The application of the hottest artificial intellig

Application of artificial intelligence in the telecommunications field and suggestions on the layout of operators

Application of artificial intelligence in the telecommunications field and suggestions on the layout of operators

20:01 source://

original title: application of artificial intelligence in the telecommunications field and suggestions on the layout of operators

in 2016, Google's alphago and Li Shishi's go duel attracted worldwide attention, pushing artificial intelligence (AI) to the forefront of technological change again. With the upgrading of data, algorithms and computing power, AI will become the most destructive technology field and infrastructure in the next decade. At present, artificial intelligence has been widely applied and deployed in the end, management and cloud sides of the telecommunications industry, and is accelerating a new round of reform in the telecommunications industry

image source: Graphic insect creativity

AI will become the most disruptive technology in the next 10 years

from the perspective of time sequence, the development of AI can be roughly divided into three stages up to now. In the first stage, since 1950, Turing, the famous "father of computer" in Britain, proposed a test method for judging whether machines have intelligence, that is, Turing test, marking the birth of artificial intelligence. The second stage was from 1980 to 1993, when the "expert system" using xcon appeared (a computer intelligence system with complete professional knowledge and experience), and artificial intelligence began to enter the industrialization stage. The third stage is since the 1990s. With the accumulation of big data and the exponential growth of computer hardware performance, the upgrading of algorithms has ushered in a new round of outbreak of artificial intelligence. According to the 2018 new technology maturity curve released by Gartner, AI has been everywhere with the rapid development of lft-d molding technology (long fiber reinforced thermoplastic material online molding technology) in European and American countries. In the next 10 years, AI technology will be used by the public, that is, the so-called AI democratization

with the upgrading of algorithms, computing power and data, AI is making breakthrough progress. At present, artificial intelligence has been widely used in machine vision, fingerprint recognition, face recognition, expert system, automatic planning, intelligent search, intelligent control, robotics and other fields. The combination of artificial intelligence and application scenarios will bring great changes and subversion to the fields of intelligent manufacturing, intelligent city, intelligent medical treatment, intelligent home, intelligent finance and so on, including telecommunications operations

on the terminal side, AI terminal products have become a new business entry

at present, artificial intelligence has spread to the end. Recently, high-end intelligence, which can only measure the total force required to destroy the sample, has begun to be equipped with artificial intelligence chips. Apple has already used the self-developed artificial intelligence bionic chip a11 bionic chip on iphonex to improve AR function and face recognition technology. Huawei's latest two generations of Kirin chips also include AI. Chinatelecom also released the "Ai terminal white paper" on May 17 world telecommunication day this year, which standardized and defined AI in terms of computing power, capability and application. Another important AI terminal is the very popular smart speaker in the market. As speech recognition is the most important technical capability of artificial intelligence, intelligent speakers based on speech recognition have become the entry-level terminals laid out by hardware manufacturers. According to the survey data of a market research company, the penetration rate of smart speakers in the United States has reached 16%, and continues to increase rapidly. Among them, Google h of Google does not make money at all. OME and echo of Amazon have occupied the absolute majority of the market. China's smart speaker market is also growing rapidly. At present, tmall genie of Alibaba and Xiaomi's small processing technology tolerance Da Ai speakers are also selling well

on the pipeline side, the impact of artificial intelligence on the future telecommunications network is ubiquitous

as the infrastructure of information and communication, the telecom network can provide powerful analysis, judgment, prediction and other capabilities through artificial intelligence technology and applications, and can empower elements, networks and services, which will greatly change the design, maintenance, operation and optimization of the telecom network. At present, the global telecommunication network is entering the early stage of intergenerational upgrading from 4G to 5g. 5g network will break the original closure and realize layered decoupling. AI will be gradually used in the network infrastructure layer, network and business control layer, operation and scheduling layer according to the detailed requirements of intelligent capabilities. At the infrastructure level, AI will provide AI accelerators for active hardware facilities to achieve different levels of training and reasoning capabilities. For example, the introduction of AI accelerators in the core data center can meet the centralized training and reasoning requirements of global strategy or algorithm models. Embedding AI accelerators in the base station can support equipment level AI strategies and applications. In the network and business control layer, priority can be given to integrating AI reasoning ability to realize intelligent control of the network and business, such as network intelligent operation and maintenance and intelligent tuning, rapid interception of malicious behavior, prevention of attacks, maintenance of network security through machine learning, etc. At the operation and arrangement level, AI engines can be deployed in the big data system to do in-depth and intelligent mining of data, so as to guide operation and maintenance and operation and realize intelligent operation

on the application side, the combination of artificial intelligence and scenarios will empower various vertical industries

with the continuous upgrading of the underlying technologies and capabilities of artificial intelligence, basic applications based on speech recognition, semantic understanding and computer vision have begun to be widely combined with various scenes of production and life, bringing many new intelligent possibilities to various vertical industries. For example, in the aspect of smart home, many consumer electronics hardware manufacturers and home appliance manufacturers have launched AI based assistants and home appliance products to realize the intelligent control and operation of various household devices. In the field of customer service, it is also widely used at present. Virtual assistants and customer service robots with language recognition and processing functions have begun to replace most of the manual customer service work. In the field of intelligent transportation and driverless driving, artificial intelligence is an indispensable technology. At present, Baidu's artificial intelligence system Apollo and application scenarios are driverless. At the same time, we also see that Alibaba's et brain is being applied in vertical scenarios such as smart city, industry, automobile, finance, home and new retail

operators' AI applications have both endogenous needs and exogenous needs

in the future, the network of telecom operators will become increasingly complex. 2g/3g/4g/5g symbiosis and multi domain coexistence will make the complexity of network operation and maintenance show a geometric growth. In the 5g era, AI will also become a native capability, making 5g networks flexible, flexible and intelligent. At the same time, operators also need to build various services and applications based on internal data, computing power and business scenarios, so as to promote AI liquidity and enhance business competitiveness. It is suggested that telecom operators should accelerate the introduction of AI technology and the creation of AI platform, and constantly improve the AI capability and product coverage level. In the user side, we should actively layout AI portals, including intelligent terminals, intelligent customs and other products. On the network side, we should strive to build a network operation and maintenance brain and improve the intelligent operation and maintenance ability for the next generation of networks. On the application side, we should pay attention to scenario driven, explore capacity open operation and ecological cooperation, actively cut into the vertical fields such as smart home and smart city, and promote the realization of AI

[source: China Academy of ICT]