Research On Satellite Internet Of Things Access Selection Algorithm Based On Multi-attribute Decision Making

In recent years,the Internet of Things(Io T)technology has developed rapidly,playing an important role in environmental monitoring,smart ho mes,smart cit ies,etc.However,the existing Io T network is based on base stat ions,and base stat ion dep lo yment cost is relat ively high and the survivabilit y is poor,so the terrestrial Io T network has many limitations.These factors have led people to seek other types o f networks to combine with the Io T techno logy,so the concept o f Satellite Io T was proposed.Satellite Io T network is a network that co mbines terrestrial Io T network with low-orbit satellite constellat ions for data transmissio n.In this network,most o f the t ime terrestrial users face the problem of select ive access under mult i-satellite coverage.However,most o f the exist ing access algorithms are single-attribute algorithms,which can not make full use o f satellite informat ion.So me researches on multi-attribute access algorithms o ften ignore the difference between the Io T network and the traditional telephone network,that is,the Io T traffic is do minated by a large number of intermittently sent small data packets,and aggregated traffic shows burst iness and autocorrelat ion.These are characterist ics that tradit ional networks do not have.Therefore,many exist ing mult i-attribute access algorithms can not adapt well to the data access problem in the context of the Internet o f Things.In response to the above problems,this paper cond ucts research fro m two aspects including network traffic characterizat ion and mult i-attribute access algorithms,and proposes a satellite access algorithm framework based on the co mbinat ion o f ON/OFF model network flow algorithm and fixed-weight TOPSIS(Technique for Order Preference by Similarit y to Ideal So lut ion,TOPSIS)algorithm.On the basis of the above framework,in order to better fit the actual network traffic,this paper proposes a non-uniform distribution service model,which makes the network traffic in different regions of the world more variable,and the distribut ion of terrestrial users faced by satellites orbit ing the ground is more co mplicated.In order to further improve the performance of the algorithm,an eliminat ing-TOPSIS access algorithm based on entropy weight method is proposed.The entropy we ight method can determine the weight through exist ing data and reduce subject ivit y,while the eliminat ion process can gradually eliminate obvious disadvantaged candidate satellites,narrow the set o f candidat e satellites,and leave the more advantageous satellites for the final select ion.Through performance simulat ion,it is verified that the improved TOPSIS algorithm is more suitable for non-uniformly distributed service traffic than the fixed-weight TOPSIS algorithm and other single-attribute access algorithms,and is more conducive to improving network performance.