TDD Deployment: VIAVI Solutions
Application Note
5G Timing and
Synchronization
Handbook for
TDD Deployment
Synchronization is one of the most critical functions of a communication system; however, in the context of 5G,
especially for Time Division Duplex (TDD) where both uplink and downlink transmission is on the same frequency,
the possibility of interference is much more significant. As a result, we see more stringent requirements for timing
and synchronization for both TDD LTE and 5G-NR. In this paper we will discuss the relationship of TDD, timing and
synchronization, and frame synchronization especially for 5G TDD deployments.
Table of Contents
With 100 MHz of bandwidth, C-Band can truly enable the enhanced mobile broadband (eMBB) use case for
5G. One thing to note is that C-band offers only Time Division Duplexing (TDD). TDD delivers a full-duplex
communication channel over a half-duplex communication link. This means both the transmitter and receiver
use the same frequency but transmit and receive traffic at different times by using synchronized time intervals.
Advances in digital signal processing and computation speed of hardware allows for TDD operations, but it does
offer some challenges. Let’s review the benefits of TDD and some of the timing and synchronization requirements
to ensure it can deliver the similar quality of RF services as Frequency Division Duplexing (FDD).
TDD turns out to be a more attractive option from spectral efficiency point of view because it requires only an
unpaired spectrum for operation which is beneficial considering the scarcity of frequency resources. Also, physical
layer features such as massive MIMO, beamforming, and precoding, that rely on channel state information (CSI)
measurement in the uplink, are more robust due to channel reciprocity.
While it brings spectral efficiency, TDD introduces a critical challenge: Timing and synchronization. Stringent timing
restrictions are imposed on a TDD system to avoid interference as both downlink (DL) and uplink UL share the
same spectrum.
1 Radio 10ms
Frame=10ms
1 Radio Frame=
1ms 1ms 1ms 1ms 1ms 1ms 1ms 1ms 1ms 1ms
10 Sub Frames
No of Slots Per
Sub Frame
Variable, Dependent on the numerology
1ms
0.5ms x2 0.25ms x4
Symbols Per
Vaiable, Depends on the numerology
Sub Frame
As subcarrier spacing changes so does the number of slots and symbols per subframes. For example, 15KHz has one
subframe of 1ms duration which is equal to one slot carrying 14 symbols. For 30KHz subcarrier spacing, one subframe
is equal to 2 slots of 0.5ms duration each and 28 symbols and so on (for normal cyclic prefix). For different type
of services, ultra-reliable low latency communication (URLLC) versus eMBB, for example, the service provider may
decide to use different slot and frame configuration. Release 15 version of 3GPP 38.213 has defined 56 slot formats
(Table 2) each of which is a predefined pattern of downlink/flexible/uplink symbols during one slot. The following
table provides a quick reference.
Slot
Format -
54
D D D D D F U D D D D D F U
gNB 1
Slot
Format -
55
D D F U U U U U U F D D U U
gNB 2
Intra-cell Interference
Interference caused within the same cell due to large timing inaccuracies. The probability of intra-cell interference is
low because in a TDD cell, different users are scheduled on different slots by the scheduler.
Inter-cell Interference
When users in the adjacent cells are scheduled on the same subcarriers but with different DL/UL slots, inter-cell
interference is a possibility, particularly if cells are not synchronized to a common clock. Figure 3 reveals the four
possible scenarios as mentioned in the paper Interference analysis and performance evaluation for LTE TDD system.
Type 1: Type 2:
Downlink UE receives interference BS receives interference from
from adjacent BS adjacent cell UE
Type 3: Type 4:
Downlink UE receives interference BS gets interference from
from adjacent cell UE adjacent BS
Transmission
Interference
In the Type 1 scenario, cell 1 is assigned on a DL time slot and the adjacent cell 2 is also assigned on DL time slot at the same time. In this scenario, both the UEs
on the cell edge receive interference from the neighbor cells.
In the Type 2 scenario, which is the inverse of Type 1, both cell 1 and cell 2 is assigned an UL time slot. This results in the reception of weak interference at the
cells from the adjacent neighbor UEs. Remember, UE power is limited compared to the gNB.
