Research On Reliable Data Transmission Mechanism Of Wireless Peer-to-peer Network On Timely Coding Digestion Strategy

Wireless sensor network is a typical data-centric distributed self-organizing network with data sensing,storage and transmission.It is useful in monitoring and data collecting of harsh environments that humans cannot access.The network nodes not only influenced by the bad environment but also limited by limited resources.In these case the node can be destroyed at anytime,data reliability will be badly affected.Therefore,how to design and realize an efficient data collection mechanism to improve the persistence of data in the network is a meaningful job.Network coding technology is generally used to solve the above-mentioned problems.A representative research in this technology is Growth Codes coding and its relevant works,these protocols can have better performance in harsh or special environments.But it also has the problem of low data collection efficiency.First of all,the Growth Codes coding will generate a large number of backups of the source data.As the backup data is decoded,the proportion of invalid backups gradually increases,it will become difficult for aggregation nodes to collect valid data.In view of the above problems,based on the research of Growth Codes,this dissertation proposes a time efficiency coding strategy and a non-cooperative degree growth coding framework to improve the efficiency in WSN and especially in sparse networks,the following work is included:(1)Redundant copies in the network will reduce the data collecting efficiency,for this problem,this article proposed a Random Feedback Digestion(RFDG)model.With three different feedback methods,designed three data collecting protocols based on broadcast,data exchange and random walk respectively.In this type of protocols,sink nodes share the information of its decoded codewords to the sensor nodes.So that the sensor nodes can delete many codewords which are redundant.These codewords will not be involved in the process of data collecting.Thereby increasing the efficiency of data collecting.Through the experiments,it can be proved that the efficiency can be improve by this type of protocols.(2)The coding degree conversion time of the codeword is the time when the coding process is moderately increased.The Growth Codes computed the best sequence of conversion time,however,this work does not strictly match the model for design,so the sequence is not suitable for the protocol based on the model proposed in this work.Based on the feedback digestion model and the three protocols,the optimal codeword conversion time series suitable for the new model is given by theoretical analysis and derivation.(3)Growth Codes data collection protocol uses a completely random route and a random data backup mechanism.This mechanism divides the data in the network into near-end and far-end data.Due to the growth code mechanism,the far-end data gradually encodes a highly complex codeword that is difficult to be decoded by the sink node.For this problem,this paper proposed a non-cooperative growth level coding framework.The nodes are layered according to the hop count of the sink node,and each node calculates a degree distribution sequence applicable to the node according to the number of neighbors and the level.Through the experiments,the non-cooperative growth level coding mechanism improves the reliability of data collection.In particular,the protocol is more conducive to the recovery of all data by the sparse network system.