| Along with the development of wire-less communication, the frequency resources become more and more lacking. The most important challenges in the field of wire-less communication are using limited frequency resources to transfer infinity information, and restraining the wire-less interfere. So, when the 3G communication system is advanced, the wire-less transferring technique became the focus of research at once. The smart antenna technic, as one part of the wire-less transferring, also became one of the key techniques in 3G researching.Virtually, the smart antenna makes use of the differences in the spatial dimension between varied signals, and then deal with them in the special dimension. Corresponding with the FDMA, TDMA, CDMA system, we can consider the smart antenna as SDMA(space division multiple access), therefore, the communication resource no longer located just in the time dimension, frequency dimension or code dimension but exhibited to the space dimension using this SDMA system. Smart antenna can division the different users based on the user signals from the different space dimension even if the signals are in the same time dimension the same frequency dimension and the same code dimension. In another words, the smart antenna makes use of the differences between the incidence angles of the useful signal and the interfere signal and then choose a proper merge weight finally, so a correct antenna incept mode is formed: the main beam aims at the expecting signal and the side lobe aims at the interference signal. Using such precept we can acquire a wire-less circuit with energy centralized between base stations and users.Smart antenna can sort three kinds. The first one is the fixed wave-packet smart antenna. This kind of antenna establishes some weights which point to different directions in advance, each beam direction is fixed, and the beam width changes with the numbers of the element. The wave-packed with better receiving signal is adopted, and along with the movement of user in areola, the base station choose different corresponded wave-packets to acquire the best receiving signal; The second one is partly self-adaption smart antenna. This kind of antenna commonly pick up the arrival direction information of user signals from the receiving array single, then form the wave-packet towards the arrival direction. The weights varied with the arrival direction. The rule of the algorithmic is maximize the expecting signal energy of user received, and limit the interference signal of other directions, so we can equate as a complex weight with variable phase and fixed altitude. Some self-adaption smart antenna has better capability than the fixed wave-packet smart antenna, but the space information of signals is not totally used, and the self-adaption range is limited, also the arithmetic of the direction angle extract is complex; The third one total self-adaption smart antenna. This kind of antenna ceaseless adjust weights by the change of the space distribution characteristic of signals in one rule. The phase and altitude of the weights both can update freely. This means can make the most of the space characteristic the user signal and interference signal to maximize the SNR of received signals when acquire convergence arithmetic, and the arrival direction of interference signals is neglected commonly.The total self-adaption smart antenna is amply discussed in this paper. There are three segments of the total self-adaption smart antenna: 1. Antenna array segment, This segment only receive signals but don't process it; but the elements number and their arrangement manner can influence the whole smart antenna system, the uniform linear array is mainly discussed, and the element number is 8. 2. A-D/D-A exchange segment, The main function of this segment are D-A exchange in upload and A-D exchange in download; 3. Smart enable segment, This segment has two constituent: the self-adaption processor and the wave-packed shaped network. The self-adaption processor can transfer the disciplinary pumping signal to the phase and altitude corresponded with each wave-packed, then supply each radicalization element to determine each directional diagram plus of the wave-packed shaped network. The wave-packed shaped network can make use of the directional diagram generated by the antenna array to realize kinds of smart antenna application. The self-adaption processor generates the altitude and phase adjusting modulus of each branch which are the important gist in the process of the wave-packed shaped network, The self-adaption processor contains the signal processor and the self-adaption arithmetic segment. The signal processor works based on the processed signal, and the self-adaption arithmetic segment form the weight number of the network by adopting the proper arithmetic the direction diagram based on the capability parameter such as the mean-square error, the SNR, the output noise power, and optimize the capability of the antenna array.The core of the self-adaption smart antenna research is the self-adaption arithmetic which can self-adjust the weight number synchronously as environment varied. Resumptively speaking, there are two big branch such as the non-blind arithmetic and the blind arithmetic. The non-blind arithmetic is a arithmetic needs a reference signal(pilot sequence or pilot channel), in such situation the receiving end know what transmitted, and the arithmetic processes the signal by confirming the channel response, then deduces each weight number by the certain rule, also the arithmetic processes the signal by confirming or adjusting the weight number gradually, to make the output and foregone input correlation ratio maximum. The blind arithmetic can achieve the smart antenna function without the reference participation.The blind self-adaption arithmetic is mainly researched in this paper, especially the blind self-adaption arithmetic which is applied in the CDMA system is mainly discussed. Because of multiple users share one wire-less channel in the CDMA system, the number of users is far greater than the number of antenna arraies, so the blind delimitation and blind extract abilities should be provide by the self-adaption shaped arithmetic, i.e. the blind arithmetic should be multiple target arithmetic, namely the smart antenna system should have multiple output ports.Firstly, the MT_LSCMA which early used in CDMA system is researched and simulinked in this paper, Such arithmetic is complex, also it has the"interference capture"problem; Secondly, LS-DRMTA is researched, Such arithmetic uses the expansion frequency array of the CDMA technology, employ the ith bit detection signal of the kth user as the prior information, so the signal waveform emitted by the kth user in the [(i ? 1)Tb ,iTb] periodic time can be deduced by expand the frequency again, then the re-expended signal can be used to update the user's k weight number. The simulation results indicates that it is precede than the MT_LSCMA both in the arithmetic complex level and the wave-packed shaped veracity. Lastly, the LS-DRMTCMA is researched, Such arithmetic combine the characteristics of the LS-DRMTA and the MT-LSCMA, it use the expansion frequency array signal and the constant modulus specialty to update the weight vectors. It has merits of LS-DRMTA, also can acquire lower error rate than LS-DRMTA, certainly the enhance of capability goes with more of the arithmetic complex.The arithmetic and simulation above all base on the reference signal less than the expected signal, but the"distance effect"is exist in the CDMA system, i.e. the reference signal may be greater than the expect signal, so a kind of improved arithmetic bring forward to settle this problem in this paper at last. It is proved that the can conquer the"interference conquer"problem primely, as a result, a new arithmetic is acquired by using the idea of SMI-LSCMA and combined with the expansion frequency array signal in the CDMA system. The simulation result validate the improved arithmetic can acquire excellent wave-packed shaped result when the reference signal is greater than the expected signal, also acquire better dynamic trace capability and higher output SNR. So the improved arithmetic is significative. |
PDF Full Text Request