In the Type 3 scenario, cell 1 is assigned DL time slot and cell 2 is assigned UL time slot. The cell edge UE in cell 1 experiences strong interference from the cell
edge UE in cell 2. This is the most serious type of interference of all the cases.
In Type 4 scenario, cell 1 is assigned on UL time slot and cell 2 is assigned DL time slot. Cell 1 experiences interference from cell 2. However, the strength of
interference is relatively low as the path loss between the cells are high due to the large separation between them.
In general, to avoid such use cases of interference all base stations in a network should be synchronized with a
common phase clock reference (e.g. UTC - Coordinated Universal Time). Per ITU-T standards recommendation, both
5G-NR TDD and LTE-TDD networks need to be phase synchronized in order to limit the end-to-end time error to
under 1.5μs. This 1.5μs comprises of 1.1μs absolute time error up to the access point and 0.4μs over the fronthaul to
the radio. Different timing synchronization solutions can be used to ensure all the radio units in the network are
synchronized which will allow the schedular at the base stations to make sure interference is minimized.
Network
B
Network
A
F (MHz)
3400 3800
BS BS
Desired links Unwanted interference
(four scenarios)
MS MS
If the two networks are phase and frame synchronized such issues can be avoided; however, it is challenging to
deploy multi operator synchronized networks. In cases where LTE networks are already deployed, this may become
more challenging as 5G-NR new frame structures bring new compatibility and performance issues. The purpose
of synchronized operation where co-channel adjacent networks or collocated adjacent channel networks are
synchronized is to prevent BS-BS and MS-MS interference scenarios. Synchronized in this scenario means more than
just have a common Coordinated Universal Time (UTC) reference; rather it also requires compatible frame structure
across operators. Frame and slot synchronization will help in avoiding performance degradation due to cross link
interference without requiring additional mitigation techniques such as additional filtering, inter-operator guard
bands, geographical separation between base stations, etc. Hence, inter-network synchronization can simplify
deployment as less coordination for cell-site radio planning is required.
In summary, for a TDD LTE or 5G-NR network (where TDD is the only option for C-band), we not only need
frequency and phase synchronization, but also frame and slot synchronization to avoid inter-network interference.
Understanding the different types of synchronization and some of the requirements and recommendations
proposed by 3GPP, ITU-T and other regulatory bodies such as ECC is essential to understand the complexity of
deploying a 5G-NR TDD network. Additionally, with the evolution of RAN to an open RAN (O-RAN) architecture,
timing and sync requirements and testing of timing and sync will be even more important because additional
delays from open interface network nodes may need to be considered for seamless 5G services.
Frequency Synchronization
Two clocks that are aligned in terms of their repeating interval (i.e. frequency) but not in terms of phase or time.
Frequency
Phase Synchronization
Two clocks that are aligned in terms of their repeating interval (i.e. frequency) and also phase (a one second
interval), but without a common time origin.
Phase
Time Synchronization
Two clocks that are aligned in terms of their repeating interval (i.e. frequency), their phase (a one second interval),
and share a common time origin
Time
2020-02-18 2020-02-18
[Link] [Link]
T=TO Time (t)
In wireless communications, the receiver does not have prior knowledge of the physical wireless channel or
propagation delay associated with the transmitted signal. Typical communication receivers use low-cost oscillators
to keep the cost of the devices manageable. These oscillators inherently have some drift. Hence, using timing
synchronization as a process by which a receiver node determines the correct instance of time at which to sample the
incoming signal and carrier synchronization as a process by which a receiver adapts the frequency and phase of its
local carrier oscillator with those of the received signal, the receiver node can demodulate received signals properly.
Synchronization definition and procedures may vary depending on the specific communication system. For example,
in terms of OFDMA, timing synchronization may consist of frame, slot, and symbol synchronizations, residual timing
tracking, first arrival path search etc. Similarly, carrier synchronization may imply integer or fractional frequency offset
estimation, etc. In 5G NR a carrier accuracy of 50 parts per billion and timing accuracy of 10μs is required. However,
for LTE/5G NR TDD this requirement goes to a more stringent 1.5 μs. For advanced features such as MIMO, location-
based services, etc., timing accuracy of a few 100ns is required. See Table 2 for timing and sync requirements, type of
synchronization, and whether absolute verses relative synchronization is needed and the effects of noncompliance.
