Grant: Alliance Grants

We propose the definition of models to reconstruct channels from limited User Equipment (UE) measurements. We will use existing feedback that has been collected from UEs at Ericsson communicating over real networks. In collaboration with the team at Ericsson we will look for different methods (e.g., channel reciprocity) to acquire labeled data sets of limited UE feedback and corresponding channel coefficients to be used in developing the reconstruction models. We will develop a simulation tool to reconstruct temporal-spatial channels based on limited CSI feedback that has been collected from UEs communicating over real networks. The models will study spectrum sharing and assignment to increase the efficiency of spectrum reuse. This includes analysis of resource and power allocation to increase spectrum efficiency and achieve other objectives, such as meeting Quality of Service (QoS) demands for different types of users. An important research item is the study of spectrum sensing methods, including spectrum awareness techniques to adapt to dynamic radio environments. We will propose new algorithms to detect and classify signals in low Signal-to-Interference-plus-Noise Ratio (SINR) scenarios. We will use a mix of real and simulated signals to extract features of different types of signals over the spatial and time dimensions. We will propose distributed algorithms that will be evaluated in terms of performance and compare the results with existing algorithms for resources allocation. We will also evaluate and compare the performance of centralized and distributed autonomous approaches for resource and power allocation.The performance as well as the complexity of such methods will be evaluated to validate their efficiency and usability.