Student:	Mustafa NAJMULDEEN
Supervisor:	Asst. Prof. Dr. Sedat GÖRMÜŞ
Department:	Bilgisayar Mühendisliği
Institution:	Graduate School of Natural and Applied Sciences
University:	Karadeniz Technical University Turkey

Title of the Thesis:	Distributed Caching for Wireless IoT
Level:	M.Sc.
Acceptance Date:	21/8/2020
Number of Pages:	37
Registration Number:	i3796

Summary:

The future Internet will be connecting billions of low power embedded devices. This

globally connected network of small embedded devices is thought to be generating most of

the traffic carried over the future Internet backbone. This new Internet containing billions of

small embedded devices are the Internet of Things (IoT). Furthermore, significant portion

of the information data generated by the IoT networks will need to be stored in data centers

contributing to the scalability issues already affecting the Internet. New and novel solutions

need to be created to prevent this data explosions impact on the Internet infrastructure.

In view of this thesis, an uncommon dispersed cache solution was proposed to reduce

the impact of the IoT traffic on the Internet backbone. In this solution, the generated data is

processed and stored within the IoT network securely via a distributed storage solution. This

solution utilizes the properties of Maximum Distance Separable codes (MDS) to create an

agreement between the network bandwidth and storage area. This solution will enable

storing the sensor data at the network edge trying to tackle the scalability challenges imposed

on the Internet by billions of IoT devices. In this study, low power wireless device networks

are partitioned into disjoint cliques with minimum overlap to reliably store the generated

sensor data within these disjoint cliques in a distributed manner. Furthermore, the proposed

wireless storage solution creates optimal clique coverage to boost the accuracies of the

caching system.

Key Words: Internet of Things, MDS Codes, Distributed Storage, Cliques and RPL