Modulation-Aware Error Coding

In contrast to wireline networks, wireless communications are strictly resourcelimited in every imaginable way. First of all, the wireless channel is a shared medium, which implies that its resources have to be split among multiple concurrent users (whether intended or not) at the same time and place. From a receiver's point of view, several unwanted signal components will be present at the antenna terminal, often several orders of magnitude stronger than the actual signal-of-interest. Even after aggressive filtering in the front-end of the receiver, a considerable amount of interferer power may still reside in the polished signal that is offered to the digital back-end. Furthermore, the transmitter is allowed to inject only a limited amount of power in the channel. All this, combined with conservative restrictions on bandwidth usage, make the wireless medium one of the most hostile environments for transporting information.

In the urge to get data as quick and as reliable as possible from one place to another, the efficiency of available resources is often pushed as far as possible towards the theoretical limit predicted by the Shannon theorem [Sha48]. Unfortunately, Shannon does not give any information on the way information should be attached to the channel, nor does it give any clue about the way to extract it again at receiver side. Consequently, chances are very likely that a transmission method which has proven its effectiveness in one application area will only make suboptimal use of the available resources if ported to another domain without further ado. This can be clarified by some simple examples.