Yao Dan, deputy general manager and chief technology officer of Ericsson of Taiwan, stated that in the 4G era, the actual service quality statistics parameters are usually monitored through Key Performance Indicators (KPIs) in order to observe the truest appearance of the provided services; In the 5G phase, in addition to the sudden increase in network traffic, different technologies such as network slicing, software-defined networking (SDN) or network function virtualization (NFV) were introduced, and only the existing traditional KPI network behavior analysis was used. , It seems unable to fully meet the large data volume analysis and new technology support required by 5G, so it needs AI platform and machine learning to assist in automatic analysis, build more intelligent networks, and improve user experience.
Based on this, Ericsson is actively developing a dedicated AI expert analysis system for telecommunications companies - EEA (Ericsson Expert Analytics). Yao Dan mentioned that Ericsson and other companies in the industry such as Google, IBM, Microsoft and other companies are focusing on the development of AI platforms. Different, and these platforms can be applied to any industry; Ericsson is focused on the development of AI and machine learning technologies that are best suited to the communications industry.
Yao Dan stated that when 5G is moving toward higher-frequency development, it needs more base station coverage. In Taiwan, it needs nearly 100,000 to 200,000 base stations. However, if the data read by the base station does not have an AI platform to judge and analyze, these The amount of big data data will be useless. However, with the introduction of AI technology, it can meet three application modes, including more intelligent optimization and management of the network, improving the customer experience, and assisting with different customer characteristics to achieve accurate marketing operations. Telecom companies to create better marketing and marketing applications.
For example, Sima Data, a subsidiary of China Mobile, analyzes demographic characteristics, mobile terminal characteristics, social communication features, outbound travel characteristics, spatial location distribution, and APP usage behaviors through mobile network data collected by telecommunications companies. The big data system. If a store owner is going to open a new store and seeks a confirmation location for the store's customer base, he can enter Sima Data, input some conditions, and then use Sima Data to compare user data and mobile services. Customer base conditions to provide a reference for the location of the proposed shop.
It is worthwhile to mention that Yao Dan believes that in the future telecommunications companies should move toward building an AI platform, because communication will be the same as electricity in the coming years and it will be inseparable from people’s lives. Therefore, users of the next generation will have The data traffic will be very alarming. If the telecom operators' own systems do not have AI capabilities, they will not be able to meet the huge data analysis and traffic demands in the future.