Research On Access Congestion Control For Massive Internet Of Things Connections

The number of Internet of Things(IoT)devices is increasing day by day and the new IoT applications emerge one after another,which bring huge con-venience for the work,life,study and social management of human.However,due to such massive IoT devices and their unique network requirements com-pared to Human to Human(H2H)communications,the current network cannot support these massive IoT access attempts.This leads to heavy access con-gestion and overload,increasing the access delay and energy consumption,de-creasing successful access probability and the radio access resource efficiency,strongly influencing the H2H communications and even causing the network collapse.And due to the wide coverage,flexible resource managements and so on,cellular networks have become key technologies to support the massive IoT connections.Hence,it is significant and essential to design a reliable and effective congestion control scheme to enhance to ability of cellular networks to handle the massive access attempts,alleviate the heavy access congestion and overload from massive IoT connections and enhance the access resource utilization.In order to solve above issues,this thesis carries out detailed research from:random access traffic,random access channel(RACH)resources and random access strategy based on the time sequence of random access procedure under cellular networks.This research aims to limit incoming random access traffic via the optimal random access arrival control,enhance the access resource effi-ciency via reasonable access resources allocation and execute cooperative ran-dom access on both sides of devices and networks via efficient random access protocol design.Then the access delay and access collisions encountered dur-ing massive IoT access attempts can be significantly decreased.The successful access probability and radio access resource utilization can also be improved.Thereby the access congestion control under the large connection of IoT ac-cess attmepts can be dramatically alleviated.The main research contents and innovations are summarized as follows:1.Random access traffic control for massive IoT connectionsThis thesis aims at the problem of high energy consumption and low data collection efficiency due to the access congestion from massive IoT devices under unmanned aerial vehicle(UAV)based data collection scenario.Firstly,the MAC layer congestion during the access protocols from devices to UAV is analyzed and the corresponding analytical mathematical models are also built.Secondly,the random access stable state is proved and derived and the opti-mal access traffic to maximize the number of successful connections is also obtained.Thirdly,the random access traffic can be optimally controlled in real time via the dynamic velocity adjustments of UAVs.Then the MAC layer ac-cess congestion and overload can be significantly alleviated,simultaneously enhancing the data collection utilization of UAVs.Finally,simulation results verify the accuracy of mathematical network connection models and proves that the proposed optimal access traffic control algorithm can alleviate the access congestion and maximize the data collection efficiency.2.Random access resource scheduling for massive IoT connectionsThis research aims to realize optimal and dynamic RACH allocation for IoT devices under different coverage levels in enhanced Machine Type Com-munications(eMTC)network.Firstly,key technologies of eMTC network pro-posed by The 3rd Generation Partner Project(3GPP)are summarized and ana-lyzed in terms of system cost and complexity,network coverage and power sav-ing mechanisms.Secondly,the random access procedure under eMTC network is studied and analyzed with comparison to traditional random access procedure under cellular network,like Long Term Evolution(LTE).Sirmultaneously,the mathematical models are built and the successful access probability,average ac-cess delay and collision probability are derived.Then the RACH resources are dynamically and optimally allocated for different coverage levels considering corresponding number of access attempts,maximum repetition of preambles and so on.The objective is to maximize the number of successful IoT con-nections.Finally,simulation results verify that the mathematical RACH model of eMTC network can accurately characterize the RACH procedure and the RACH resource allocation can be dynamically and optimally adjusted accord-ing to the access arrivals under different coverage levels to maximize random access performance.3.Cluster-based group paging for massive IoT connectionsThis research aims to solve the access congestion and overload under the group paging mechanism where devices cooperate with each other and form the paging group.The two-step network access framework is introduced and divide the paging group into more clusters with further cooperate among devices.Si-multaneously,communications between devices and base station are separated into inner cluster data collection and cluster header data transmission where cluster header upload the data on behalf of the whole cluster.Then the low cost and high access capacity characterized IEEE 802.11ah network is utilized for the inner cluster data collection phase in order to relieve the transmission con-flicts and collisions within the clusters.Afterwards,mathematical models for both the inner cluster data collection and cluster header data transmission are established in terms of successful access probability and average access delay.And the optimal clusters under one paging group is also derived to maximize the access performance of proposed congestion control scheme.Finally,sim-ulation results verify that the mathematical access models can well match the simulated access results and the proposed dynamic cluster adj ustments can keep the highest successful access probability and the lowest average access delay under different access scales,alleviating the access congestion and overload from massive IoT access attempts.4.Adaptive base station cooperation based congestion control mechanism for massive IoT connectionsAiming at the access congestion for massive IoT devices under multi-ple base stations,this research introduces the biology-based attractor selection mechanism(ASM)and design an adaptive base station cooperation scheme.With external network conditions considered,massive IoT access attempts are adaptively separating into different base stations via providing appropriate base station selection decisions for IoT devices.This can effectively alleviate the access congestion and overload from massive IoT devices.Firstly,an access traffic estimation algorithm is proposed to ensure that base stations can learn their access traffic load with the utilization of random access resources and so on.Secondly,with available RACH resources and the access traffic load of base stations taken into consideration,the activity of ASM is derived and the nonlinear differential equation between the activity and the base station selec-tion probability is designed.Then the load-sharing ratio among multiple base stations would be adjusted via the stable state of ASM in order to maximize the access performance,realizing the adaptive access off-loading among differ-ent base stations.Finally,simulation results prove that the proposed attractor selection based base station cooperation scheme can significantly alleviate the access congestion and overload with high successful access probability and ra-dio access source utilization.