Optimizing vertical common subexpression elimination using coefficient partitioning for designing low complexity software radio channelizers

The complexity of finite impulse response (FIR) filters used in the channelizer of a software defined radio (SDR) receiver is dominated by the complexity of the coefficient multipliers. A method for designing low complexity channel filters by optimizing vertical common subexpression elimination (VCSE) using coefficient partitioning is presented. Our algorithm exploits the fact that when multiplication is implemented using shifts and adds, the adder width can be minimized by limiting the shifts of the operands to shorter lengths. Design examples of the channel filters employed in the digital advanced mobile phone system (D-AMPS) receiver show that the proposed method offers considerable full adder reduction over VCSE methods.