The reliability of data delivery is a major concern in low power loss networks as they are deployed of devices of low energy, processing power and memory. RPL (Recognition of Prior Learning) is the routing protocol developed for these networks by the ROLL, working group. RPL is based on Destination Oriented Direct Acyclic Graph (DODAG). DODAGs have more than one path to the destination unlike the normal tree structure. Mac protocol supports the routing protocol to discover paths to the destinations using the link matrices. In this paper, the authors are comparing the three protocols: RBMAC and a random protocol on DODAG in terms of performance: drop rate and throughput.