Sudden cardiac arrest, caused primarily by ventricular fibrillation, is one of the leading causes of mortality in the Western world. There is a compelling need for risk stratification to identify patients at risk for sudden cardiac arrest. Cardiac alternans, a recognized harbinger of sudden cardiac arrest, manifests as a beat-to-beat alternation in action potential duration (cellular level) or in electrocardiogram morphology (whole heart level). Although much progress has been made to understand the mechanisms of alternans, predicting and control of alternans, especially at the heart level, remain great challenges. Current approaches to predict cardiac alternans based on restitution properties of the heart are either too simple to be valid or too complex to be useful. In this work, we developed a reduced order model from the amplitude equation to investigate dynamics and control of alternans in cardiac fiber, i.e. beyond single cell level. Detailed bifurcation and stability analyses were carried out to illustrate complex spatiotemporal patterns of alternans and the limitations in feedback control due to spatial effect.