laitimes

MWC2022: Qualcomm released Snapdragon X70 AI to fully enhance 5G performance

MWC2022: Qualcomm released Snapdragon X70 AI to fully enhance 5G performance

Jiwei Network report (text/ Zhang Yiqun) On the 28th, the 2022 MWC (Mobile World Congress) kicked off in Barcelona, Spain, due to the epidemic two years after the farewell, MWC resumed offline for the first time.

As the highlight of the MWC in the past year, Qualcomm announced the launch of a new generation of 5G modem and RF solution - Snapdragon X70 at this year's MWC.

The world's first modem and RF system integrating 5GAI processors, supporting 10Gbps5G transmission speeds, and the world's only modem and RF system that can support all 5G commercial frequency bands from 600MHz to 41GHz have become the three highlights of Snapdragon X70, and under the full empowerment of AI, 5G performance has been significantly improved

According to Nan Mingkai, senior director of product marketing at Qualcomm Technology, qualcomm's newly launched Snapdragon X70, while inheriting the leading performance of previous generations of products, has also added many expansion functions, in addition to the three leading advantages, it has also significantly enhanced its 5G AI capabilities, transmission speed, network coverage, reduced latency, flexibility and energy efficiency, and continued to expand Qualcomm's leadership in 5G modems and RF systems.

AI enhances four capabilities

By integrating a 5G AI processor in the baseband for the first time, the Snapdragon X70 has brought a series of performance improvements under the blessing of AI, which is mainly reflected in four aspects:

First, AI-assisted channel status feedback and dynamic optimization. The terminal will detect the channel status in real time and report to the base station, according to the channel status reported by the terminal, the base station will select a more appropriate modulation method and better time-frequency resources for the terminal in a timely manner when dispatching downstream. With the blessing of AI, the prediction and feedback of channel state will be more accurate, and the base station will be better able to dynamically optimize, thereby helping to improve throughput and other key performance indicators.

According to Nan Mingkai, through the simulation data, it can be found that in the context of burst data traffic, AI-assisted channel status feedback and optimization can achieve a downward throughput improvement of 20%, 16% and 24% for the edge of the cell, the middle section of the cell and the center of the cell, respectively. In addition, AI-assisted channel status feedback and optimization not only improves throughput, but also optimizes transmission efficiency, improves spectrum utilization, and saves transmission power.

MWC2022: Qualcomm released Snapdragon X70 AI to fully enhance 5G performance

The second is the world's first AI-assisted millimeter wave beam management. Millimeter waves use antenna arrays for signal transmission and reception, while millimeter waves are very easy to combine with beamforming techniques. The introduction of AI technology in mmWave beam management enables smarter troubleshooting and prediction of uncertain and complex environments, helping to optimize mmWave beam focus and orientation for higher transmission efficiency, better network coverage, and link robustness.

"Simulation test results show that AI-backed launches have achieved a 28% improvement in overall network coverage compared to user terminal (UE) launches without AI support, which is a considerable improvement. Nan Mingkai said.

The third is the choice of AI-assisted networks. In actual deployment, the operator's network environment is very complex, with a variety of operating frequencies and different network standards, which affects the flexibility and security of the network. Through AI, network patterns can be intelligently identified and monitored, so that certain dropped or risky connection states can be predicted and avoided, thereby achieving better mobility and link robustness, reducing stuttering while improving the user experience.

The fourth is AI-assisted adaptive antenna tuning. This technology is mainly applied to the RF front end, through AI technology to intelligently detect the position of the user's hand holding the terminal, and dynamically tune the antenna in real time, thereby supporting higher transmission speeds and energy efficiency, as well as wider network coverage.

A 10Gbps peak is not the upper limit

The Snapdragon X70 released this time also shows excellent performance in supporting network connectivity.

According to Nan Mingkai, it is mainly reflected in three aspects: one is extremely high transmission rate; the other is excellent network coverage; and the third is ultra-low latency, which requires a series of standardized features and innovative performance support.

