Performance Research For NGB Access Network Upstream Channel Under Noise And Intelligent Diagnosis Study

Radio&TV is the most popular carrier of transmitting information and also is themost important platform of propagating public opinions. It will play an important rolein the national information technology infrastructure in the future. With the rapiddevelopment of IT, more and more access to information channels has beenestablished and more and more rich information obtained by the masses. Especially inearly2010, the nation has made a plan about accelerating the telecommunicationnetwork, broadcast network and internet network integration strategic plan and clearlymade a network integration schedule. In order to adapt to the new situation andmeet the needs of the masses on digital media and information services, we mustintroduce advanced techniques and concepts to transform the existing TV system. Thenext generation broadcast network comes into being. The next generation broadcastnetwork is a product of digital TV broadcasting technology and digital informationtechnology integration, and it consists of telecommunication network, computernetwork and cable television network. Users can access to a variety of information inaccordance with their own needs, and it fundamentally change the master-slaverelationship "you broadcast and I receive" between radio and television and form anew media services. At the same time, in order to meet the needs of construction anddevelopment of the next generation broadcast network and break the high-endequipment monopolized by foreign business, the theory and method related to test anddetection has become a hot research in the domestic, and a variety of specialinstrument development have already been included in the schedule.In this paper, based on the needs for development of the next generation cablebroadcast network monitor system, we deeply studied noise characteristics and faultdetection. In view of downlink channel performance research has been mature, thispaper will focus on: detailed derivation and build of the next generation cablebroadcast network uplink channel noise model, the research and development ofOFDM base-band system simulation platform and uplink channel failure analysisdetection system using neural network technology. The main innovation works are asfollows:Access network technology and multi-carrier communication is the coretechnology for the next generation broadcast networks. Based on the analysis andstudy of the related access network technique protocol standard, this dissertation focuses on the characteristic function model including Gaussian noise, narrowbandcontinuous wave noise, pulse noise and BER expression, and the performance of nextgeneration cable television network based on the OFDM technique. UsingMatlab/Simulink software, we firstly constructed some modules such as the signalsource, modulation and demodulation, OFDM and BER calculation. Then base-bandOFDM systems simulation platform is formed using typical noise model andmodeling technology. Finally the simulation results are obtained by simulating theOFDM systems and analyzing the performance of communication system at theplatform. It lays solid foundation for further researching the next generation broadcastnetwork performance and developing special broadband network monitor system withindependent property.Based on the above research results, BP neural network technique can be used inthe area of the next generation cable broadcast and TV network for upstream channelfault diagnosis. Spectrum characteristics of upstream channel are used as neuralnetwork inputs. Upstream channel spectrum data obtained by actual measurement isused to train the neural network so that noise and interference type can be obtained byanalyzing the neural network output. After a lot of simulations, the most suitablelearning algorithm and activation function of the BP neural network can be obtainedto be used for upstream channel fault diagnosis of the next generation cable broadcastand TV network. The corrections of the weights were also derived in this thesis basedon the most suitable algorithm and activation function.Based on noise model expression derived by this thesis, upstream channel faultdiagnosis identification system for the next generation cable broadcast and TVnetwork can be built as well as cable broadcast and TV communication multi-carriersystem simulation platform. We designed a scheme of upstream channel noisemonitoring system for the next generation cable broadcast and TV network, which hasbeen embedded into the related products of the Deli Electronics Limited Company. Itplays a core technical role for the final product. Proved by actual application, theaccuracy of the fault diagnosis system based on BP neural network is more than85%.It verifies the correctness of the noise model, simulation platform and fault diagnosisalgorithm. We hope to contribute to the development of the next generation high-endbroadcast and TV network instrument for a special purpose with independentintellectual property rights.