Describing an IoT network physically and looking beyond the three-layer structure.
Fleshing out IoT system design with additional processes and new structures.
How IoT is impacted at the device and hardware level.
IoT system design concerns network-enabled devices and the connections between them.
In many ways, IoT is a well-established field, but there are still exciting developments percolating down to designers as materials and technologies take new leaps. As overall device connectivity continues to burgeon, space for new and improved systems of design arises in conjunction. IoT system design will push for greater performance, higher speeds, reduced cost/power consumption, and improved insights for greater efficiency in practically every industry imaginable.
The Size and Shape of IoT Networks
IoT systems operate across multiple layers, with a high degree of flexibility in constructing the communication network between objects using various standards and protocols. Given the wide amount of configurations available to IoT systems, it’s valuable by beginning with some analysis as to what variables can impact the architecture. Design is influenced by two independent factors:
- The scale of the network indicates the size of the area within which it communicates. A local device that exists on the periphery of network architecture could be considered small-scale, as it has little connection with other nearby nodes. Meanwhile, a smart building could be considered large-scale (depending on the scope), as its domain includes many IoT devices and communications between them.
- How tightly or loosely distributed IoT devices are within an area indicates the density. Density is also a strong reflection of the bandwidth constraints and latency on the network.
IoT is by all accounts a wide umbrella term, and defining the parameters of the network helps to shape the infrastructure of the design. Superficially, IoT can be defined across three layers, those being sensors/actuators, a network layer, and an application layer; these three layers capture the core functionality, being perception, communication, and interactivity, respectively. To further maximize this architecture, IoT systems need to embrace some additional criteria that extend beyond the base layers:
- Software interfaces - At a high level, languages need to operate with generality to promote ease of use as well as avoid translation conflicts. Furthermore, standardization between languages drives greater accessibility, unlocking additional configurations for system design.
- Location - The concept of absolute and proximal location is important in any network, but takes on something of new meaning with IoT due to increases in network-capable portable electronics within recent years. Sensors probe their local environment, and combined with other devices, form a dynamic mapping of particular characteristics germane to the purpose of the network. The location of these sensors may or may not be dynamic as well, and designers can better represent the parameter under consideration by accounting for the future positioning of these devices.
- Data - Data streams pass through multiple stages of the design when traveling from sensor to interface. To bridge different formats, data will require standardization. The network framework will also need to encompass the collection/transmission of data, whether that’s continuous, discrete, or event-driven.
Expanding IoT System Design at the Edges
Sometimes the three-layer model of IoT architecture is lacking in description of the actual processes occurring within the system. To that end, some designers focus on breaking down the nebulous “network” layer into more appropriate transport and processing layers:
- The transport layer is the intermediate between sensory data acquisition and its processing step. Using a myriad of different protocols, data is sent from the device level to a central hub. Depending on the needs of the system, e.g., range, power consumption, latency, etc., certain network protocols may be favored, disfavored, or entirely ill-suited for adequate performance.
- The processing layer represents the enmeshment of big data systems with IoT. Powerful computing that literally and figuratively weighs down IoT devices can be utilized server-side and passed back to devices via the connecting network. Data can also be stored to study historical trends and help predict future network behavior.
Long-term storage carries its challenges, however. Data, even in aggregate, need to either be sensitively handled to protect user identities or scrubbed of any personal information. With this need, some architectures also consider the importance of the overall security of the system as its own layer that provides valuable encryption and decryption of valuable proprietary and potentially sensitive data.
More recently, a new paradigm is growing in IoT system design. Termed edge computing (more generally, it’s known as fog computing, i.e., a decentralized cloud), this solution aims to push more of the processing capabilities to devices. There are a few advantages to this method, primarily that information sent across the network can be smaller in size, allowing for more flexibility in protocol deployment. In addition, edge computing can actualize the data in a hybridized manner, acting more rapidly on preprocessed data relevant to the sensors and actuators while continuing to pass information to the cloud for more comprehensive analysis. On the downside, edge computing is vastly limited in storage and processing capabilities on its own, and extreme low-power networks may find the energy cost of preprocessing detrimental relative to any gains in speed or responsiveness.
Start Planning IoT at the Hardware Level
Support for IoT system design must be incorporated at all levels of the development lifecycle; while the focus on IoT generally weighs network architecture greatly, equal importance should be placed on the hardware itself. Faster speeds and rise/fall times for components require careful planning during board layout. This extends to the materials used in board fabrication, as shops and manufacturers can work alongside design teams to furnish low-loss materials especially conducive to high-speed design.
For a comprehensive solution to board design, layout, simulation, and prototyping, Cadence’s suite of PCB design and analysis software offers teams a fully encapsulated service for board design. With OrCAD PCB Designer, engineers and layout designers can quickly develop a board for both simulation and manufacturing using powerful, yet easy-to-use tools.
Leading electronics providers rely on Cadence products to optimize power, space, and energy needs for a wide variety of market applications. To learn more about our innovative solutions, talk to our team of experts or subscribe to our YouTube channel.