For the mobile communication service, the most important delay is the end-to-end delay, that is, the delay of the data packet from the transmitting end to the receiving end correctly received at the transmitting and receiving ends of the established connection. According to different service models, the end-to-end delay can be divided into one-way delay and backhaul delay. The one-way delay refers to the delay from the transmitting end to the other receiving end through the wireless network. The backhaul delay refers to the data packet. The delay from the transmitting end to the destination server receiving the data packet and returning the corresponding data packet until the transmitting end correctly receives the response data packet.
The existing mobile communication is mainly human-to-human communication. With the miniaturization and intelligence of hardware devices, future mobile communication has more high-speed connection applications between "people and things" and "objects and things". Machine Type CommunicaTIon (MTC) has a wide range of applications, such as mobile medical, vehicle networking, smart home, industrial control, environmental monitoring, etc. It will promote the explosive growth of MTC system applications, and a large number of devices will access the network to achieve real The "Internet of Everything" brings infinite vitality to mobile communications. At the same time, a wide range of MTC system applications will bring new technical challenges to mobile communications, such as real-time cloud computing, virtual reality, online games, telemedicine, intelligent transportation, smart grid, remote real-time control, etc. There is a higher demand for delay, and the existing LTE system cannot meet this demand and needs to be studied.
This paper mainly introduces the delay demand of future MTC services, analyzes the existing delay of LTE system, and expounds the key technologies to reduce delay.
MTC service delay requirement analysisIn the future, the MTC data transmission delay will be further reduced. When the communication response time is faster than the system application time constraint, a real-time communication experience can be obtained. The time constraints for the four typical applications are given below:
â— The human muscle response time is 0.5s~1s, which means that when a person clicks on a connection, if the connection can be established in 0.5s, people can realize real-time web browsing experience.
â— Hearing: When the sound signal can be ready to receive within 70ms~100ms, people can realize real-time call. Considering the speed of sound waves, this means that when two people are more than 30m away, they cannot rely on sound waves for real-time communication.
â— Vision: The resolution of human vision generally does not exceed 100 Hz, which means that as long as the image update rate is not less than 100 Hz (delay does not exceed 10 ms), people can get a seamless video experience.
â— Haptic: In this respect, real-time is required, and the delay is limited to several ms. The applications involved include the use of mobile 3D targets, virtual reality, business security control in intelligent transportation, and smart grid.
The industry has proposed to reduce the end-to-end delay of existing systems by more than 5 times, and considers RTT (Round Trip TIme) on the order of 1 ms when considering the requirements of the 5th generation mobile communication system. Real-time games, M2M, sensor alarms, or event detection scenarios should be the focus of research. Some scenarios require no more than 100ms of delay. Among them, sensor-based alarm or event detection scenarios have a latency requirement of at least 2ms.
Therefore, in the ultra-low latency scenario, the MTC system delay needs to consider the millisecond-level air interface delay.
LTE system existing delay analysisThe target of ITU-R for transmission delay setting is 10ms for one-way delay. The LTE/LTE-A system meets the ITU delay requirement with a certain margin, and the one-way packet transmission delay is less than 5ms. The downlink downlink data is transmitted by the Physical Downlink Shared Channel (PDSCH) and the uplink data is transmitted by the physical uplink shared channel (PUSCH).
In the LTE FDD system, on the subframe n, the base station uses the Physical Downlink Control Channel (PDCCH) to schedule downlink data transmission, and the terminal feeds back ACK/NACK information on the subframe n+4, and the base station receives the processing delay. The minimum time is 1ms, and the base station can perform data retransmission scheduling on the subframe n+5 at the fastest. As shown in Figure 1, the single transmission time is 1ms, and the minimum time for one retransmission is 5ms.
Figure 1 downlink data transmission
In the LTE FDD system, when the terminal has a data transmission requirement, it needs to wait for the subframe n of the schedule request (Schedule Request, SR) to be configured, and the terminal sends the scheduling request information to the base station in the subframe n, and the base station is in the subframe at the fastest. The uplink data scheduling authorization information is sent on the n+2, and after receiving the uplink data scheduling authorization information on the subframe n+2, the terminal transmits the corresponding uplink data in the subframe n+6, and the base station feeds back in the subframe n+10. The ACK/NACK information is sent to the terminal, and the terminal retransmits the uplink data on the subframe n+14. As shown in FIG. 2, the data transmission request is completed until the data transmission is completed, regardless of the time of waiting for the scheduling request subframe. The delay of a single transmission is 6ms, and the time of one retransmission is 14ms.
Figure 2 uplink data transmission
Low latency technical analysisIt can be seen from the analysis of the existing LTE air interface delay that the main factors affecting the air interface delay are the data transmission duration, the data transmission resource request waiting time, and the feedback delay caused by data processing. For these factors, there are the following four types of air interface reduction. Extended plan.
Reduced data transfer duration
The existing LTE system performs data scheduling in units of subframes, and the LTE subframe length is 1 ms. Therefore, the minimum data transmission duration is 1 ms. In order to reduce the data transmission duration, there are two possible solutions. One is to reduce the length of the subframe, such as redesigning the subcarrier spacing and the number of OFDM symbols included in one subframe, so that the duration of one subframe is shortened, thereby reducing the data transmission duration. For example, the length of the subframe is compressed to 1/4 of the length of the existing LTE subframe, that is, 0.25 ms. If the compression of the corresponding processing time is considered, the specific compression effect is as shown in Table 1, and can be compressed by about 75%.
Table 1 delay compression ratio
Another scheme is to perform data scheduling transmission in units of OFDM symbols. In this case, the minimum data transmission length is 1 OFDM symbol, and the length of one OFDM symbol is 66.67 ηs according to the OFDM symbol length of the existing LTE. The time is proportionally compressed. The specific compression effect is shown in Table 2. Compared with the existing 1ms data transmission, it can be compressed by about 92%. If the frame structure is further modified, such as the subcarrier spacing change, the length of the OFDM symbol can be further reduced. , to achieve lower latency compression.
Table 2 delay compression ratio
In addition, enhanced HARQ feedback also helps reduce retransmission delay. The traditional HARQ only feeds back the ACK/NAK information. The enhanced HARQ can additionally feedback the received BER estimation information. Combined with the information and the channel inverse status information, the scheduler can be more targeted in terms of redundancy version selection and MCS selection. The probability of correctly decoding the data after retransmission is greatly improved, thereby further reducing the data transmission delay.
Reduced latency due to data transfer resource requestsIn the LTE system, when the terminal has data transmission requirements, the scheduling request needs to be sent first, and the base station can allocate resources to allow the terminal to perform uplink data transmission. This process causes the uplink data transmission delay to be significantly larger than the downlink data transmission delay, as shown in Table 3. Show. In addition, sending a scheduling request to configure a resource for transmitting data by the terminal additionally increases the delay. Therefore, if the base station can pre-allocate the resource terminal, the terminal directly transmits data on the pre-allocated resource when there is data transmission, thereby reducing the scheduling request process. Therefore, the uplink data transmission delay is equivalent to the downlink data transmission delay, so that the uplink data transmission delay compression is about 17%, and the retransmission delay is 36%, and the data transmission delay reduction scheme can be combined with the foregoing data transmission delay reduction scheme. Further reduce the uplink data transmission delay.
Table 3 Comparison of uplink and downlink data transmission delays
Shenzhen Sunbeam New Energy Co., Ltd , https://www.sunbeambattery.com