MWC2022: Qualcomm released Snapdragon X70 AI to fully enhance 5G performance

In terms of downlink, the Snapdragon X70 supports peak rates up to 10Gbps and supports downstream four-carrier aggregation, which refers to four carriers completely based on the Sub-6GHz band, such as two FDD carriers plus two TDD carriers. In the millimeter wave band, the Snapdragon X70 can support aggregation of up to 8 carriers.

On the upside side, the Snapdragon X70 supports peak rates of up to 3.5Gbps. Upstream carrier aggregation is supported, especially across the FDD and TDD bands.

"This is a very valuable feature, and it is also a technology that is currently of interest to domestic and foreign operators and actively promote deployment." The upstream emission switching based on carrier aggregation uses the transmission of FDD as a supplement to obtain the effect of continuous transmission in the time domain, thereby maximizing the use of mobile phone transmission power, greatly enhancing the network coverage at the edge of the cell, and significantly improving the peak rate in the center of the cell. Nan Mingkai said.

In addition, Nan Mingkai said that the Snapdragon X70 is equipped with Qualcomm 5G ultra-low latency kits, which include standardized and non-standardized technologies, especially in the non-standard parts with Qualcomm's unique innovations.

In the Snapdragon X65, the peak transmission rate of 10Gbps has been achieved, and this time in the Snapdragon X70, there is no significant increase in the peak rate, does this mean that the X70 has limited room for improvement?

Nam Ming Kai said that this is not the case, in fact in June 2021, at the 2021MWC Barcelona, the Snapdragon X65 achieved a connection rate demonstration of more than 10Gbps.

"In terms of hardware capabilities, Snapdragon 5G modems and RF systems support transmission speeds that far exceed 10Gbps. In terms of bandwidth, Snapdragon 5G modems and RF systems support a maximum bandwidth of up to 300MHz in the Sub-6GHz band and up to 1GHz in the millimeter wave band. We can do a very high in hardware specifications, but at present in the software has not been matched according to high specifications, the reason is also very simple - we provide 5G modems and RF systems to support mobile terminals, in actual use needs to cooperate with the operator's network deployment needs, considering the current 5G network deployment related technical specifications and connection requirements, we have made the current planning of the peak rate of 5G products. It should be emphasized that the 10Gbps peak rate supported from Snapdragon X65 to X70 can be achieved in real scenes, and we have also achieved such extremely high 5G rates in field tests. Nan Mingkai said.

Extreme flexibility for easy deployment

The improvement of the Capabilities of the Snapdragon X70 brings extreme flexibility to operators in deploying networks.

Nan Mingkai further introduced that the application of modems and RF systems requires interconnection between multiple manufacturers, which is a highly standardized field, which also means that in the main technological innovation, industry guidance and innovation along the unified track are required: first of all, 3GPP is pre-researched, after multi-party discussion, industry standards are formed, and the entire industry is screened according to the goals and routes specified in the standards, and planned according to the needs of operators.

MWC2022: Qualcomm released Snapdragon X70 AI to fully enhance 5G performance

Nan Mingkai stressed that in the time point and introduction rhythm of the entire feature planning, the industry should pay special consideration to the operator's network deployment plan. Each operator's network deployment requirements are different, with differences in geographical and spectrum resources, the need to create competitive advantages, and the need for cost control. From Qualcomm's point of view, it must be able to meet the network deployment needs of operators in different regions and at different stages of development, so Qualcomm has made corresponding innovations, especially the introduction of an upgradable architecture that combines software and hardware, which supports software upgrades and support new features according to the needs of different operators after hardware deployment.

In order to provide operators with the ultimate flexibility, the Snapdragon X70 implements a number of leading technical features:

One is millimeter wave independent networking (SA). In the early stage of 5G networking, most operators chose the non-independent networking (NSA) model, and the 5G base station was anchored to 4G and followed the 4G core network; while SA was relative to the NSA, 5G SA did not need a 4G core network. MmWave-only networking (mmWave-only) does not need to be networked with the Sub-6GHz band, and can only use the millimeter wave spectrum to network separately.

