Research On Chaotic Spreading Spectrum And Wirless Channel Pararmeter Estimation In CDMA Systems

CDMA (Code Division Multiple Access) is one of kinds of the new and mature wireless communication technique based on spreading spectrum. The basis element of CDMA is to construct a kind of spreading sequence which has excellent correlation property. At the same time, it must have enough coding sequences to meet the need of the multitudinous users. The properties of the chaotic sequences are the auto-correlation is d function and cross-correlation is zero, additionally, because the chaotic system is highly sensitive to original condition, so there are a great many codes. Thus, replacing traditional spreading sequences by chaotic sequences will make the best use of the spreading spectrum communication technique. It is not that any chaotic sequence can be used in spread-spectrum communication, although chaotic sequence has lots of excellence.For instance, unbalance spreading sequence, It should arose carrier wave leak, thus we need search for appropriate sequences.The chaotic sequence is sensitivity to initialize value, so the work of search for the appropriate sequence is complex.The main work of this dissertation is recapitulated as follows.In this dissertation, a chaotic sequence generator is designed which can produce chaotic sequence that has excellent correlation property and sidelobe property, it can be used in CDMA communication system to enhance system performance. This sequence can be proved to be chaos which has positive Lyapunov exponent. And the new chaos sequence has more complex chaotic attactor and dynamics behavior by reconstrust chaos phase space. The new chaotic spreading sequences, tranditional spreading sequence and periodicity chaos spreading sequence are used in CDMA communication system to analyse the BER of the system.Simulation result indicate that the system which uses the proposed chaotic spreading sequence has the lowest BER.Wireless channel parameter estiamtion is another content of this dissertation. Channel parameter (channel coefficients and delay parameter) is estmited by Unscented transformationand UKF which is combine with Kalman filter. Multiuser and near-far effict is consided in this simulation. The result show that UKF can restrain near-far effict in signal detector, so radio channel parameter can be estimated effectly in multiuse and multipath condition.