An modified hybrid invasive weed optimization for array antenna beamforming

To reduce the desired signal cancellation at high input SNR as well as improving the precision of antenna array beamforming, an modified hybrid invasive weed optimization (MHIWO) algorithm is proposed in this paper. A new distribution variance takes the place of the exponential function used in the original invasive weed optimization(IWO), thus can balance the global convergence and the local convergence well. Additionally, the t-distribution function is used as the spatial distribution of the offspring seeds. Simulations are performed based on sphere, rastrigin and ackley functions to demonstrate the high precision and high convergence rate of the proposed method as compared with IWO, modified adaptive dispersion invasive weed optimization (MADIWO) and MHIWO. The viability of the improved algorithm is also illustrated in the application of antenna array beamforming. When compared with typical beamforming, MHIWO algorithm can suppress the interference successfully and is not inclined to falling into local optimal solution. Furthermore MHIWO based beamforming can obtain high output SINR even at high input SNR.