Research On Blind Equalization For Short Burst Signal Based On Data-reusing Method

High Frequency (HF) communication is a great active embranchment of information science field. Throughout recent years of the development of HF communication, the short burst mode of the transmissive manner gets more and more popular. The signal transited by this mode has a shorter duration, so its anti-interference ability is stronger, and gets more and more applications in reality. At the same time, the factors of limited bandwidth and multi-path lead to the Intersymbol Interference (ISI) in the received symbol, which have an awful effect on the communication quality. So we need to use equalization technic to reduce this ISI. But when the receiver has no training sequence, here blind equalization have to be used to process the signal. So, the research of blind equalization for the short burst signal becomes one of the hotspot in radio communication field.The short burst signal has a shorter duration, less symbol, and the symbol number is hundreds commenly, sometimes only dozens. This condition proposes higher requirements for the equalization in non-cooperative communication. After conclude and summarize the present research of blind equalization, this paper lays emphasis on the Constant Modulus Algorithm based on data-reusing method (DR-CMA), and work hardly on its convergent causation, and its convergent characteristics. Based on above work, this paper designs a new blind equalization algorithm for the short burst signals. The main work and innovation of this paper can be summarized as follows:1. Aiming at the Constant Modulus Algorithm's convergent causation by data-reusing method, this paper compartmentalizes its convergent process for three phases, and analyze particularly every phase's convergent causation. This results indicate that the DR-CMA's convergent direction will emergent fault in the transition phase, if this fault can be modified in the subsequent process, the algorithm will convergent at the right direction.2. Aiming at the Constant Modulus Algorithm's convergent characteristics by data-reusing, Based on the minimum entropy algorithm, this paper propose an analysis method of convergent characteristics base on the minimum entropy algorithm. And analyze deeply these characteristics from the entropy point of view. The research shows that, the smaller of the short burst signal's entropy, the better of the equalization's performance.3. Based on above research, and aiming at the problems of the data-reusing CMA based on affine projection when solved the short burst signal's blind equalization, such as the high complexity, and so on, this paper designs a data-reusing CMA based on likely-affine projection, this algorithm combine the idea of affine projection and the advantage of CMA, reduce its operative complexity, pick up its convergent speed, improve its convergent performance. The analysis and experiment shows that, the data-reusing CMA based on likely-affine projection has the higher efficiency, faster speed, better performance.4. Aiming at non-cooperative communication, this paper designs a blind equalization scheme for the short burst signal. and programed the data-reusing CMA based on likely-affine projection in software according to the demands, Furthermore, practical signal tests show that this scheme can recover the information of received 8PSK signal which only have 150 symbols, and in another way, it also prove that this blind equalization is feasible and efficient.