(DO NOT USE) EP271V HT23 Internet of Things and Artificial Intelligence


Wireless sensor networks (WSNs) the essential infrastructure of the Internet of Things (IoT). WSNs are networks of wireless nodes equipped with sensing capabilities for a huge variety of applications, such as body monitoring, autonomous vehicles, healthcare, industrial automation, or smart grids. The focus of the course is on communication and data analysis protocols and algorithms for networking, signal processing and machine learning in WSNs. The course presents the essential design and performance analysis methods for networking and machine learning by WSNs.


Learning outcomes

After passing the course, the student should be able to

  • give an account of the central wireless network protocols for IoT system design
  • give an account of central machine-learning methods for wireless IoT systems
  • design machine-learning methods for wireless IoT systems
  • theoretically characterize performance for wireless communications protocols and machine-learning methods with distributed datasets

in order to

  • understand and explain which design options there are for a specific wireless communication system
  • understand and explain which design alternative is available for a specific machine-learning algorithm with distributed datasets
  • be able to give arguments for which type of performance should be prioritised for designing wireless IoT systems and machine-learning methods
  • understand and explain design alternative for machine learning for specific datasets distributed over a wireless IoT system.



Course main content

The course focuses on wireless networks and machine-learning methods for wireless Internet of Things (IoT). The course starts with an introduction of applications of wireless IoT. Thereafter, methods for wireless communications protocols with an emphasis on analytical performance analysis are treated. In the course, machine-learning algorithms that can be executed on wireless IoT systems are analysed, where data and computations are distributed. The interplay between wireless network and machine-learning is analysed based on theoretical methods.


  • Introduction
  • The wireless channel 
  • MAC
  • Routing
  • IoT architecture
  • IoT communication standards
  • Centralized Machine Learning
  • Distributed Machine Learning
  • Federated Learning
  • Summary


Algebra, Analysis and Probability theory, including documented proficiency in English corresponding to English B

Recommended prerequisites

  • Knowledge in one variable calculus, 6 higher education credits, equivalent to completed course SF1625/SF1673/SF1685.
  • Knowledge in computer communication, 6 higher education credits, equivalent to completed course IK1203/EP1100.
  • Knowledge in probability theory, 6 higher education credits, equivalent to completed course SF1900-SF1935.
  • Knowledge in signals and systems, 6 higher education credits, equivalent to completed course EQ1110/EQ1120.
  • The upper secondary course English B/6.


The following books are only for reference. The course's book will be distributed in pdf

  • G. J. Pottie and W.J. Kaiser, “Principles of Embedded Networked Systems Design” Cambridge, 2005
  • W. Dargie and C. Poellabauer, “Fundamentals of Wireless Sensor Networks”, Wiley, 2010


  • NL1 - Assignment, 1.0, grade scale: P, F
  • INL2 - Assignment, 1.0, grade scale: P, F
  • INL3 - Assignment, 1.0, grade scale: P, F
  • TEN1 - Examination, 4.5, grade scale: A, B, C, D, E, FX, F

The problems of the exam will contain theoretical parts.

Requirements for final grade

4.5 points based on written exam and 3 points on homework assignments.

Offered by

EES/Network and Systems Engineering


Carlo Fischione


Carlo Fischione <carlofi@kth.se>

Add-on studies

EP2200 Queuing Theory and Teletraffic Systems

EP2590 Wireless Networks


Course syllabus valid from: Fall 23
Examination information valid from: Fall 23

Course summary:

Date Details Due