Philip A. Regalia
Catholic University of America, Washington, DC
The turbo decoding algorithm [1–7] ranks unquestionably among the major practical achievements in modern telecommunications, validating the theoretical error correction capabilities of realizable codes predicted from Shannon's work .
The essence of the algorithm is to harness information exchange between lower-level receiver functions, thus demonstrating the performance advantages achievable by combining traditionally compartmentalized functions into a bigger picture. In contrast to global optimization procedures which can prove unwieldily complex, the “turbo principle” aims instead to couple basic building blocks boasting computational efficiency into iterative algorithms that transform information exchange into progressive performance improvements. The standard recipe for turbo receivers, in which extrinsic information is extracted from one block to usurp the role held for a priori probabilities within another, can be given countless voicings depending on which elements are combined; choices include synchronization, equalization, and decoding in the traditional monouser paradigm, and can be expanded to user separation, key negotiation, and access control in multiuser settings.
The intent of this chapter is to relate measurable progress in a particular instance of iterative receiver design, namely turbo equalization which combines equalization and error correction decoding. ...