A PIPELINED ADAPTIVE LATTICE FILTER ARCHITECTURE

The stochastic gradient adaptive lattice filter is pipelined by the ap plication of relaxed look-ahead. This form of look-ahead maintains the functional behavior of the algorithm instead of the input-output mapp ing and is suitable for pipelining adaptive filters. The sum and produ ct relaxations are employed to pipeline the adaptive lattice filter. T he hardware complexity of the proposed pipelined adaptive lattice filt ers is the same as for the sequential filter, and is independent of th e level of pipelining or speedup. Thus, the new architectures are very attractive in terms of implementation. Two pipelined architectures al ong with their convergence analyses are presented to illustrate the tr adeoff offered by relaxed look-ahead. Simulation results supporting th e conclusions of the convergence analysis are provided. The proposed a rchitectures are then employed to develop a pipelined adaptive differe ntial pulse-code modulation codec for video compression applications. Speedup factors up to 20 are demonstrated via simulations with image d ata.