电讯技术
電訊技術
전신기술
Telecommunication Engineering
2015年
9期
972-977
,共6页
尚欢欢%周林%赵华侨%高火涛%刘克刚
尚歡歡%週林%趙華僑%高火濤%劉剋剛
상환환%주림%조화교%고화도%류극강
阵列信号处理%波束形成%方向图综合%稀疏约束%混合范数%凸优化
陣列信號處理%波束形成%方嚮圖綜閤%稀疏約束%混閤範數%凸優化
진렬신호처리%파속형성%방향도종합%희소약속%혼합범수%철우화
array signal processing%beamforming%pattern synthesis%sparse representation%mixed norm%convex optimization
在阵列信号处理中为抑制噪声和干扰,一般都要求波束有比较低的旁瓣和零陷,同时,需要具有良好的稳健性。为此,提出了一种新的混合范数约束的波束综合算法。该算法运用稀疏表示方法,对波束的主瓣和旁瓣分别使用不同的范数约束,同时对波束的零点和敏感度因子进行约束,用MAT-LAB( Matrix Laboratory)的凸优化工具包进行求解。仿真实验表明,设计得到的波束具有更低的旁瓣和零陷,最高旁瓣电平约为-28 dB,而最低零陷深度可达-50 dB;算法求解过程简单,与现有算法相比能够容忍较大的幅相误差,具有良好的稳健性;同时,阵列效率与经典算法相比提高了8 dB左右。
在陣列信號處理中為抑製譟聲和榦擾,一般都要求波束有比較低的徬瓣和零陷,同時,需要具有良好的穩健性。為此,提齣瞭一種新的混閤範數約束的波束綜閤算法。該算法運用稀疏錶示方法,對波束的主瓣和徬瓣分彆使用不同的範數約束,同時對波束的零點和敏感度因子進行約束,用MAT-LAB( Matrix Laboratory)的凸優化工具包進行求解。倣真實驗錶明,設計得到的波束具有更低的徬瓣和零陷,最高徬瓣電平約為-28 dB,而最低零陷深度可達-50 dB;算法求解過程簡單,與現有算法相比能夠容忍較大的幅相誤差,具有良好的穩健性;同時,陣列效率與經典算法相比提高瞭8 dB左右。
재진렬신호처리중위억제조성화간우,일반도요구파속유비교저적방판화령함,동시,수요구유량호적은건성。위차,제출료일충신적혼합범수약속적파속종합산법。해산법운용희소표시방법,대파속적주판화방판분별사용불동적범수약속,동시대파속적영점화민감도인자진행약속,용MAT-LAB( Matrix Laboratory)적철우화공구포진행구해。방진실험표명,설계득도적파속구유경저적방판화령함,최고방판전평약위-28 dB,이최저령함심도가체-50 dB;산법구해과정간단,여현유산법상비능구용인교대적폭상오차,구유량호적은건성;동시,진렬효솔여경전산법상비제고료8 dB좌우。
In order to suppress noise and interference in array pattern synthesis lower side lobe level( SLL) and nulls are required,during the beamforming robustness of the algorithm is vital simultaneously. This pa-per proposes a new method based on the mixed norm constraint. This algorithm adds different norm con-straint respectively for the main lobe and the side lobe by using the theory of sparse representation,at the same time adds constraint on the nulls and the sensitivity and makes use of the toolkit of the convex optimi-zation in Matrix Laboratory( MATLAB) to find the problemˊs solution. The designed beam pattern has low-er SLL whose value is -28 dB and the value of nulls is -50 dB. The proposed algorithm is easy to get the problem solved and can bear the array errors compared with the general algorithms,so it has a good robust-ness. At the same time the arrayˊs efficiency has been improved 8 dB compared with those of the conven-tional algorithms.