Design Principles for Connected Devices

Lecture notes
March 2024

1 Introduction
Data transfer from sensors creates a stack for communication to an application or service, following a specified
protocol at each stage. Data refers to device data received for monitoring or actions. Data stack includes layers
like Application, Presentation, Session, Transport, Network, Data-link, and Physical. Data-adaptation actions can
be related to privacy, security, consolidation, aggregation, compaction, and fusion. An action can be a gateway
action using different protocols.

Following are the key terms which need to be understood to learn the design principles of connected devices for
IoTs:
1. Layer refers to a stage during a set of actions at which the action is taken as per a specific protocol or method.
2. Physical layer refers to a layer at transmitting-node or at the receiving node for the data bits.
3. Application layer refers to a layer for transmitting or receiving the data bits of an application.
4. Level refers to a stage from the lowest to the highest.
5. Domain refers to a set of software, layers or levels having specific applications and capabilities.
6. Gateway refers to software for connecting two application layers, one at the sender and the other at the
receiver.
7. IP stands for Internet Protocol version 6 (IPv6) or Internet Protocol version 4 (IPv4) for the network layer.
8. Packet means packaged data-stack which routes over the network.
9. Protocol Data Unit (PDU) is a unit of data which is specified in a protocol of a given layer which transfers
from one layer to another.
10. Maximum Transmission Unit (MTU) is the largest size frame or packet or segment specified in octets (1 octet
= 1 byte = 8 bits) that can be sent in a packet or frame-based network such as the Internet.
11. End-point device or node denotes the one that provides connectivity to a coordinator or router.
12. Coordinator denotes the one that connects to a number of end-points as well as routers in a star topology
and forwards the data stack from one attached end point/router to another.
13. Router refers to a device or node capable of storing paths to each destination to which it has logical links.

2 IoT/M2M Systems, Layers and Designs Standardization

Internet Engineering Task Force (IETF), an international body initiated actions for addressing and working on the
recommendations for the engineering specifications for the Internet of Things. IETF suggests the specifications for
the layers, and the engineering aspects for the IoT communication, networks and applications.
International Telecommunication Union for Telecommunication (ITU-T) suggested a reference model for IoT do-
main, network and transport capabilities for the IoT services and the applications at the application and application-
support layers.
European Telecommunication Standards Institute (ETSI) initiated the development of a set of standards for
the network, and devices and gateway domains for the communication between machines (M2M). ETSI proposed
high-level architecture for applications and service capabilities.

1
2.1 IoT Layers and Deployment Templates
An IoT system comprises of the following components:
1. Device: An IoT device allows identification, remote sensing, actuating and remote monitoring capabilities.
2. Resource: Resources are software components on the IoT device for accessing, processing, and storing sensor
information, or controlling actuators connected to the device.
3. Controller Service: Controller service is a native service that runs on the device and interacts with the web
services.
4. Web Service: Web services serve as a link between the IoT device, application, database and analysis com-
ponents. Web service can be either implemented using HTTP and REST principles (REST service) or using
WebSocket protocol (WebSocket service).
IoT Level-1 : A Level-1 IoT system has a single node/device that performs sensing and/or actuation, stores
data, performs analysis and hosts the application. Level-1 IoT systems are suitable for modeling low-cost and
low-complexity solutions where the data involved is not big and the analysis requirements are not computationally
intensive.

IoT Level-2 : A Level-2 IoT system has a single node that performs sensing and/or actuation and local analysis.
Data is stored in the cloud and application is usually cloud-based. Level-2 IoT systems are suitable for solutions
where the data involved is big, however, the primary analysis requirement is not computationally intensive and can
be done locally itself.

IoT Level-3 : A Level-3 IoT system has a single node. Data is stored and analyzed in the cloud and application
is cloud- based. Level-3 IoT systems are suitable for solutions where the data involved is big and the analysis
requirements are computationally intensive.

IoT Level-4 : A Level-4 IoT system has multiple nodes that perform local analysis. Data is stored in the cloud
and application is cloud-based. Level-4 contains local and cloud- based observer nodes which can subscribe to and
receive information collected in the cloud from IoT devices. Level-4 IoT systems are suitable for solutions where
multiple nodes are required, the data involved is big and the analysis requirements are computationally intensive.

IoT Level-5 : A level-5 IoT system has multiple end nodes and one coordinator node. The end nodes that
perform sensing and/or actuation. Coordinator node collects data from the end nodes and sends to the cloud.
Data is stored and analyzed in the cloud and application is cloud-based. Level-5 IoT systems are suitable for
solutions based on wireless sensor networks, in which the data involved is big and the analysis requirements are
computationally intensive.

IoT Level-6 : A level-6 IoT system has multiple independent end nodes that perform sensing and/or actuation
and send data to the cloud. Data is stored in the cloud and application is cloud-based. The analytics component
analyzes the data and stores the results in the cloud database. The results are visualized with the cloud-based
application. The centralized controller is aware of the status of all the end nodes and sends control commands to
the node.

