Группа авторов

Artificial Intelligent Techniques for Wireless Communication and Networking


Скачать книгу

Although each sector is individually reshaping and allowing for unique ideas, 5G and AI integration will be truly revolutionary. In reality this integration is central to our small sensor edge definition, which includes in-home computing, edge cloud and 5G, to create an integrated intelligent system and service networking tool. Substantial amounting data, be it via on-device AI or gradual edge cloud storage over moderate 5G, can be distributed close to its source. Online AI is important to process data closer to the source because it provides efficient and effective output such as security, consistency and reliability as well as support for knowledge in scale [22].

      The topic has several introductions. They all serve a different purpose and offer a different perspective on these rapidly changing 5G communications. The goal of this review is to provide the reader with 5G usable artificial intelligence and to assist mobile users. The main contributions of this work are

      1 Provided a complete groundwork of integrated services of 5G in AI and AI in 5G;

      2 Provided a clear road map of artificial intelligence and 5G in the industrial space;

      3 Described the role of artificial intelligence in the mobile networks along with research challenges.

      This survey is clearly differentiated from other recent surveys by the above listed points. The paper is structured as follows: Along with this detailed introduction, the complete study of AI in 5G and 5G in AI are reviewed in Section 2.2. Section 2.3 discusses artificial intelligence and 5G in the industrial space. Section 2.4 briefly describes the roles of AI in mobile networks followed by a conclusion in Section 2.5.

      5G would be able and stronger than the conditions characteristic to host many more smart devices, particularly when about 41.6 billion Internets of Things smart devices are to be used by 2025. This is especially critical and many want a secure cloud access for efficient movement, such as driverless car systems and industrial sensors [12].

      GSMA, the mobile network trading body, has identified wireless technology as providing three pillars of next-generation connectivity:

       enhanced mobile broadband (eMBB),

       ultra-reliable,

       low-latency communications (URLLC) and massive machine-type communications (mMTC).

Schematic illustration of the AI in next generation networks.

      Figure 2.3 AI in next generation networks.

      Multiple devices based on high-speed and low-speed bandwide will be needed for the supporting equipment, both require online analytics in real-time. The large number of information generated requires the combination of 5G and Artificial Intelligence (AI) (Figure 2.3) [6].

      2.2.1 5G Services in AI

      AI can exist at every angle in the cloud environments, multi clouds, and mobile network of the potential. We also see significant suppliers of Technological innovations, such as NVIDIA, making significant contributions in 5G-based networks in connectivity, the Internet of Things (IoT) and other edge environments [14, 20].

      2.2.1.1 Next-Generation Edge Convergence With AI Systems on Chip

      2.2.1.2 Massive Device Concurrency Replenishing AI Data Lakes in Real Time

      5G can handle up to one million competing boundary devices per square kilometer, which is larger than 4G networks in a magnitude order. The last scale would enable companies, in a shifting paradigm known as multiaccess leading computing, to regularly obtain large amounts of data from cellphones, sensors, heating systems and other mobile devices. As 5G networks start to overwhelm water data structures with new telephone data globally, AI application analysts and data scientists can create more advanced analytical data.

      2.2.1.3 Ultra-Fast, High-Volume Streaming for Low-Latency AI

      5G connection times are substantially less as 4G, as small as 1 ms vs. the 50 ms 4G feature. As a result, 5G has far higher transmissions and processing rates than 4G, which is 20 Gb/s or 5–12 Mb/s for 4G. The stronger relation between the bandwidth of the system and the 5G power amplifier comes from the capacity to simultaneously transmit several data streams between the ground station and the borders. These fun working methods allow 5G to serve AI DevOps pipelines staff in low latency, from data intake, planning, modeling, training in real-time streaming scenarios. In addition, when coupled with its much lower latency, the faster download speeds of 5G will allow analysts to obtain, clean and evaluate much more information in a much shorter period of time.

      2.2.2 AI Services in 5G

      2.2.2.1 Distributed AI

Bar graphs depict the service providers achieving benefits through AI.

      Figure 2.4 Service providers achieving benefits through AI.

      2.2.2.2 AI for IT Operations (AIOps)

      To ensure that 5G provides much faster, safer and more RF-efficient connectivity than before, AIOps will be necessary. For the virtual machines in the network and multi-cloud management suites which tackle 5G networks and applications from one end to the next, AIOps technologies will have to be central. At least from the end-to-end through 5G ecosystems, AIOps drives consistent quality of service. AI-based controls can ensure continuous and accurate configuration of RF channels and other network infrastructures to support improvements in service quality, traffic patterns, and application tasks. They also encourage reliable alarm control, installation and healing and the optimization of the subscriber’s interface.

      2.2.2.3 Network Slicing