Study of Routing Protocols for Low Power Internet of Things (IoT) Devices

Abstract

Internet of Things (IoT) refers to the interconnection of everyday objects around us and sharing of information among them through the Internet. This information can be used in various applications starting from smart agriculture to smart healthcare and so on. IoT plays a critical role in controlling and automating the whole process making the computation and communication faster. The basic technology used in most of the IoT devices is wireless sensor networks. Every IoT device is embedded with sensors within it that enable it to sense, process the data, and communicate to other things. Thus, the underlying architecture of IoT demands the study and understanding of WSN for its efficient functioning. This thesis focuses on the study of routing protocols developed for IoT based applications. Routing protocols play a critical role in sending the information gathered from the sensors or objects to the required destination for further processing. The information after processing can be used in various applications for helping humans to improve their lifestyle. To start with the research, a Scalable Survival Path Routing (SSPR) protocol has been designed to meet the efficiency of a scalable wireless sensor network. The protocol uses weight based clustering technique to select cluster heads to route data from source to the destination via them in a WSN. This enables to design a scalable network and the protocol performance is increased for larger network. SSPR protocol has been simulated and analysed using NS2 simulator.
While moving towards the routing protocols for IoT, it was observed that the SSPR protocol does not satisfy all the requirements of IoT applications like it does not support Ipv6 packets, cannot be coded to IoT hardware devices and adaptable to all types of traffic patterns. So a composite routing technique named as Modified Routing Protocol for Low Power and Lossy Networks (MRPL) is proposed for such a network. The proposed protocol combines three metrics: Expected Transmission Count (ETX) of a link, Hop count (HC) between source and destination and Available Energy (AE) of a node to select a routing path from source to destination. The path selected for transmission would be the path with less number of packet transmissions, less number of intermediate hops and maximum available energy. ETX metric ensures that good quality links will be selected for data transmission, HC ensures that shortest path to the destination would be selected and available energy ensures that a node does not die out in the process of data transmission. Clustering is a proven solution to be an efficient method to enhance the scalability and v lifetime of any network, a Clustered Approach in RPL (CARPL) has been proposed. The proposed protocol uses a weight based clustering algorithm to perform well even in a larger network. The nodes in the network send their data to their respective cluster heads inside each cluster using one-hop communication. The cluster heads are responsible for aggregating these data and transmitting it to the destination using multi-hop communication. Using this method the scalability as well as the network lifetime is increased.
MRPL and CARPL has been analysed using COOJA simulator in Contiki OS as this simulator supports the features required for IoT like it supports Ipv6 packets, RPL for low power and lossy networks standardized by Internet Engineering Task Force (IETF) and also it supports the code to run inside IoT specific devices like sky motes. It is observed that our proposed protocols outperform the existing approaches in performance metrics like packet delivery ratio, network latency, throughput, and energy consumption.