Othman Blog

Appearance

Introduction to IoT - Lecture 5: IoT Communication Models and Levels ​

IoT Communication Models ​

IoT communication models define how devices in a network exchange data. Common models include:

  • Request-Response
  • Publish-Subscribe
  • Push-Pull
  • Exclusive Pair

Request-Response Communication Model ​

  • In this model, the client sends a request to the server, and the server processes it and sends back a response.
  • Common in REST APIs and web-based applications.
  • Server fetches data, prepares a response, and returns it to the client.

Publish-Subscribe Communication Model ​

  • Involves publishers, brokers, and subscribers.
  • Publishers send data to topics managed by a broker.
  • Subscribers subscribe to these topics and receive data when published.
  • Publisher and subscriber are decoupled.

Push-Pull Communication Model ​

  • Producers push data to queues.
  • Consumers pull data from these queues.
  • Producers and consumers are decoupled.
  • Queues buffer messages and handle mismatched data rates.

Exclusive Pair Communication Model ​

  • Bidirectional, full-duplex model with a persistent connection between client and server.
  • Once connected, both ends can exchange messages.
  • Connection remains open until explicitly closed.

IoT Levels & Deployment Templates ​

IoT systems generally consist of:

  • Device
  • Resource
  • Controller Service
  • Database
  • Web Service
  • Analysis Component
  • Application

IoT Level-1 System ​

  • Single device/node handles sensing, storage, analysis, and application.
  • Ideal for low-cost, low-complexity systems with limited data.

IoT Level-2 System ​

  • Single device performs sensing and local analysis.
  • Data is stored in the cloud, and the application is cloud-based.
  • Suitable for systems with large data but non-intensive analysis.

IoT Level-3 System ​

  • Single device with cloud-based storage, analysis, and application.
  • Designed for large data and intensive computational needs.

IoT Level-4 System ​

  • Multiple nodes perform local analysis.
  • Data is sent to the cloud; observers (local/cloud) can subscribe to updates.
  • Suitable for complex, multi-node systems with heavy data and analysis needs.

IoT Level-5 System ​

  • Multiple end nodes + one coordinator node.
  • End nodes sense data; coordinator aggregates and sends it to the cloud.
  • Ideal for wireless sensor networks with complex analytics.

IoT Level-6 System ​

  • Multiple independent end nodes send data directly to the cloud.
  • Cloud handles storage, analytics, and visualization.
  • Central controller manages node status and control commands.

Web Services ​

  • A web service is a software interface accessible over the internet.
  • Uses standard formats (like XML or JSON) to send/receive API data.

Amazon Web Services (AWS) ​

  • A suite of cloud computing services by Amazon.
  • AWS is spread across 9 global regions, each with multiple Availability Zones.
  • Offers services for computing, storage, databases, networking, and more.

Website

Amazon Web Services for IoT ​

AWS provides various services for implementing scalable IoT systems:

  • EC2
  • AutoScaling
  • S3
  • RDS
  • DynamoDB

Amazon EC2 ​

  • Infrastructure-as-a-Service (IaaS) for scalable virtual computing.
  • Allows launching virtual machines for IoT apps.

EC2 Python Example ​

python
# Python program for launching an EC2 instance
import boto.ec2
from time import sleep

ACCESS_KEY="<center access key>"
SECRET_KEY="<center secret key>"

REGION="us-east-1"
AMI_ID = "ami-d0f5bfb9"
EC2_KEY_HANDLE = "<enter key handle>"
INSTANCE_TYPE = "t1.micro"
SECGROUP_HANDLE = "default"

conn = boto.ec2.connect_to_region(
  REGION, aws_access_key_id=ACCESS_KEY,
  aws_secret_access_key=SECRET_KEY
)

reservation = conn.run_instances(
  image_id=AMI_ID, key_name=EC2_KEY_HANDLE,
  instance_type=INSTANCE_TYPE,
  security_groups=[SECGROUP_HANDLE, ]
)

Amazon S3 ​

  • Scalable cloud storage for raw IoT data.
  • Stores sensor data, logs, multimedia, etc.
  • Reliable and redundant infrastructure.

Amazon Auto Scaling ​

  • Automatically adjusts the number of EC2 instances.
  • Vertical scaling: Upgrading resources.
  • Horizontal scaling: Adding more nodes.
  • Useful for dynamic scaling of IoT platforms.

Amazon RDS ​

  • Cloud-based relational database service (MySQL, Oracle, etc.).
  • Provides scalable, managed SQL databases.
  • Ideal for structured IoT data storage.

Amazon DynamoDB ​

  • NoSQL, fully managed, high-performance database.
  • Designed to handle massive data volume and real-time access.
  • Great for unstructured IoT data.

AWS Core Services by Category ​

Compute ​

  • Amazon EC2
  • AWS Lambda
  • Auto Scaling
  • AWS Elastic Beanstalk
  • Amazon ECS / ECR
  • Amazon Lightsail
  • AWS Batch

Networking ​

  • Amazon VPC
  • Amazon Route 53
  • AWS Direct Connect
  • Elastic Load Balancing

Storage ​

  • Amazon S3
  • Amazon EBS
  • Amazon CloudFront
  • Amazon Glacier
  • Amazon EFS
  • AWS Snowball
  • Storage Gateway
  • AWS Snowmobile

Database ​

  • Amazon RDS
  • Amazon DynamoDB
  • Amazon Redshift
  • AWS Database Migration Service
  • Amazon ElastiCache