Network Performance Enhancement of Multi-sink Enabled Low Power Lossy Networks in SDN Based Internet of Things

Software Defined Network (SDN) brought revolution in the network field with the partnership of Academia and Industry. SDN bridges the gap to overcome issues of IoT deployment, optimization and better utilization of network resources. The escalation in resource congestion in Wireless Sensor Networks (WSNs) can usually lead to scalability, data computation or storage, and energy efficiency problems with only a single sink node for data acquisition. Internet of Things (IoT) has resource and energy constraints for WSN devices. Low Power and Lossy Networks (LLNs) ought to be optimized for traffic with multiple sinks. RPL routing has constraints to support this approach. However, RPL inherits the ability to offer features like Auto-Configuration, Self-Healing, Loop avoidance, and detection. These features of RPL can be transformed into the improved performance of a WSN by increasing the number of sinks with a linear increase of data transmitting nodes in the network. Further, to mitigate the escalated computing needs, edge computing has emerged as a new paradigm to resolve SDN-enabled IoT and localized computing needs. This study proposes an SDN-based solution to the interconnectivity of resource constraint LLN devices with edge computing routers in mesh and cluster topological scenario using RPL as IoT routing protocol. Performance evaluation concerning different routing metrics and objective functions: Minimum Rank with Hysteresis Function (MRHOF) and Zero (OF0) are analyzed. COOJA simulator is used for emulation of random as well as linear grid topologies for the creation of WSN static nodes. Simulation results confirm that the gradual increase of a number of nodes from 16, 32, 48, 64 and a simultaneous increase in sinks nodes as 1, 2, 3, 4 respectively in LLN network reflects the desired advantages with the stable network.