Congestion Control for 6LoWPAN Wireless Sensor Networks: Toward the Internet of Things

The Internet of Things (IoT) is the next big challenge for the research community. The IPv6 over low power wireless personal area network (6LoWPAN) protocol stack is considered a key part of the IoT. Due to power, bandwidth, memory and processing resources limitation, heavy network traffic in 6LoWPAN networks causes congestion which significantly degrades network performance and impacts on the quality of service (QoS) aspects. This thesis addresses the congestion control issue in 6LoWPAN networks. In addition, the related literature is examined to define the set of current issues and to define the set of objectives based upon this.
An analytical model of congestion for 6LoWPAN networks is proposed using
Markov chain and queuing theory. The derived model calculates the buffer loss probability and the number of received packets at the final destination in the presence of congestion. Simulation results show that the analytical modelling of congestion has a good agreement with simulation. Next, the impact of congestion on 6LoWPAN networks is explored through simulations and real experiments where an extensive analysis is carried out with different scenarios and parameters.
Analysis results show that when congestion occurs, the majority of packets are lost due to buffer overflow as compared to channel loss. Therefore, it is important to consider buffer occupancy in protocol design to improve network performance.
Based on the analysis conclusion, a new IPv6 Routing Protocol for Low-Power and Lossy Network (RPL) routing metric called Buffer Occupancy is proposed that reduces the number of lost packets due to buffer overflow when congestion occurs. Also, a new RPL objective function called Congestion-Aware Objective
Function (CA-OF) is presented. The proposed objective function works efficiently and improves the network performance by selecting less congested paths. However, sometimes the non-congested paths are not available and adapting the sending rates of source nodes is important to mitigate the congestion.
Accordingly, the congestion problem is formulated as a non-cooperative game framework where the nodes (players) behave uncooperatively and demand high data rate in a selfish way. Based on this framework, a novel and simple congestion control mechanism called Game Theory based Congestion Control Framework
(GTCCF) is proposed to adapt the sending rates of nodes and therefore, congestion can be solved. The existence and uniqueness of Nash equilibrium in the designed game is proved and the optimal game solution is computed by using
Lagrange multipliers and Karush-Kuhn-Tucker (KKT) conditions. GTCCF is aware of node priorities and application priorities to support the IoT application requirements. On the other hand, combining and utilizing the resource control strategy (i.e. finding non-congested paths) and the traffic control strategy (i.e. adapting sending rate of nodes) into a hybrid scheme is important to efficiently utilize the network resources. Based on this, a novel congestion control algorithm called Optimization based Hybrid Congestion Alleviation (OHCA) is proposed. The proposed algorithm combines traffic control and resource control strategies into a hybrid solution by using the Network Utility Maximization (NUM) framework and a multi-attribute optimization methodology respectively. Also, the proposed algorithm is aware of node priorities and application priorities to support the IoT application requirements.