Abstract: In the perceptual layer of the Internet of Things,introducing IPv6protocol is the key research direction．The length of the data load between IPv6protocol and IEEE802.15.4protocol does not match that the upper level data packets can’t directly exchange data with the underlying protocol．The fragmentation and restructuration algorithm through adaptation layer protocol was used to solve the problem of payload length incompatibilities between the above two kinds of protocols．In order to analyse algorithm performance,fragmentation and assembly algorithm of 6LoWPAN,the improved fragmentation and assembly algorithm of 6LoWPAN and another improved fragmentation and assembly algorithm of 6LoWPAN combining each-hop ARQ algorithm were mathematically modeled and simulated．Firstly,fragmentation and assembly algorithms were simulated and analyzed through the network energy consumption and delay effects of the main performance indexes which contain bit error rate and the number of hop．Secondly,a novel adaptive algorithm framework which aims at optimizing network performance and saving network resources was proposed,so that the adaptation layer can choose different fragmentation and assembly algorithms under the changed network environment．

Key words: IPv6adaptation layer,6LoWPAN,Fragmentation and assembly algorithm