Department of Computer Science and Technology

Course pages 2019–20

Mobile and Sensor Systems

Principal lecturers: Prof Cecilia Mascolo, Dr Robert Harle
Taken by: Part II CST 50%, Part II CST 75%
Past exam questions

No. of lectures: 12
Suggested hours of supervisions: 3
Prerequisite courses: Operating Systems, Concurrent and Distributed Systems

Aims

This course will cover topics in the areas of mobile systems and communications, and sensor systems and sensor networking. It aims to help students develop and understand the additional complexity introduced by mobility and sensing, including energy constraints, communication in dynamic networks and handling measurement errors. The course will be using various applications to exemplify concepts.

Lectures

  • Introduction to Mobile Systems. MAC Layer concepts. Examples of mobile systems, differences with non mobile systems. Introduction to MAC layer protocols of wireless and mobile systems.

  • Mobile Infrastructure Communication and Opportunistic Networking. Description of common communication architectures and protocols for mobile and introduction to models of opportunistic networking.

  • Introduction to Sensor Systems, MAC Layer concepts and Internet of Things. Sensor systems challenges and applications. Concepts related to duty cycling and energy preservation protocols. Emerging concepts and communications protocols for the Internet of Things.

  • Sensor Systems Routing Protocols. Communication protocols, data aggregation and dissemination in sensor networks.

  • Machine Learning for Mobile Systems and Sensor Data Mobile and wearable sensing. Machine Learning on Sensor Data. On-device Machine Learning.

  • Mobile Sensing: Systems Considerations Considerations of energy preservation. Local computation vs cloud computation. Storage and Latency.

  • Privacy in Mobile and Sensor Systems. Concepts of mobile and sensor systems privacy. Privacy and sensor based activity inference. Mobility prediction and privacy.

  • Localization I. Basic concepts. Proximity, trilateration, triangulation, ToA, TDoA. Examples including ultrasonic, mobile networks, UWB.

  • Localization II. GNSS, RSS fingerprinting/WiFi positioning.

  • Tracking. Kalman filtering, particle filtering, Pedestrian Dead-Reckoning as an example.

  • Mobile/Wearable health sensing. Health as a key driver for wearables. High availability and low fidelity (wearable) vs low availability and high fidelity (clinical). Sensing challenges (low power, noisy, poor contact). PPG case study. False positives and the base rate fallacy.

  • Robots and Drones Concepts related to control, communication and coordination of robotic systems.

Objectives

On completing the course, students should be able to

  • describe similarities and differences between standard distributed systems and mobile and sensor systems;

  • explain the fundamental tradeoffs related to energy limitations and communication needs in these systems;

  • argue for and against different mobile and sensor systems architectures and protocols.

  • Understand typical error sources for sensing and be aware of techniques to minimise them.

  • put concepts into context of current applications of mobile and sensor systems as described in the course.

Recommended reading

The course is based mainly on research papers cited in each lecture. The following books, however, contain some of the more traditional concepts.

* Schiller, J. (2003). Mobile communications. Pearson (2nd ed.).
* Karl, H. & Willig, A. (2005). Protocols and architectures for wireless sensor networks. Wiley.
Agrawal, D. & Zheng, Q. (2006). Introduction to wireless and mobile systems. Thomson.