Shulou(Shulou.com)11/24 Report--

Teacher: one side... On one side... . It means that two things are going on at the same time. Please make a sentence.

Xiao Ming: the base station sends and receives.

Teacher: hmm? Your father is a communications engineer, right?

Hello, everyone. I'm mayday.

In this issue, let's talk about one of the most basic topics in the field of wireless communications: duplex.

The so-called duplex refers to the two-way data transmission between two communication devices. Specifically, duplex technology includes two modes: full-duplex and half-duplex.

Full duplex means that two-way data transmission can be carried out at the same time. In other words, both sides of the communication can receive the data sent by each other while sending data, sending and receiving data in parallel.

Half-duplex can be much more crude, sending and receiving can not be carried out at the same time, can only be carried out in turn: you can not receive it when you receive it, and you cannot send it when you receive it. This is the mode of our common walkie-talkies.

Full duplex, did we use it? After five generations of development, mobile communication technology can be said to be at its best. So, when our base station and mobile phone interact, we must use full duplex, right?

At this time, our two most commonly used terms are FDD (frequency division duplex) and TDD (time division duplex). So are they full-duplex or half-duplex?

For FDD, we use two segments of spectrum, one specifically used as a base station to send signals to mobile phones, also known as downlink, and the other as a mobile phone to send signals to base stations, also known as uplink. In order to prevent interference between downlink and uplink, there must be a certain isolation band between the two segments of the spectrum, which is called "duplex bandwidth".

Thus it can be seen that the downlink and uplink of FDD are half-duplex, and they combine to form a "pseudo full-duplex" system at the cost of doubling the spectrum resources. This is like the lane on the road, each lane can only be one-way, but the combination of lanes in different directions can achieve two-way traffic.

For TDD, it is true that the spectrum takes up only one segment, but uplink and downlink can only be used in turn. In other words, when the base station is sending data, the mobile phone can only receive it quietly, and if you want to send the data, you can only wait until the sending time for you is up.

Isn't this a real half-duplex? Our commonly used 5G band is in TDD mode, but the switching time between upstream and downlink is very short, which is millisecond, and we can't feel it at all. Therefore, TDD uses microscopically fast half-duplex to achieve "pseudo full-duplex" macroscopically.

Can't we transmit and receive at the same time on the same spectrum and achieve real "simultaneous full duplex on the same frequency"? In this way, the spectral efficiency is directly doubled! Isn't it beautiful to have national security and folk music?

However, the problem that everyone can see for so many years is still unsolved, among which there must be great difficulties that are extremely difficult to solve.

To achieve full duplex is tantamount to two trains running at high speed in the opposite direction on the same track, and the result is self-evident.

The reason for this is that it will cause huge interference when sending and receiving at the same time in the same frequency band. This includes not only the self-interference sent by the base station to its own reception, but also the interference between the base station and the base station, the interference between the base station and the mobile phone, and the interference between the mobile phone and the mobile phone.

Therefore, everyone can only focus on increasing the lanes, pushing up the frequency band, widening the carrier bandwidth and doubling the number of transceiver channels.

For example, from 2G to 5G, from low frequency (less than 1GHz) to intermediate frequency (less than 6GHz), and then to millimeter wave or even terahertz, the channel bandwidth extends from a few megabytes to tens of megabytes, hundreds of megabytes and even terahertz, and the number of transceiver channels also extends from single channel to dual channel, 4 channels, 8 channels, 32 channels, 64 channels and even 128channels.

As for full-duplex technology, although it was widely discussed in the initial stage of 5G standardization and regarded as one of the key technologies of 5G, it was finally put on the shelf because it was difficult to implement.

Let the wireless communication use the real full-duplex, has become the biggest dream of the communication person buried deep in the heart.

Realistic compromise: subband full-duplex star shift, 5G has been commercial for several years, the technical standard of the second half of 5G: 5G-Advanced standard is in full swing.

Full-duplex, once again entered everyone's field of vision.

This is because, with the penetration of 5G industry applications to the industrial field network, the need for the network to support both ultra-large uplink bandwidth and ultra-low latency highlights, the current FDD and TDD models are difficult to resist.

For example, video surveillance, electronic fencing, machine vision and other applications in the factory are mainly large uplink services. The bandwidth requirements of multiple terminals range from hundreds of Mbps to Gbps; industrial AR with a delay of less than 10 milliseconds, AGV cooperative handling requires a delay of less than 5 milliseconds, and machine motion control requires a delay of less than 4 milliseconds.

