Research On Communication Modeling,Cross-layer MAC And Routing Protocols For Electromagnetic Nanonetworks

Electromagnetic nanonetwork is composed by a large number of nano-nodes,which range from a few hundred cubic nanometers to a few cubic micrometers.Due to the limitation of size,nano-nodes have to communicate at Terahertz(THz)band(0.1-10 THz).Compare to the traditional wireless sensor networks,nanonetworks are expected to expand the capabilities,and promise new solutions for serveral applications in military,environmental and biomedical fields.However,due to the physical constraints of nano-node and the peculiarities of communication channel in the THz band,traditional communication technologies are not suitable for Nanonetworks.In this thesis,the main features of Nanonetworks,such as high node density,high path loss of THz band,energy limitation and low capabilities of nano-node,are comprehensively taken into consideration to investigate the communication modeling and cross-layer protocol design for Nanonetworks.The outcomes of this thesis can not only complete the basic fundamentals of Nanonetworks,also emphasize on the design of higher layer protocols.Based on the existing THz channel attenuation model: 1)a multipath interference and coverage model is proposed by stochastic geometry methods;2)Aiming at the centralized and distributed network topology and Time Spread On-Off(TS-OOK)modulation scheme,a scheduling receiver driven-based MAC protocol is proposed;3)Combining with artificial intelligence algorithms,a deflection routing algorithm based on reinforcement learning is presented;4)Modeling the achievable throughput of energy-harvesting nanonetworks in THz band to provide theoretical basis for the design of Nanonetworks.Specific work and results are as follows:1.The multi-path interference and coverage area of beamforming nano-controller is modeled for 3D nanonetworks based on stochastic geometry,providing the theoretical basis for communication protocol design.First,a 3D blockage model is introduced with beamforming nano-controller based on the peculiarities of THz channel.Then,the Line of Sight(LoS)and Non-Line of Sight(NLoS)interference form nano-nodes and nano-controller are modeled by stochastic geometry methods,and the corresponding Singl to Interference plus Noise Ratio(SINR)is obtained.Finally,by combining the features of 3D nanonetwork structure and beamforming antenna,the coverage model of nano-controller is presented.2.To address the high density,extremely energy and resources constraints of nano-nodes,a novel scheduling receiver driven-based MAC protocol(SRD-MAC)is proposed.SRD-MAC can effectively reduce the interference and collision probability,and improve the performance of nanonetworks.In the protocol,nano-node can divide a time frame into several time slots,and allocate corresponding Receiver Driven-based Slot(RDS)by its own ID.Based on the interference and coverage models,two different communication methods are designed for centralized and distributed network strucutres.The energy consumption,delay and throughput of the proposed protocol are studied.The simulations are conducted with the comparsions of other MAC protocols.From the results,it can be concluded that the proposed SRD-MAC has the best performance.3.A multi-hop deflection routing algorithm based on reinforcement learning(MDR-RL)is proposed based on the peculiarities of nano-nodes and energy harvesting system,to adapt to the fluctuation in energy of nano-node and improve the throughput.Firstly,new routing and deflection tables are implemented in nano-nodes,so that nano-nodes can deflect packets to other neighbors when route entry in routing table is invalid(caused by energy or buffer restricts)to improve the packet delivery probability.Then,an energy prediction scheme is introduced to enhance the deflection decision-making process.One forward updating scheme and two feedback updating schemes based on reinforcement learning are designed to update the tables by utilize the deflection probability,loss probability,hop count to destinations and energy status of nano-nodes.The proposed algorithm can adapt to the change of nano-nodes and network traffics.The simulation results show that the MDR-RL using on-policy updating scheme performes better than the other updating schemes and flooding algorithm.4.Throughput is a key element for nanonetworks.In order to understand the performance of nanonetworks under different parameters and help researchers to design nanonetworks under different scenarios,the maximum achievable throughput of electromagnetic nanonetworks in THz band is comprehensively investigated.This model valuable for researchers to configure different parameters according to different network environments and requiremnets to realize more efficient nano-networks.On the one hand,the peculiarities of the THz band channel are taken into consideration by capturing the impact of spreading loss and molecular absorption loss.While,the two energy harvesting states of nano-node is modeled.Based on the proposed energy efficiency and spectrum efficiency,and combine the Shannon capacity theorem,the throughput is constrained by energy harvesting rate and maximum transmission power.Finally,the accuracy of the model is validated by experimental simualtion.