Computationally efficient approaches for blind adaptive beamforming in SIMO-OFDM systems

In orthogonal frequency division multiplexing (OFDM) based transmissions over single-input multiple-output (SIMO) wireless channels, adaptive beamforming can be employed at the receiver side to combat the effect of directional interference. To save channel bandwidth, there is strong incentive to use blind algorithms that attempt to restore properties of the transmitted digital signals. Among these, the recursive least-squares constant modulus algorithm (RLS-CMA) is of considerable interest due to its fast convergence and good interference cancellation properties. However, since a distinct copy of the RLS-CMA must be run on each individual sub-carrier in OFDM applications, this approach may entail considerable computations. In this paper, we investigate frequency interpolation schemes to reduce the computational complexity of the SIMO-OFDM beamforming system based on the RLS-CMA. These approaches, which exploit the coherence bandwidth of the broadband wireless channels, divide the subcarriers into several contiguous groups and apply the RLS-CMA to a selected subcarrier in each group; the weight vectors at other frequencies are then obtained by interpolation. We show through simulations that an M-fold reduction in complexity can be achieved where M, the number of sub-carriers in each interpolation group, depends on the characteristics of the radio channel and OFDM system.