Building tomorrow’s IoT applications using three layer model

29 December, 2016
Building tomorrow’s IoT applications using three layer model

Patrick Steiner is Lead Architect at Red Hat.

The Internet of Things is changing business IT and holds great potential for companies. By analysing data from networked devices, they can automate their business processes, increase productivity and lower costs. This requires a scalable, reliable and secure IT infrastructure, which should be based on standardised components and protocols and operate on three different layers: the device layer, the gateway layer and the datacentre layer.

The Internet of Things networks intelligent devices of all kinds, such as sensors, mobile devices, machines or vehicles with each other and with the cloud. The analysis of IoT data offers great opportunities to companies, they can make decisions faster, optimise business processes or develop new applications or even business models.

In this way, the Internet of Things impacts nearly every field, energy, health, transportation, retail, hospitality, manufacturing and financial services.

This opens up a broad spectrum of new potential applications, ranging from intelligent building technology, automated lighting, energy management, intelligent manufacturing systems, optimised solutions for inventories, logistics and supply chain management to remote monitoring of vital patient data.

The size and public nature of the Internet of Things, however, also involves technological challenges. Network and system architects have to optimise the IT infrastructure in order to meet the demanding requirements of IoT in terms of scalability, reliability and security.

The Internet of Things poses entirely new challenges for scalability. In the IDC study, Worldwide Internet of Things 2013–2020 Forecast, the market researcher states that more than 220 billion devices will be connected via Internet of Things by 2020. A single intelligent system could collect and analyse billions of data objects from millions of different endpoints. This will place unprecedented demand on processor performance, storage and networks.

IoT-based applications and automated business processes also place higher demands on availability of the system. Many intelligent systems are used for mission-critical applications, and system failures can lead to lowered productivity, dissatisfied customers or drop in sales. The same is true for emergency services, medical applications and monitoring solutions. In these cases, a system failure can endanger property, the environment, people’s health or even lives.

Distributed IoT solutions create large-scale security challenges, since the systems are networked over the Internet and use processing capacity and storage resources from the cloud. That is why companies need to expand their security infrastructure to efficiently protect themselves from data loss, theft and ever more sophisticated denial-of-service attacks.

This infrastructure includes comprehensive authentication, authorising and auditing features. These build confidence, regulate access to resources and ensure compliance with legal guidelines and company policies. Companies should use powerful encryption methods in order to protect their intellectual property and customer data from theft.

Layer model

Intelligent IT solutions meet the requirements of IoT systems in terms of scalability, reliability and security. The solutions are based on a hierarchical model with a device layer, control layer and datacentre or cloud layer. The solutions also use standardised protocols and components.

The device layer covers the high number of intelligent devices, including mobile devices, wearables, sensors, control devices and autonomous machines and appliances. Communication between the devices and the control points is based on standard network protocols, either cable or wireless. Open messaging standards are also used for routing raw data and exchanging control system information. The device layer also includes gateways that allow for interoperability with older and proprietary devices.

The gateway layer serves as a link between the devices and datacentre or the cloud. It collects and saves data from the devices and sends them to the datacentre. Conversely, it also sends control information to the devices, all based on open messaging standards. The gateway layer also serves as a cache for data required for tactical analysis or regulatory standards.

In addition, the gateway layer plays a central role for application of business rules management systems, since it saves real-time data monitored by business rules management in order to speed up processes. This layer recognises patterns in the data and functions according to predefined business rules. In addition, it distributes workloads and automates routine services for applications such as routing, transforming, distributing and aggregating data.

The datacentre and cloud layer is responsible for operating applications and services, analysing data and adapting business rules based on historical trends. It stores data for long-term analysis, contains most user and management interfaces and offers a virtual environment for complex tasks, distributed computing and business analytics functions. Furthermore, aggregated data from the control layer is collected and examined in this layer and business rules are distributed downstream.

One alternative to the three-layer model outlined here is the two-layer model, in which devices are connected directly to the datacentres or the cloud. This model is ideally suited to consumer applications that require less bandwidth and that do not need a gateway layer for the distribution of workloads.

Flexibility

The multi-layer model and open standards allow systems to utilise all the advantages of the Internet and the cloud, scalability, reliability, security and flexibility. By using standard interfaces, companies gain a variety of protocols and products and avoid becoming dependent on any single manufacturer. They can utilise different components from different providers and freely combine these into one system that perfectly meets their requirements.

Scalability

The distributed model allows for maximum scalability and lowers costs. The intermediate control layer reduces the load on datacentres and the cloud by processing raw data, carrying out business logic and communicating with devices. Since it also reduces device demands on processor performance and storage, more simple, cost-effective devices can be used. The elements of the control layer can be geographically distributed in order to maximise performance, uniform distribution of computing, network and storage resources and cost-effectiveness, lower costs for network bandwidth.

Reliability

The hierarchal topology is robust on every layer and between layers. Architects can implement redundant elements and domains in order to eliminate single points of failure and deliver high quality service. A system with reliable network, computing and storage components ensures continual availability of mission-critical applications.

Security

The multi-layer model offers multiple levels of security. Architects can implement different security measures at the individual levels, for example encryption, secure authentication, authorisation, and auditing, as well as patch and configuration management, in order to protect the system from attacks and threats.

This type of flexible, high-performance system can master all the challenges that the Internet of Things poses with its millions of networked devices and huge data volume. As a result, companies can transform raw data into valuable information and create the conditions for long-term success in their business.


Key takeaways

  • Solutions are based on a hierarchical model with device layer, control layer, datacentre or cloud layer
  • The device layer includes intelligent devices like mobile devices, wearables, sensors, control devices, autonomous machines, appliances
  • The gateway layer serves as a link between devices and datacentre or the cloud
  • The gateway layer plays a central role for application of business rules management systems, since it saves real-time data monitored by business rules management in order to speed up processes
  • The datacentre and cloud layer is responsible for operating applications and services, analysing data and adapting business rules based on historical trends
  • By using standard interfaces companies gain a variety of protocols and products and avoid becoming dependent on any single manufacturer
  • The hierarchal topology is robust on every layer and between layers
  • Architects can implement redundant elements and domains in order to eliminate single points of failure and deliver high quality service

Red Hat’s Patrick Steiner explains how an open standards multi-layer application stack can meet requirements of IoT in terms of flexibility, scalability, reliability, security.


2016 Awards Banner

Latest Whitepapers

Whitepaper

Whitepaper

Whitepaper

Whitepaper


Gigamon Advert