Research On Clustering Spectrum Sensing And Sharing For Cognitive Railway Mobile Communication Network

With the increasing demand of the data-carrying capacity between train and ground, and more demand is put forward by passengers to use e-business, IM communication and online entertainments on the train, the high-speed railway mobile communication system faces the problem of limited type of service, un-guaranteed QoS and low transmission rate, due to the scarcity of licensed spectrum resource.Cognitive radio is a new paradigm of designing wireless communication systems which aims to cease the contradiction between limited wireless spectrum resource and low efficiency static spectrum allocation mode. This provides a feasible solution to solve the problem of scarcity of licensed spectrum resource in high-speed railway mobile communication system: the key application(such as train control and dispatch communication) is carried by GSM-R/LTE-R, and high volume data communication, which beyond the carrying capacity of them, is carried by spectrum hole which is sensed by cognitive user, with certain spectrum sharing scheme.In the dissertation, due to the complexity of railway communication, serious electro-magnetic interference caused by the electrified railway and the railway operation characteristics, the cognitive railway mobile communication networks(CRMCN) model is proposed. The dissertation mainly study the clustering scheme, spectrum sensing based on clustering, and spectrum sharing algorithm in CRMCN. The main contributions of this dissertation include:A combined weight clustering algorithm is proposed, termed as CRMCN-WCA. Considering that the large number of nodes(CR users) in a cell, the diverse types of nodes and the obvious differences of nodes’ mobility, the metrics, including the ideal degree, transmission power, node power, clustering stability and node type have been considered as weight metrics to elect the cluster-heads. Meanwhile, the scheme of cluster maintenance and gateway node selection is studied. The CRMCN-WCA has good performances in terms of network stability and clustering maintenance cost for the case of a lot of CR users in high speed moving.According to lack of prior knowledge of PU signal and high demand for practicability of sensing algorithm, a cooperative clustering spectrum sensing algorithm based on threshold closeness factor(TCFS) is proposed. According to the energy detection theory, if the collected energy is located in the “decision” region, the CR user sends 1 bit or 2 bit self-adaptive quantization to report its sensing result based on TCFS and likelihood decision. Otherwise, the CR user sends nothing. Performance analysis show that the average number of sensing bits decreases without noticeable loss in sensing performance(in terms of the missing probability and the false alarming probability) while using this algorithm.A repeated game based spectrum sharing algorithm is proposed. The problem of maximize the total rate profit of spectrum sharing of two CR users is solved with the algorithm when the channel competition happened in a cluster. The “a cooperative trigger strategy” is adopted to obtain the Pareto optimality. Next, the convergence behavior of the total rate profit for spectrum sharing is analyzed for considering the impacts of CR uses’ transmission power, noise power, discount factor and convergence coefficient. The fairness of CR user is also considered in this algorithm.The achievements of the dissertation will provide theoretical foundations and technical support for efficient network construction and evolution to the high-speed railway broadband mobile communication network.