What is the big difference between 5G and 4G in terms of demand?
High speed, a large number of connections, low latency. Then all the newly added technologies must serve the above three needs.
First of all, let’s take a look at what technologies 5G lacks compared to 4G. The answer is one: Turbo code.
At 0:45 am on November 17, 2016, US time, in the discussion of the 5G shortcode solution at the 87th meeting of 3GPP RAN1, Huawei’s Polar code solution eventually defeated the US LDPC and France’s Turbo 2.0 solution, became the 5G control channel eMBB (Enhanced Mobile Broadband) scene coding final scheme.
Secondly, let’s talk about some new technologies that 5G has added compared to 4G.
That is millimeter wave, the frequency is approximate. Since the frequency bands below 3G have been used up, there are no excess frequency bands for 5G use, and the bandwidth used by 5g is often several hundred M, so it can only develop in the high-frequency direction. But the higher the frequency, the greater the loss, which brings about a problem. As the frequency increases, the coverage will get worse.
Is there any way that we can use high-frequency millimeter waves without reducing or even enhancing the coverage performance?
Yes, it is a massive MIMO.
Massive MIMO. Everyone knows MIMO, that is, multiple antenna ports transmit and receive at the same time, which brings diversity gain. In the 4G era, 4 antennas and 8 antennas are generally used more. So what is massive MIMO? As the name implies, many antennas that transmit and receive at the same time. How much is this “many”? The answer is as follows:
That is to say, when the 3Ghz frequency is used, the base station can use up to 256 antennas to transmit and receive simultaneously. When using 70Ghz frequency point, up to 1024 antennas can be used. With so many antennas, the effect is much better than 4G. After all, the cost is not cheap. So it is estimated that the MU-MIMO mode will be in actual use. Massive MIMO is well, but it also brings some problems. For example, when there are more and more antennas, the beam will become narrower and the coverage area will be affected. What does it mean? Please see the picture below,
In the above picture, from left to right are the beams of one antenna, two antennas, and a lot of antennas. We can see that when only one antenna is used, the signal is evenly covered from all directions. Mobile phone 1 2 3 The received signal is equal. When two antennas are used, the signal coverage has a certain degree of directivity, and the signal coverage directly below is stronger than the left, right, and bottom. When using many antennas, the beam becomes more concentrated and the coverage area becomes a big sword. The 2nd mobile phone can get stronger coverage, but the 1st and 3rd mobile phones cannot receive the signal.
So is there a way to solve the above problem?
The answer is still: yes. Please see Beam Management.
Beam Management. The principle of this function is simple: The base station sends specific things like reference signals in all directions. The terminal detects and gives feedback to the base station so that the base station knows its direction.
LDPC coding. 5G abandons the turbo code used in 4G and replaces it with LDPC. Why change it? How is LDPC code better than Turbo? There are two main reasons: 1)The most important reason is that Qualcomm is awesome. 2) Because Turbo coding introduces operations such as interleaving. When the code length is longer, the complexity increases and the time delay becomes very large. However, as we said in the beginning, low latency is one of the three requirements of 5G, so Turbo seems a little weak. LDPC is different. Due to the sparseness of its check matrix, its decoding algorithm has a shorter delay, which has obvious advantages over Turbo when it comes to long codes. Therefore, 5G abandons Turbo and uses LDPC which sounds right.
UL Waveform. In 4G systems, OFDMA is used in the downlink, and SC-FDMA is used in the uplink. It is due to the high peak-to-average ratio of OFDM and stricter requirements on equipment hardware. To reduce the cost of mobile phones. After discussion, someone decided to use SC-FDMA instead of OFDM for the uplink transmission of the 4G system.
So in the 5G area, what kind of multiple access was used? After the R15 agreement came out this month, I found no many changes in this area. A major change is that the uplink supports OFDM and DFT-S-OFDM. In addition, the sub-carrier spacing in 4G is fixed at 15Khz, However, because 5G is at high frequencies, the usable bandwidth is very wide, so a new term is introduced: numerology.
In 5G, the sub-carrier spacing is not fixed at 15Khz in the 4G area, but variable, but one RB still has 12 sub-carriers, and this has not changed.
Subframe Structure. We know that in 4G, a radio frame is 10ms, a subframe is 1ms, and a slot is 0.5ms. In 5G, the length of the radio frame and subframe has not changed, and it is still 10ms and 1ms. But the slot length has become configurable, and its value depends on two parameters: μ and slot configuration. See the below picture:
When μ is 0 and slot configuration is 0, 1 radio frame contains 10 subframes, 1 subframe contains 1 slot, and 1 slot contains 14 symbols.
When μ is 1, the slot configuration is 0, it becomes the following situation: 1 subframe contains 2 slots, and each slot has 14 symbols, That is to say, 1 subframe contains 28 symbols.
By analogy, when μ is set to 5 and slot configuration is set to 0, a subframe can have up to 32*14=448 symbols, which is ten times more than symbols. This brings the possibility of exponentially increasing the speed.
5G introduces the concept of self-contained subframes, that is, the HARQ cycle is from the minimum 4ms in the 4G era. Shorten it to within 1ms. This provides some help for the ultra-low latency.