Soft message relaying through chaotic analog coding

Two message forwarding strategies are widely considered in wireless relay channels: amplify-forward (AF) and decode-forward (DF). Amplify-forward is a mechanism of forwarding the reliability information extracted directly from the channel reception, without exploiting the channel code that exists in the data packet, whereas decode-forward makes use of the channel code but forwards the hard-decisions instead of the soft reliability information. Early work has shown the benefits of combining AF and DF through soft relaying, where the soft reliability message is deduced at the output of the channel decoder (at the relay node), rather than at the output of the channel. To further enhance soft forwarding requires effective protection of the soft messages. This paper exploits "chaotic turbo codes," a special type of analog error correction codes based on the renowned butterfly effect of chaos, and on the powerful parallel concatenation structure of digital turbo codes. Although the general-purpose maximum-likelihood (ML) decoder is directly applicable, the true excitement comes from a maximum a posteriori (MAP) decoding algorithm that is specifically designed for relay applications. The result is a soft-input soft-output encoder that provides the much needed protection to the soft messages (that is to be relayed), and a soft-input hard-output decoder that makes best direct estimate on the digital bits underlying the soft messages (which is the ultimate goal). The new analog-encode-forward (AEF) strategy is shown to be both simple and effective for soft relaying in user cooperation.