Multi-stream information transmission over K-user interference networks

In this paper, communication over K-user interference networks is discussed. Each user encodes information as a number of redundant binary data streams which are concurrently transmitted over the channel at various power levels. The reception is performed in two stages: in the first stage, the data streams are separated by an iterative detector; they are individually error-control decoded in the second stage. Conditions for decoding convergence in terms of the density of data streams per power level are formulated. The technique is illustrated through the use of several examples, including one-to-many and many-to-one interference channels. It is proven that optimal degrees of freedom can be achieved for a number of interference channels and channel parameters.