The browser version you are accessing is too low. To provide better experience, it is recommended that you upgrade the browser toEdgeBrowserOr, it is recommended.GoogleBrowser

Internet of Things and Data Pipes

Release Date:2018-08-20  Author:By Zhang Xunniu, Wang Shaofei  Click:

We live in an era of highly developed information technologies. Various new technologies emerge one after another, constantly affecting our work and life. As the infrastructure of this era, the internet and mobile internet continuously integrate all sorts of emerging technologies, and are developing towards the internet of things (IoT), which undoubtedly excite us the most. IoT itself is not a new concept. It was first proposed in the 1990s, but developed slowly. However, its great goal of interconnecting everything has always been the development direction of various technologies. The IoT-based smart life, smart city, and smart earth are gradually becoming a reality. Nowadays, IoT has been recognized as the third wave of the development of the world’s information industry after the waves of computer and internet. IoT has become a competitive technological highland that technology giants want to occupy.
It can be seen that IoT has two core features: connections and data. Centered around connections and data, IoT widely integrates a large number of existing technologies, involving telecom, big data, artificial intelligence, data mining, cloud computing, automation, electronics, and materials. In terms of connections, IoT extends the functions and scope of traditional telecom networks to a wider physical world. In terms of data, IoT allows the access of a considerable variety of massive devices, greatly expanding the source channels of network information data. According to statistics, 20% of the world’s total amount of recently created, obtained, and copied data is from IoT, and is growing at the fastest pace. With gradually ubiquitous connections and richer data being converged, IoT is becoming a new driving force for the advancement of various technologies.
IoT is a complex giant system that can be divided into different parts in different dimensions. In the dimension of hierarchical network architecture and based on the data processing procedure, IoT is divided into three layers: the information perception layer for comprehensive information perception, the data transmission layer for reliable data transmission, and the data application layer for big data storage and intelligent processing.

Generation of IoT Data
The information perception layer or the nerve terminal of IoT, is key to IoT information data sources. This layer is connected to a large number of application devices that integrate a variety of information sensing and data collection components, including RFID readers, temperature sensors, humidity sensors, acoustic and photoelectric sensors, and cameras. These components allow this layer to have a certain level of intelligence for dynamically sensing context information of the devices themselves and their surroundings. This layer is the main producer of IoT data. The rich data generated by this layer is the foundation for IoT applications and also a prerequisite for continuous development of IoT and growing popularity of IoT applications.
Although the information perception layer is the main producer of IoT data, it does not have sufficient data storage and processing capability, but can process data locally and pretty simply. The data generated by this layer needs to first be transmitted to the data processing center through the network, then collected, stored, and analyzed, and finally exchanged and shared. Only through this procedure can the IoT data have its value. 

Transmission of IoT Data
IoT is an extension and expansion of traditional telecom networks, and also an integration of traditional network technologies. Meanwhile, the inherent characteristics of IoT also facilitate the emergence of many new telecom technologies.
In terms of data transmission, the biggest challenge lies in the device access network. The devices at the information perception layer of IoT are of many kinds and large difference. Some are very big, some are very small, some are located on both sides of a road with convenient transportation, some are deployed in a remote wasteland, some seldom send messages and their message transmission speeds are very slow, some frequently interact with others and send messages as fast as hares run when going into action, some consume a large amount of energy, and some consume a small amount. These diverse characteristics make it highly complex for the devices to access IoT. There is no unified network access mode that can fully meet IoT requirements for device access.
In general, wireless is the primary access mode of IoT devices, and wireline is their auxiliary access mode. In the wireless access mode, there are short-range wireless communication technologies such as Wi-Fi, Zigbee, and Bluetooth, as well as long-range wireless communication technologies such as 2G, 3G, 4G, LoRa, 5G (NB-IoT). Short-range wireless access is generally applicable to indoor devices, while long-range wireless access is more suitable for outdoor devices. Long-range wireless access technologies support low power consumption, medium and low bandwidth, large capacity, and long distance, meeting the access requirements of most IoT devices. LoRa and NB-IoT, two major wireless access modes, can meet these requirements. LoRa is led by the enterprise alliance, while NB-IoT is mainly promoted by telecom operators. With the advent of 5G, the wide deployment of mobile communication base stations with NB-IoT functions makes the competition between LoRa and NB-IoT increasingly fierce.
In terms of backbone network data transmission, IoT brings a significant impact on data traffic. IoT requires higher data bandwidth to accommodate more data being transmitted, and more flexible data routing and forwarding rules to intelligently forward data. With new network design ideas represented by software-defined networking (SDN), the traditional data communication network architecture is being reconstructed. New data communication technologies including quantum communication also provide new options for data transmission in the IoT era.
In terms of data communication protocol stack, the underlying physical layer protocol is determined by each physical network. Different physical networks use different underlying physical layer protocols, each of which operates in its own way. The traditional TCP/IP protocol stack still dominates all network layer protocols. At the transmission and application layers, the HTTP protocol for the application-oriented northbound interface is still in a dominant position, but for the equipment-oriented southbound interface, MQTT, COAP and other protocols designed for IoT can meet device requirements for low power consumption and low bandwidth transmission.
In terms of data communication security, the data security requirements of IoT involve every aspect at each layer. At present, there is no standard integrated solution. In the near future, it will be a general practice to use traditional HTTPS, TLS, and DTLS technologies to meet the security requirements of their respective applications.