2.2 Modified OSI Model for the IoT/M2M Systems

OSI protocols are a family of information exchange standards developed by the ISO and ITU-T. The seven-layer
OSI model is a standard model for designing communication networks. Various models for data interchange modify
the model for simplicity. The IETF suggests modifications for IoT/M2M. Data communicates from device to
application end, with each layer processing received data and creating a new data stack. In-between layers process
data from the bottom functional layer to the top layer.

Gather + Enrich + Stream + (Manage + Acquire + Organise +Analyse) = IoT Applications

2
Figure 1: Seven-layer generalised OSI model (on left) and IETF six layer modified OSI model for IoT/M2M (in
the middle)

The application layer 6 now includes new applications and services, with the application-support layer 5 using
protocols like CoAP for network communication. The application-support layer includes processes for data man-
agement, acquisition, organization, and analysis. Changes are also made at data-link layer 2 and physical layer
1, with new layers including data-adaptation and physical cum data-link. The data-adaptation layer includes a
gateway for network and web communication.

2.3 ETSI M2M Domains and High-level Capabilities

A domain specifies the functional areas. High-level architecture means architecture for functional and structural
views. The ETSI network domain has six capabilities and functions: (1) M2M Capabilities, (2) M2M service
capabilities, (3) M2M management functions, (4) Network management functions, (5) CoRE network(3G and IP
Networks), and (6) Access network(LPWAN, WiFi).

The ETSI device and gateway domain has the following functional units:

1. Gateway between M2M area network, and CoRE and access network, possessing M2M service capabilities
and applications.
2. M2M area network (for example, Bluetooth, ZigBee NFC, PAN, LAN).
3. M2M Devices.

3 Communication Technologies
The model’s physical cum data-link layer comprises a local area network/personal area network, which uses either
wireless or wired communication technologies for IoT or M2M device connectivity. Figure below illustrates con-
nected devices’ data communication to the local network and gateway.
Figure 3. displays the number of devices in an IoT or M2M devices network, including local area networks.
Connectivity methods include RF, Bluetooth Smart Energy, ZigBee IP, ZigBee NAN, NFC, 6LoWPAN, or mobile.
Tens of bytes communicate at an instance between the device and local devices network.

Standard IoT Communication Technologies include : Bluetooth, ZigBee, BLE, WiFi, Z-wave, RFID, Cellular,
SigFox, Ethernet, NFC, LPWAN, LoRaWAN(SigFox, LoRa).

IoT Protocols : *discussed in Chapter 1.

3
Figure 2: ETSI M2M domain architecture and its high-level capabilities, and its correspondences with six layers of
modified OSI and four layers of ITU-T reference model.

4 Data Enrichment, Data Consolidation and Device Management at

Gateway
A gateway at the data-adaptation layer performs various functions such as data privacy, security, enrichment,
consolidation, transformation, and device management. IoT or M2M Gateway consists of data enrichment, con-
solidation, device management, and communication frameworks. The device layer, the lowest layer in the ITU-T
reference model, has device and gateway capabilities. The gateway connects M2M area networks to CoRE and
access networks, enabling devices to communicate and network with the web using message transport protocols
and web communication protocols. It also includes data management and consolidation.

4.1 Data Management and Consolidation Gateway

Gateway includes the provisions for one or more of the following functions: transcoding and data management.
Following are data management and consolidation functions:

Transcoding : Transcoding is the process of converting data into the required formats for an IoT device, using
software to adapt and change protocols, formats, or codes, ensuring their acceptable representation at the server.

Privacy : Privacy is an aspect of data management and must be remembered while designing an application.
The design should ensure privacy by ensuring that the data at the receiving end is considered anonymous from an
individual or company.

Secure Data access : Access to data needs to be secure. The design ensures the authentication of a request for
data and authorisation for accessing a response or service.

Data Gathering and Enrichment : IoT/M2M applications involve actions such as data-gathering (acquisition),
validation, storage, processing, reminiscence (retention) and analysis. Data gathering refers to data acquisition
from the devices/devices network. Data enrichment refers to adding value, security and usability of the data.

Data Dissemination : Consider the following three steps for data enrichment before the data disseminates to the
network as aggregation, compaction and fusion.

4
Figure 3: Connected devices 1st to ith connected to the local network and gateway using the WPAN or LPWAN
network protocols.

Data Source and Data Destination : ID: Each device and each device resource is assigned an ID for specifying
the data of source and a separate ID for data destination. Address: Header fields add the destination address.

Figure 4: IoT or M2M gateway consisting of data enrichment and consolidation, device management and commu-
nication frameworks at the adaptation layer.

4.2 Device-management Gateway

Device Management (DM) involves provisioning unique device IDs, activating, configuring, registering, deregis-
tering, attaching, and detaching devices, and accepting subscriptions for resources. It also includes device fault
management and guidelines for dealing with device faults. In IoT/M2M applications, a DM server interacts with
devices through a gateway, managing device parameters and settings. In low-power loss environments, devices
communicate with a gateway instead.