Why is it difficult for both FDD and TDD modes to meet the requirements of large bandwidth and low latency at the same time? Let's talk about it.

As the use of spectrum division has long been determined in history, it is impossible to change without taking all the currently applied systems offline, so different frequency bands are actually strongly bound to FDD or TDD duplex mode.

The binding relationship between frequency band and duplex mode FDD frequency band is characterized by low frequency band, less available bandwidth and limited rate that can be provided. For example, 900m uplink and downlink each has 35m bandwidth, 1800m uplink and downlink each has 75m bandwidth, these few valuable resources have to be distributed to a number of operators, each is even more stretched, although the coverage is good, but the network speed can not go up.

Although the rate is limited, FDD mode has an outstanding advantage, that is, the uplink and uplink data are sent on their own independent spectrum, which can basically be sent when there is data, without requiring card time like TDD, so FDD can achieve a relatively short delay.

The TDD spectrum is on the contrary, the frequency band is generally higher and the available bandwidth is large. For example, on 3.5GHz, Unicom and Telecom each have 100m of bandwidth; on 2.6GHz, Mobile enjoys 160m of bandwidth alone.

These TDD large bandwidth carriers can achieve high uplink or downlink rates by setting different uplink and downlink time slot ratios, but the delay is difficult to reduce due to the half-duplex characteristics of TDD itself.

Although we do not feel the impact of delay, the communication between machines in the factory is extremely sensitive to it. And such a stringent delay requirements are still rigid, can not be achieved to work.

If we can combine the advantages of TDD and FDD in the same frequency band, can't we support large bandwidth and low delay at the same time?

So someone thought, isn't the bandwidth of your TDD spectrum large? I just cut inside the TDD carrier and divided it into two sub-bands (called subbands). Both subbands are still in TDD mode, but the configuration of uplink and uplink time is opposite. In this way, I receive when you send, and I send when you receive. Doesn't this have the temperament of FDD?

In this way, through subband division and time slot allocation, we can integrate the technology of FDD in the TDD carrier at a small cost, and realize the pseudo "full duplex" in the TDD carrier.

TDD and subband full duplex but this kind of pseudo "full duplex" is essentially the stitching of TDD and FDD technology, actually does not achieve the improvement of spectrum efficiency, but only a small step in the long march of full duplex, so it is called "subband full duplex", referred to as SBFD (Subband Full Duplex).

How to divide the subbands? From a purely technical point of view, the upstream and downlink can be divided into half and half, so that the downlink rate is balanced; you can also draw more downlink subbands so that you can achieve a large downlink rate; or you can achieve a large uplink speed by delimiting more uplink subbands.

From the point of view of demand, what the general public needs for video browsing is a large downlink speed, but the requirement for delay is not high, and there is no demand for full-duplex sub-band; while in factories, applications such as data reporting, surveillance cameras and machine vision need large uplink, and a large number of control applications require low latency, so subband full-duplex is useful in industrial scenarios and needs to be configured as the above line subbands.

As for the need to divide several subbands, two from the point of view of use is enough, but in fact this mainly depends on the interference.

If you want to deploy a full-duplex subband operator whose spectrum is adjacent to other operators, it is best to keep the adjacent spectrum as it is for downlink, and try to keep the uplink subband as far away as possible to minimize interference.

Specifically, if there is a spectrum of neighboring operators on both sides of the spectrum, it is recommended to divide two downlink subbands and one uplink subband, and put the uplink subband in the middle and configure it according to the sandwich mode of downlink + uplink + downlink; if there are only other operators on one side, it is better to divide one downlink subband and one uplink subband.

The upstream and downlink subbands divided in the "sandwich" form are on the frame structure. In order to be compatible with the existing terminals, the traditional frame structure of DFFFU can be maintained. The first time slot is full downlink, the middle three time slots are configured with subbands and the upstream and downlink are flexibly scheduled as needed, and the third time slot is full uplink.

How to eliminate the interference of the two configurations of the upper and lower subbands? Self-interference in the system is the core problem that must be solved in subband full-duplex.

Because the subbands of full-duplex uplink and downlink subbands are close to each other, there is no duplex interval of tens of M in the uplink and downlink bands of traditional FDD, which will lead to serious transceiver interference.