Nan Mingkai said that millimeter wave independent networking is the first commercial technology on the Snapdragon X70, which can bring advantages to operators, for example, emerging operators with only millimeter wave spectrum resources, no Sub-6GHz spectrum resources or service providers in vertical industries can quickly deploy networks by using the support of millimeter wave independent networking.

The second is multi-SIM card technology. The technology was first demanded in the Chinese market and is now a standardized feature in Europe and the rest of the world. Snapdragon X70 supports global multi-SIM functions including dual SIM dual standby (DSDS) and dual sim (DSDA), supporting independent communication between the main card and the secondary card - for example, when the secondary card makes voice calls, the main card can continue to surf the Internet; when the main card is used to play games, the secondary card will not cause the main card to play the game and the incoming call will not cause the main card game application to stutter.

The third is comprehensive spectrum aggregation functions, including NR-DC and EN-DC. Among them, NR-DC is a 5G dual-connection technology, which can support operators to aggregate a large number of spectrum resources in the millimeter wave and Sub-6GHz bands, which can not only improve connection rate and efficiency, but also ensure connection stability. EN-DC is based on the NSA model, using LTE and 5G for dual connectivity.

The fourth is to support the global millimeter wave band and sub-6GHz band, covering the sub-6GHz spectrum from 600MHz to 6GHz, and the millimeter wave spectrum from 24GHz to 41GHz. The Snapdragon X70 is the only modem and RF system in the world that supports all 5G commercial bands from 600MHz to 41GHz.

Fifth, the upgradable architecture. Both the Snapdragon X70 and X65 support upgradeable architectures, enabling commercial adoption of 3GPP Relese 16 features. Different operators have different deployment schedule requirements, leading operators also have a choice in the order of overall network deployment, Qualcomm hopes to help the commercial use of the new features of The First 16 through the Snapdragon X70.

60% increase in outstanding energy efficiency

MWC2022: Qualcomm released Snapdragon X70 AI to fully enhance 5G performance

From Snapdragon X50 to X65, qualcomm's previously released four-generation 5G modem and RF system has made many optimizations in terms of energy consumption. The Snapdragon X70 has further improved its energy efficiency.

According to Nan Mingkai, the Snapdragon X70's energy efficiency is mainly reflected in three aspects.

First, the Snapdragon X70 will improve energy efficiency by 60% by introducing the 3rd generation Qualcomm 5GPowerSave.

Secondly, the Snapdragon X70 enhances the matching of the wireless system through AI-assisted adaptive antenna tuning technology, using the same energy to achieve a higher signal-to-noise ratio, and only a lower transmission power is required to transmit the same data. Even at the same signal strength, the order of modulation and demodulation can be raised, improving spectral efficiency and helping the terminal save power consumption. In the case of transmitting the same packet, not only the transmission time is reduced, but also the transmit power of the signal is reduced, thereby reducing the power consumption.

Third, the Qualcomm QET7100 broadband envelope tracker supports envelope tracking of 100MHz bandwidth in the 5G band, which is very leading in the industry and can save transmit power.

MWC2022: Qualcomm released Snapdragon X70 AI to fully enhance 5G performance

Today, 5G is accelerating globally, reports and data from relevant research institutions show that 200 operators around the world have launched 5G commercial services, and more than 285 operators are investing in the deployment of 5G, and it is expected that from 2020 to 2025, the shipment of 5G smartphones will exceed 5 billion.

In the nearly three years since the beginning of 5G commercialization, Qualcomm has launched 5 generations of 5G modems and RF solutions, namely Snapdragon X50, X55, X60, X65, X70, which continue to promote the popularity of 5G commercial process and the performance of 5G mobile terminals. (Proofreading/Andrew)

Read on