Page History

Goal of this topic is to decentralise/distribute access control mechanisms across a local IoT network.

Currently, Access Control in IoT environment is rather centralised, meaning that Access Control gateways aren't neccesarily near IoT devices, which leaves a big attack plane for hackers.

In this topic, the ultimate goal will be to bring Access Control mechanisms as close as possible to sensors in an IoT system. In order to achieve that, multiple challeges arise - access rights synchronisation, building trust between modules in a distributed system, etc.

Your solution will be developed and integrated as part of COSYLab Smart Home framework.

Note: This topic is only available as P1/P2 or Master's Thesis.

"Smartify Access Control in the IoT environment by analysing user behaviour patterns"

Make use of the collected data from sensors in a Smart-Home and define events that could be of interest to assigning Access Rights to the users in the system.

This topic requires analysis of the user behavior patterns(waking up time, leaving to work, bed time) in the IoT environment, description of the patterns through system events and application of these events to the composition of access rights.

This solution will be integrated with an already existing Smart-Home management framework, present on COSY:Lab.

Note: This topic is only available as P1/P2 or Master's Thesis.

Transmissions over radio may suffer interference both from out-of-band and in-band transmissions. Your task is to choose modulation schemes and develop models for interference for them, both within the same modulation scheme and across schemes. Target metrics focus on the demodulated signal and thus the disturbance that interferences cause in the demodulator: effects on the spectral content, noise floor, transmitter/receiver synchronization, etc. You evaluate your models both in theory and in practice, i.e. mathematically and through an implementation in GNU Radio.

The Kibana analysis software provides convenient ways to analyze datasets, but it lacks features that help with modeling tasks to (e.g.) describe arrival and service processes from the time series attributes of data. Your task is to remedy this situation. This includes a requirement analysis from modeling tasks performed in the literature, an actual software implementation for Kibana, and a showcase study based on example sensor backhaul data provided by a large Austrian IoT network.

The ESP8266 offers an interesting WiFi-enabled small-scale embedded platform. Your task is to extend the MicroPython firmware available for it to enable its wireless interface's promiscuous mode. This lets an ESP8266 act as a passive WiFi sensing and measurement device. Based on your implementation, you design a distributed experiment with multiple devices that makes use of your new feature, e.g. track WiFi beacons from smartphones across geographical areas in a privacy-preserving manner (paper).

Radio channel capacity is a key resource to operate wireless networks. Mobile operators spend a lot of money (typically in the range above 1 bn EUR) to obtain spectrum licenses, which give them exclusive rights to use particular frequency bands. In reality, these frequencies are utilitzed to the full extend only in densely populated areas. In other regions (e.g. rural areas) operators do not need their entire licensed spectrum. Instead of static licenses, the proposed topic focuses on load based licensing schemes: Your task is to create a simple simulation of a dynamic "licensing process" which takes into account the current network load. It is expected to implement a simple assignment algorithm, e.g. a license broker in Matlab. The model should be agent based and object orientated.