Research On Fast Network Coding Technology Over Data Streams In Wireless Sensor Networks

In recent years,the Wireless Sensor Networks(WSNs)have been developed rapidly with the advancement of new technology and the cost reduction of manufactory.Due to the ease of deployment and versatility of sensor nodes,WSNs are widely used in various fields such as industrial manufacturing,smart home appliances,smart agriculture,medical care,and disaster relief.The main task of the sensor node is to collect,save,and aggregate data.And the further data analysis operations are usually done in the data center while the data analysis results are fed back though the network to the user.The explosive growth of sensor data and the unreliability of wireless links bring enormous challenge to the design of efficient communication and storage mechanisms.And with the popularization of the Internet and the Big Data application of upper layer,the demand for WSNs data's reliability and low delay transmission becomes urgent.Reliable and fast data delivery has always been as an important design goal of WSNs.Researchers have applied various coding methods on data transmission and redundant storage for coping with the complex working environment of WSNs and disaster recovery after WSNs data collection.However,the additional overhead of these codecs would bring the challenges to the fast data delivery.In most existing transmission models of WSNs,data transmission and redundant storage are as two separate modules,which make wireless sensor data multiple encoding and two decoding in the process of collection and transmission to analysis,which leads to a large amount of computing resources waste and data delivery delay.In addition,because of some invalid data in WSNs occupy a large amount of codec resources,and the traditional decoding method increase this effect,it is difficult for the traditional transmission architecture to achieve the needs of real-time computing.This research is attempting to study the fast transmission model of wireless sensor data from collection to analysis processing.Combined with network coding,the coding method,routing strategy and decoding scheme which belong to the fast model are studied.The main research work and innovations are as follows:(1)This study analyzed the delivery process of wireless sensor data based on coding,and proposed an integrated transmission scheme based on network coding.This scheme improves the efficiency of computing resources in the whole transmission operation by reducing the times of encoding and decoding in the process of sensor data collection and analysis.The method proposed in this study by experimental analysis is significantly modified in terms of data delivery delay compared with the classical method.(2)This study proposed a fast coding scheme of network coding.Firstly,this research analysis the linear correlation of random linear coding,the upper triangular matrix coding and Vandermonde matrix coding,discussed the linear correlation and application range of the above coding strategies.The designed weights Vandermonde echelon fast coding scheme can ensure the coding matrix linear independence and distinguish the coding according to the real-time requirement of the packet.The fast coding scheme significantly reduces the time required for large-scale node operations,and it can prevent invalid data takes up the data center in WSNs computing resources.Experimental results showed that the proposed fast coding algorithm is more efficient and more accurate compared with other coding algorithms.(3)The dissertation proposed a fast multipath routing scheme of sensor networks based on network coding,and designed network coding based routing schemes for wireless sensor networks and mobile wireless sensor networks respectively: the gradient constrained flood routing algorithm and the fast routing scheme based on ant colony and network coding.This paper took advantage of the special conditions in the two environments to establish fast routing,and increased the coding opportunities of network coding to achieve a full mix of data streams,which can improve network disaster tolerance for achieving fast data delivery.The experimental results proved that our two fast multipath routings have better performance.(4)The dissertation proposed a fast parallel decoding scheme of wireless sensor data in the data center,which regards the server as the minimum calculation unit.The proposed scheme accelerated the delivery time of wireless sensor data from acquisition to analysis,and made the decoding in data center no longer limited to the computing power of single server theoretically.The experimental results showed that the proposed parallel decoding scheme has better performance.