Null Space Learning in Cooperative MIMO Cellular Networks Using Interference Feedback

We present schemes for acquiring the null space of the interference channel between a User Equipment (UE) and an interfering Base Station Group (BSG) in Cooperative Multi-point cellular networks, whose only required network information is the interference levels at the UE. Specifically, the interfering BSG transmits a sequence of learning signals which inflicts interference on a UE served by a neighboring BSG. The UE treats interference as noise, measures its overall interference plus noise power, and feeds this value back to its serving BSG. Then, the latter distributes this information to the interfering BSG, from which it learns the null space of the interfering channel. We also present a null space tracking algorithm, whose performance includes an inherent tradeoff between the accuracy of the null space learning and the inflicted interference during learning, and characterize analytically and via simulations its performance under channel variations and noisy measurements. The proposed algorithms do not affect the transmission protocol between the UE and the serving BSG, do not add any signaling to the control channel between them, and do not require any protocol changes from the UE side.