1. Absolute TE – Time difference between a node and Primary Reference Time Clock (PRTC) which is the Grand
Master time reference. It can be measured using precision timing protocol (PTP) for 5G-NR TDD system, ITU-T
recommends 1.1 microseconds up to the access point.
2. Relative TE – Time difference between inputs into two radio units. As identified in Table 4, meeting relative
TE requirements is essential for advanced features including carrier aggregation, MIMO, CoMP, and location-
based services.
UE
GM/PRTC
Relative TE
UE
TAE ≤ 65 ns
UE
UE TAE ≤ 130 ns
UE
Inter-band
Non-Contiguous Intra-band Contiguous
TAE ≤ 260ns TAE ≤ 130 ns
UE UE
Another component to understand is the Time Alignment Error (TAE), which is the time difference between two
antenna ports, measured over the air using GPS or a common timing source as a reference. Over the air TAE
measurements limits differ between two RUs, carriers or antenna ports (for MIMO) as shown in Figure 8.
Although timing and carrier synchronization are necessary for successful communication, they cannot provide
a common notion of time across distributed nodes. Clock synchronization is the process of achieving and
maintaining coordination among independent local clocks to provide a common notion of time across the
network. For example, GPS receivers have been used as the most common time synchronization source at the
cell-sites in the past. This may not be a cost-effective option in 5G.
Ethernet is backwards compatible which allows for commodity equipment, enabling greater convergence of
access networks, and enabling statistical multiplexing which will help lower the aggregate bit-rate requirements.
Use of standard IP/Ethernet network switching/routing will also make functional virtualization and overall network
orchestration relatively easy. The challenge is that Ethernet is not synchronous. In the brave new world of 5G
(either eCPRI or O-RAN), the synchronization plane will be carried independently over an Ethernet layer and will
not be restricted to specific protocol. Global positioning system (GPS), precision time protocol (PTP), synchronous
Ethernet, or something similar can be used for timing and synchronization.
Powered GPS
Splitter
GPS
Core NodeB
Backhaul
EPC CU DU RU
NGC Backhaul Midhaul Fronthaul
Figure 9 - GPS based synchronization. For 4G BBU and RRH are synced using CPRI
To fully understand the exact time of day at the satellite receiver, we need to be able to compensate for the delay
between the time when the satellite sends the time of day message and when that message arrives at the satellite
receiver. However, this becomes challenging because satellites are not stationary above us.
The challenge is handled as follows. All satellites periodically transmit an ephemeris. The ephemeris of a satellite
is a mathematical description of its orbit. All satellite receivers calculate an accurate position of where they are.
This calculation is called conducting a survey and uses the mathematical technique trilateration which is similar
in concept to triangulation. Once an accurate position is calculated, in other words once the survey is complete,
then the delay between the satellites and the satellite receiver can be computed to “correct” the time of day that
it was received.
Position
To accurately perform this calculation and establish an accurate position, a minimum of four satellites, as shown
in Figure 10, is required. There are four variables to account for – longitude, latitude, altitude, and time, hence
the need for four satellites. The longer a survey runs the more accurate the position calculated will be. The more
accurate the position of the satellite receiver, the smaller the time error between cells and the lower the chance
that overlapping cells will interfere with each other.
FTN Mux
Grand Master FTN Mux
Clock
Multiple options to meet the stringent phase and time synchronization requirements exist, i.e., to choose the
S-plane configuration. The intent is to ensure all nodes are synchronized to the PRTC source. The location of the
source may vary depending on the network topology, cost, and application. By using a grand master clock synced
to a satellite source and a combination of boundary clock and slave clocks, network nodes can be aligned to a
common time and phase. Options include:
1. Install GNSS/GPS at all cell site. This can be expensive and may not be a practical option in some cases.
(Figure 12, Case 4)
1. GNSS at some RAN or Transport sites with full timing support (FTS) from the transport network using
PTP1588v2. Every node in the transport network must have a boundary or transparent clock. Careful placement
of GNSS and reference clocks is required to cover the network within the timing budget (Figure 12 case 1 to 3)
Other options such as assisted partial timing support can also be implemented with appropriate consideration for
the network topology and cost.
