Computationally efficient smart antennas for CDMA wireless communications

Abstract

The analysis in this paper concerns the performance of smart antenna algorithms when used in code-division multiple access (CDMA) wireless communication systems. Complex pseudonoise (PN) spreading, despreading, and pilot-aided channel estimates in the cdma2000 reverse link are some of major characteristics that are different from those in the IS-95 CDMA systems. These different features are included in our analysis. Four computationally efficient smart antenna algorithms are introduced: 1) smart antenna based on maximum output power criteria without Lagrange multiplier; 2) smart antenna based on maximum signal-to-interference-plus-noise output power ratio (SINR) criteria with eigenvector solution; 3) smart antenna based on maximum SINR output criteria without eigenvector solution; 4) more simplified smart antenna based on maximum SINR output criteria without eigenvector solution. Algorithms (1) and (4) require only 4M computational instruction cycles per snapshot where M is the number of antenna array elements. Algorithms (2) and (3) require M/sup 2/ and (4M+2M/sup 2/) operations per snapshot, respectively. These computational loads are significantly smaller than those of typical eigenvalue decomposition blind detection approaches. Bit error rates (BERs) resulting from these algorithms are evaluated through simulation. A double spike power delay profile with equal or unequal power is used. Also, a cluster of interfering users and scattered interference users are considered. For BER comparisons, antenna diversity using equal gain combining is also analyzed. The four smart antenna algorithms show significant capacity improvement compared to the antenna array diversity using equal gain combining under the double spike power delay profile with equal power and scattered interference environments.