Design of predictive IIR filters via balanced model truncation

Recently the design of polynomial predictors has had great attention from many filter designers. The results produced have indeed found many applications in industry. These new filters consist primarily of finite impulse response (FIR) systems. Other flavours of infinite impulse response (IIR) have also been developed. In this paper we look at the potential offered by the balanced model truncation (BMT) in designing a new IIR favour of polynomial predictors.