In general, the signals sent and received by the base station share a common antenna, and the strong signals transmitted will be received directly, resulting in the flooding and blocking of the weak signals from the mobile phone that should have been received. In addition, during the internal processing of the base station, there will also be coupling interference between the RF transceiver links.

Self-interference suppression has space domain, RF domain, digital domain and other means, multi-pronged, multi-stage elimination.

Self-interference cancellation in spatial domain, radio frequency domain and digital domain the simplest means of self-interference suppression in spatial domain is the separation of transceiver antennas. Transmitting and receiving can effectively prevent the transmitted signal from entering the receiving antenna by using separate antennas and adding a plurality of barriers between the two antennas. Coupled with the beam nulling technology of the transmitting antenna in the direction of the receiving antenna, the interference can be further reduced.

There are two ways of interference suppression in radio frequency domain: subband filter and radio frequency interference cancellation.

By adding a subband filter in the base station, the downlink subband can filter the signal of the uplink subband through the filter, and the upstream subband can filter the signal of the downlink subband through the filter. This method is relatively simple, but it is not flexible to adjust the filter bandwidth and will increase the insertion loss.

Radio frequency interference cancellation is by collecting a copy of the known downlink transmitted signal and transmitting it to the uplink receiver, and then canceling it by constructing the opposite signal. This method is more complex and expensive.

Radio frequency interference cancellation radio frequency domain interference suppression can be realized by evaluating the required interference suppression capability and choosing one or a combination of two methods.

The idea of interference suppression in digital domain is similar to that in radio frequency domain. The residual interference is further reduced by introducing an auxiliary radio frequency channel in the radio frequency domain and converting it into a digital signal, and then constructing an opposite signal to cancel it in the digital domain.

Through the self-interference cancellation in spatial domain, radio frequency domain and digital domain, the self-interference can be suppressed to a slightly reduced but acceptable level of sensitivity.

Solved the self-interference, that is, a single base station can work normally, but in actual deployment, there can not be only one base station and one terminal, but multiple base stations need to form a network and serve multiple different terminals at the same time. This involves a more thorny problem: cross-link interference.

In order to eliminate the cross-link interference, it is necessary to design the corresponding interference measurement mechanism to know ourselves and the enemy, and transmit the known interference characteristics, and then eliminate the interference by beam zero notch, interference suppression combining and other technologies. This process is more complex than the sub-interference cancellation in a single base station, which is still under research in the industry.

In order to take the first step smoothly, we should move from easy to difficult and step by step. First of all, we can deploy the subband full-duplex microstation in the intelligent factory, which has less power and is relatively easy to isolate from the outdoor macro station.

In the follow-up, we consider the networking between multiple subband full-duplex base stations, and finally we try to solve the networking between the subband full-duplex macro station and the existing large downlink macro station. With the progress of solving the problem of networking interference, the industrial ecology will naturally mature.

The full duplex of standardization subband has been established in 3GPP R18 and is currently in the stage of SI (Study Item). Theoretical and engineering research has been carried out in an all-round way.

China Mobile leads the standardization of full-duplex technology and packages it into a series of UDD (Unified Division Duplex) technologies. Among them, S-UDD (Single carrier UDD, single carrier UDD) refers to subband full duplex. Similarly, Samsung packaged the technology as XDD (cross division duplex).

Although some progress has been made in the current research, the technology is still far away from formal commercial use. According to the R18 study, the pace of R19 standardization, the relevant agreements are not expected to be frozen until 2025, commercial is expected to be after 2026.

In 2026, there are only three years to go before 6G. Therefore, in order to successfully promote the commercial implementation of subband full-duplex technology, we must focus on the compatibility of existing terminals. Because the upgrading of the base station side is usually easier to promote, while the popularity of the terminal industry chain is more lagging behind.

After realizing the full duplex of the subbands mentioned above, we can go further, make the subbands overlap and study how to make a small amount of spectrum at the overlap do the same frequency full duplex at the same time. In the next step, we will push the entire carrier to simultaneous full-duplex on the same frequency. This is a step-by-step process.

In any case, subband full-duplex will serve as an important milestone leading to simultaneous same-frequency full-duplex, giving play to the value of connecting the past and the future at the turn of the 5G and 6G era.

The difference between FDD and TDD will eventually become the dust of history.

This article comes from the official account of Wechat: wireless Deep Sea (ID:wuxian_shenhai), author: mayfly acquisition

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.