| Low Power Wireless Personal Area Network (LoWPAN) mostly uses IEEE802.15.4standard, which has low data rate, low battery power and other characteristics. LoWPAN is widely used in smart home, agricultural monitoring, health care and other fields. In the Internet of Things, LoWPAN is made up of large number of sensor nodes. In order to connect the nodes in Internet to the nodes in LoWPAN, IETF working group proposes IPv6over Low-Power Wireless Personal Area Networks (6LoWPAN) protocol to achieve the IPv6packets delivery on LoWPAN.Compared with the traditional protocol stack,6LoWPAN protocol add the adaption lay between the network layer and data link layer. This layer is responsible for fragmentation and reassembly of the packet. Because6LoWPAN link is lossy, for a IPv6packet, the more the fragments are, the smaller the probability of the whole packet reaching to the sink. Conversely, the less the fragments are, the larger the number of bits each fragment occupy, which causes the larger probability of loss in wireless link. Therefore, choosing an appropriate number of fragments has an important impact on the throughput of the network.This paper studies data transmission scheme for6LoWPAN aimed at improving the throughput. It begins with a brief analysis of the6LoWPAN related technologies, including IEEE802.15.4standard, IPv6technology, fragmentation and reassembly, header compression, routing strategies of adaptation layer. Then analyze the multi-path routing technologies, including protocol elements, current research. On this basis, we propose a packet fragmentation scheme based on multi-path routing. The main work and innovation are as follows:1. Derive the formula of the number of transmissions, latency and throughput of a IPv6packet from source to sink.2. Use IPv6packet fragments as optimization variable and throughput as objective function, to build a mathematical model based on multi-path routing for optimizing throughput.3. Simulate the proposed multi-path routing algorithm on MATLAB, to verify the correctness of formula derivation, analyze the impact of the number of fragments on throughput. |
