| Since the massive machine-type communications(m MTC)was classified by International Telecommunication Union as one of three application scenarios in fifth generation(5G)mobile communication system,the dense nodes communication(DNC)with high density of machine communications,whose typical application scenarios are massive machine-to-machine(M2M)communications and m MTC,has drawn much attention from researchers all over the world.The wireless mobile communication networks(WMCNs),which are the most ubiquitous communication system,have been considered as the most competitive solution for DNC.However,because of the differences between DNC and mobile broadband communications,the WMCNs can hardly meet the low resource consumption demand of DNC.To solve this,the dissertation investigates problems relating to the communication reliability,information redundancy problem and device resource constraints in DNC,which lead to the high consumption of radio resource and device resource.Firstly,to address the M2 M communication reliability problem,a random subchannel selection(RSS)based M2 M communication method is proposed.Secondly,to reduce the information redundancy of DNC based on store-carry-and-forward relaying(SCFR)in WMCNs,a SCFR communication method is introduced based on the optimal stopping theory.Lastly,to solve the resource constrains problem in DNC devices(i.e.M2 M nodes),an iterative signal processing stopping criterion based on equations satisfaction probability(ESP)and a corresponding algorithm are provided,and then an iterative signal processing chip using the proposed algorithm and probability processing technology(PPT)are designed and fabricated.The main contributions of this dissertation are as follows.1.A RSS based M2 M interference avoiding scheme are proposed to increase the reliability of M2 M communications in DNC.Focusing on enhancing the M2 M communication reliability,the proposed scheme can reduce the M2 M interference by allowing source devices to select several orthogonal subchannels autonomously,such that it does not produce any uplink signaling overhead or occupy any uplink radio resources.Thus,by increasing M2 M signal-to-interference-noise ratio and communication reliability,the proposed method can enhance data rate and spectrum efficiency,lower the transmission delay and energy consumption,and improve system throughput.2.Two optimal stopping relay selection rules aiming to improve the transmission delay and energy consumption are proposed,based on which,SCFR relay selection protocols are also designed.In order to lower the transmission delay and energy consumption,two stopping rule problems are constructed and solved,and then two stopping rules are proposed.Considering the features of DNC,two relay selection protocols are designed based on proposed stopping rules.Compared to traditional relay selection protocols,such as epidemic routing protocol and random relay selection protocol,the proposed protocols can reduce M2 M transmission delay and energy consumption,as well as information redundancy.3.An ESP and PPT based iterative signal processing stopping criterion and a PPT algorithm are proposed to meet the performance and resource consumption demand of signal processing in M2 M devices.Focusing on finding a PPT stopping criterion and closing the gap between different PPT algorithms,this dissertation first defines a PPT graph model and its updating rules to model a sort of signal processing problems that can be viewed as solving linear equations,based on which the stopping criterion and algorithm are proposed.Furthermore,the proposed algorithms can reduce the processing delay and the number of transistors used in circuits when providing the same performance as traditional algorithms.4.An isomorphic graph based signal processing chip designing method is proposed to lower the resource consumption of signal processing in DNC,and a PPT chip is designed and fabricated.By finding the isomorphic graph and designing the reusable building block,the complexity of designing a PPT chip is reduced and a PPT based LDPC decoder is designed and fabricated.Then,the dissertation designs the input and output interfaces,such that the incompatibility problem between PPT based signal processing chips and current communication systems can be addressed.At last,the decoder chip is fabricated using 0.35μm CMOS technique.Experimental results show that,its energy consumption is lower than its digital counterpart,and it is the PPT decoder that owns the largest scale and highest coding gain to data,which makes it useful for M2 M devices. |