Storage and Application of IoT Data
IoT-based applications are the driving force for sustainable IoT development. There are many simple and local IoT islanding applications, which involve simple data types and small amount of data, and can hardly be commercially used on a large scale and produce desired industrial effects. The influence is extremely limited.
The emergence of new data storage and processing technologies, such as big data storage, large data analysis, cloud computing, and artificial intelligence, meets IoT requirements for big data storage and intelligent processing and greatly accelerates the pace of IoT development.
In terms of big data storage, distributed cloud storage systems such as hadoop distributed file system (HDFS), distributed column-oriented storage system (HBase), Amazon S3 cloud storage, and Microsoft Azure storage can meet IoT requirements for large-scale data storage.
In terms of big data processing, big data processing frameworks such as MapReduce, Spark, and Storm can implement offline and real-time analysis of large-scale IoT data, explore more potential value in massive IoT data, and promote the launch of more IoT applications.
In terms of data exchange and sharing, related standardization organizations in the industry have been committed to specifying and standardizing IoT data models and service procedures to overcome the pain points of IoT application fragmentation and islanding. Relevant architectures and standards include the LWM2M architecture proposed by the Open Mobile Alliance (OMA), the oneM2M architecture provided by oneM2M, an international IoT standardization organization co-founded by multiple standards development organizations, and the IoT device standards proposed by the Open Connectivity Foundation (OCF) that contains Microsoft, Intel, Samsung, Qualcomm and CISCO. Putting forward these architectures and standards gives powerful impetus to the development of IoT.

ZTE's ThingxCloud IoT Platform
ThingxCloud is a new-generation IoT platform rolled out by ZTE. By carrying upper-layer applications, connecting lower-layer devices, and generating data on its own, this platform empowers IoT, facilitates ecosystem growth, and creates a new IoT co-building, sharing, and win-win mode.
Based on the advanced ICT PaaS platform developed by ZTE, ThingxCloud allows IoT applications to be deployed as micro-services. The platform supports service orchestration, dynamic scaling, and horizontal expansion, greatly facilitating IoT service deployment and system self-adaptation in different application scenarios. 
The ThingxCloud platform supports major global IoT standards and specifications such as LWM2M and OneM2M. For terminal devices, the platform supports multiple access protocols such as MQTT, COAP, and HTTP. Through SDK development packages, the ThingxCloud platform simplifies terminal access, so that any terminals can access the platform in any access mode. The platform opens data service API interfaces to application services, provides multiple external data services, and simplifies data interaction and its application development process.

In the future, the ThingxCloud platform will continuously integrate ZTE's advanced big data, artificial intelligence, and data mining platforms, to constantly increase its capabilities and push the development of innovative IoT applications.

[Abstract] IoT data, data pipes, data mining, cloud computing, Storage, ThingxCloud IoT platform, SDN