IoT Architecture

The fundamental components of the Internet of Things (IoT) consist of sensory devices, remote service invocation, communication networks, and context-aware processing of events. 

 

IoT aims to depict a cohesive network of intelligent entities, including both smart things and humans, who have the ability to interact with each other universally and ubiquitously, and can take charge of operating these entities when necessary.

Inter-connectivity among entities is an essential requirement in a distributed system, and the IoT serves as a prime example. 

 

An all-encompassing system architecture for the IoT must ensure the correct functioning of its components, with dependability being the most crucial design feature. Additionally, it should establish a connection between the physical and virtual domains. 

 

In order to accomplish this, it is necessary to carefully contemplate the architecture of failure recovery and scalability. Furthermore, because to the prevalence of smartphones, mobility and frequent changes in location have become essential aspects of IoT systems. 

 

Therefore, advanced designs must possess a certain degree of adaptability in order to effectively manage the dynamic interactions within the entire ecosystem. 

 

A number of research organizations have put up reference architectures for the Internet of Things (IoT). 

• The IoT-A project, referred to as the European Lighthouse Integrated Project, is dedicated to the creation and validation of an integrated Internet of Things network architecture and its supporting components. 
• The IoT-i project, which is connected to the previously described IoT-A project, aims to promote IoT solutions and identify requirements and interests in this field. IoT-i aims to accomplish strategic objectives, including the development of a unified strategic and technical vision for the IoT in Europe, which considers all fragmented sectors of the IoT domain comprehensively. Additionally, IoT-i seeks to contribute to the establishment of an economically sustainable and socially acceptable environment in Europe for IoT technologies and related research and development activities. 

 

The Figure illustrates a schematic of our enhanced iteration of a reference architecture for the Internet of Things. This architecture displays distinct service and presentation layers. The service layers encompass event processing, analytics, resource management, service discovery, message aggregation, and Enterprise Service Bus (ESB) services, which are built upon the communication and physical levels. API management, a crucial component for designing and distributing system services and web-based dashboards (or similar smartphone applications) for maintaining and accessing these APIs, is also incorporated into the architecture.

 

Given the significance of device management, the enforcement of security and privacy at several levels, and the capability to uniquely identify things and regulate their access, these components are prioritized separately in this design.

---

SOA-based Architecture

---

Service-oriented architecture (SOA) is crucial in the Internet of Things (IoT) for service providers and users as it ensures interoperability among heterogeneous devices. 

 

A generic SOA consists of four layers: Sensing, Network, Service, and Interfaces. 

 

These layers are designed to be loosely coupled and reused, allowing for easy maintenance and reliability in case of component failure. 

 

SOA has been extensively used in Wireless Sensor Networks (WSN) due to its modular design and abstraction level. 

 

It can enhance interoperability and scalability among IoT objects, simplifying user interaction with different layers and protocols. 

 

Additionally, SOA allows for the creation of diverse and complex services by composing different functions of the system, enhancing modular composability and addressing the heterogeneous nature of IoT.

---

API-Oriented Architecture

---

Service-oriented solutions traditionally use SOAP and RMI for describing, discovering, and calling services. However, due to overhead and complexity, Web APIs and Representational State Transfer (REST)-based methods have emerged as promising alternatives. 

 

These methods reduce resource requirements, especially for smart devices and sensors with limited resources, by replacing large XML files with lightweight data-exchange formats like JSON. Building APIs for IoT applications helps service providers attract more customers and enable multitenancy through security features like OAuth.

Research has proposed methods like Simurgh, which describes devices, sensors, and human services using web API notation and API definition languages. A two-phase discovery approach is also proposed to find sensors that provide desired services and match certain features. A service-broker layer, FOKUS, exposes APIs for shared access to the OpenMTC core.

Research suggests shifting from service delivery platforms (SDPs) to web-based platforms, which can reduce service discovery complexity and runtime overhead. 

 

Developers and business managers should focus on developing and sharing APIs early in their application development lifecycle to create an open-data environment for collaborative information gathering, sharing, and updating.

---

Happy Exploring!