A mathematical framework for statistical QoS and capacity studies in OFDM networks

This paper presents a mathematical modeling framework for studying the capacity of multi-cell orthogonal frequency division multiplexing networks in single- and multi-service scenarios with various quality of service (QoS) constraints. The framework is built around a feasible network load concept which relates cell resource utilization and interference-dependent resource demand generated by traffic. The feasible load problem is formulated for a general irregular network with non-uniform traffic distribution. Unlike in earlier works, the proposed model explicitly takes into account interference which is load- and service-dependent and models the service constraints on a time scale longer than the typical scheduling interval. A computationally efficient stochastic model is proposed for a regular network deployment with a user distribution pattern repeated over cells and then enhanced with QoS constraints. As an example application, the capacity region problem is studied and numerical results are presented for a realistic network setup. The framework can further be used, for example, for studying radio resource management algorithms or optimizing QoS parameters with respect to an operator QoS policy.