TSC
TDD 5G NR Synchronization
As mentioned above, with TDD deployments, in addition to frequency and phase synchronization, to avoid intercell
interference a compatible frame structure should be used between collocated networks with adjacent frequency
assignments, or adjacent networks sharing the same frequency or adjacent channels. Essentially carriers must
prevent simultaneous UL and DL transmission occurrence. I.e., at any given moment in time, either all networks
transmit in DL or all networks transmit in UL adopting a single frame structure for all TDD networks involved as well
as synchronizing the beginning of the frame across all networks. Refer to ECC Report 296 and its recommendations
for more details.
To summarize, DL spectrum may leak onto the adjacent channels. For FDD, this is acceptable since UL and DL
channels are separated by a guard band. For TDD, UL and DL share the same channel. Any DL spectral imperfection
may thus create interference in the UL signal of the adjacent operator, especially when the two cells are at the
boundaries of each other.
Hence, if two 5G networks operating in adjacent channels are not synchronized, an additional guard band of 25MHz,
as well as extra filters on the emitters may need to be provisioned.
Interference
NR Carrier 1 DL DL DL DL UL
With synchronization
The VIAVI T-BERD®/MTS-5800 (100G) along with CellAdvisor® 5G (CA5G) can perform all required timing and
synchronization test for all types of 5G networks. They measure throughput, delay, packet jitter, timing, and frame
synchronization to ensure backhaul, midhaul, fronthaul, and air interface meet designed network specifications.
For fronthaul test applications, the VIAVI T-BERD/MTS-5800:
y Signal strengths
y CNO map spectrogram plots line of sight to satellites as they move around the orbit over time
RRH
RIU
T-BERD/MTS-5800
RRH DU CU
PTP/SyncE
T-BERD/MTS-5800
Validate Frequency and Time Error versus UTC Over the Air
Using a VIAVI CellAdvisor 5G, an RF engineer or a technician can quickly validate the over the air frequency and
time errors, ensuring synchronization conforms to the +/- 1.5µs vs UTC. This can be tested for the adjacent channel
network as well.
Using the CellAdvisor 5G with NSA Signal Analysis to perform over the air tests revealed synchronization issues at
one site. These sync issues, in turn, were causing poor performance not only in the immediate coverage area but
also in the surrounding cells which were correctly synchronized. As shown in Figure 18, Physical Cell Identifier (PCI)
326, Timing Errors (TE) and Offsets are within specs, whereas neighboring sites with PCI 138 and 139 are completely
outside the specification requirements.
A lack of synchronization, as found with this provider’s site, results in lower QoS, which can cause lower user data
throughputs, inability to perform proper handover from cell to cell, and in some cases, even prevent users from
connecting to the network due to cross link interference caused by uplink slots interfering with adjacent cells slots.
Upon arming the RF engineers with this information, the service provider was able to identify the source of the
synchronization issue and brought it to resolution quickly. Figure 19 shows before and after measurements.
Figure 18 - Over the Air Timing Error measurements at two neighboring cell-sites using a CA5G
vs.
Figure 19 - Before and after timing and sync measurements using CA5G
Conclusion
Synchronization is fundamental to the performance of a cellular network and the services it offers. Both 3G and
4G cellular technology required frequency synchronization primarily to prevent interference when cells overlap.
However, 5G cellular technology and 4G LTE-TDD require phase and frame synchronization, with much stricter
synchronization requirements. Further, the need for validation of timing error has become an essential test for cell
site installation and maintenance. This holds true for slot and frame synchronization for TDD deployments as well.
VIAVI Solutions is the industry leader in test and measurement and offers the most comprehensive timing and sync
validation solution. With the fully integrated VIAVI portfolio of cloud-enabled instruments and systems, software
automation, and services for network testing, performance optimization, and service assurance, operators and their
partners can be assured of a smooth network roll-out and sustainable network lifecycle.
[Link]









