In the last few years, Wireless Sensor Network (WSN) emerges and appears as an essential platform for prominent concept of Internet of Things (IoT). Their application ranges from so-called "smart cities", "smart homes" over environmental monitoring. The connectivity in IoT mainly relies on RPL (IPv6 Routing Protocol for Low Power and Lossy Network) -a routing algorithm that constructs and maintains DODAGs (Destination Oriented Directed Acyclic Graph) to transmit data from sensors to root over a single path. However, due to the resource constraints of sensor nodes and the unreliability of wireless links, single-path routing approaches cannot be considered effective techniques to meet the performance demands of various applications. In order to overcome these problems, many individuals and group research focuses on multi-path solutions for RPL routing protocol. In this paper, we propose three multipath schemes based on RPL (Energy Load Balancing-ELB, Fast Local Repair-FLR and theirs combination-ELB-FLR) and integrate them in a modified IPv6 communication stack for IoT. These schemes are implemented in OMNET++ simulator and the experiment outcomes show that our approaches have achieved better energy efficiency, better end-to-end delay, packet delivery rate and network load balance compared to traditional solution of RPL.