IOT Solution Architecture Styles


Chandramouli Srinivasan

There are many ways to architect the Internet of Things implementations for enterprises. CIOs must consider security, privacy, cost, ease of access, agility and performance to determine the best architecture for each enterprise.

Enterprises will build and adapt their Internet of Things implementations based on a combination of these five main architecture styles:

  • Thing-centric. Things are smart on their own and store most of their data on-board. Things are self-sufficient and communicate to the Internet only for centralized coordination and analysis.
  • Gateway-centric. The gateway houses the application logic, stores data and communicates with the Internet for the things that are connected to it. Things don’t have to be as smart because the gateway provides these resources.
  • Smartphone-centric. The smartphone (or any mobile device) houses the application logic, stores data and communicates with the Internet for the things that are connected to it. Things don’t have to be as smart because the smartphone provides these resources.
  • Cloud-centric. The cloud will act as the central connection hub, power analytics, and provision data storage. Things don’t have to be as smart because the cloud will provide these resources
  • Enterprise-centric. Things are behind a firewall and are geographically collocated. There is little need to extend out to the external Internet.


Each architecture has its own advantages and disadvantages. These architectures are designed to be style models that most enterprises will want to combine according to their needs. The reason why the names of each of these architectures are appended with “centric” (for example, cloud-centric) is that we expect that most enterprises will not pursue a pure implementation. For example, an enterprise might favor a smartphone-centric architecture, but may still rely significantly on cloud resources.

Enterprise CIOs and IT leaders should use these steps as a way of thinking about how to implement these architectures:

  1. Find the architectures that fit your use cases. Use the criteria in the Choosing the Right IoT Architectures section. Expect to have different use cases that require different architectures within the same enterprise.
  2. Choose or build an IoT platform that can support these chosen architectures — (ideally, all architectures, even the ones you won’t adopt immediately).
  3. Consider emerging technologies that may eliminate the advantages and disadvantages of an architecture style. For example, high-performance messaging protocols (for example, Data Distribution Service remove the latency in the cloud to provide real-time communications as if the machines were locally close. The cost of computing, storage, and communications will also be an emerging factor. For example, a decreasing cost of hardware against a rising cost of communications would influence an enterprise toward a thing- or gateway-centric style, as opposed to a cloud-centric style.

Choosing the Right IoT Architectures by Prioritizing Constraints

To properly evaluate which architecture styles fit best, enterprise CIOs and IT leaders should consider the following criteria. There is no right answer. Often, what is perceived as an advantage in some situations (for example, using cloud resources to remove cost and complexity from things) is actually a disadvantage in other situations (for example, connecting to the cloud is problematic or less secure).

The Priority constraints that needs to managed are

  1. The cost of hardware, software, and data.
  2. Connectivity & technical requirements based on reliability and quality of service; Real-Time Performance.
  3. Data and Security.
  4. Users and Implementations complexity.