Dynamic MIMO Multiple-Carrier Multiple-Access: Adaptive Radio-Resource Allocation Under Realistic Constraints

This paper proposes a multiple-access (MA) radio-resource-allocation (RRA) scheme under realistic constraints: dynamic space-frequency-division multiple-access (DSFDMA), which is characterized by collectively and simultaneously considering the bandwidth assignment (subcarrier-clustering), power-allocation, beamforming vectors and user´s data-transmission-latency requirement. This scheme optimizes the multi-user MIMO multi-carrier systems through a weighted-mutual-information maximization. Applicable to diverse configurations of user-terminals, the DSFDMA conducts the RRA in a cognitive manner to address the system requirements in both of bandwidth efficiency and users´ data queue-length balancing (user-fairness). It is modeled and derived with a multiple-user MIMO-setup and characterized by an iterative gradient-based algorithm. Two major realistic constraints are introduced to the modeling: single-user power limits and the maximum number of spatial-beams supported and separable by the base-station receiver. Even though the optimization problem under such realistic constraints is not a convex or concave problem, through a multistage optimization, the designed algorithm exhibits a robust and superior performance, which is verified by a large number of Monte Carlo simulations. The scheme also possesses valuable scalability and generality.