Utility-based adaptive multi-frame iterative algorithm for resource scheduling in OFDMA networks

This paper focuses on the downlink resource scheduling with quality-of-service (QoS) provisioning in orthogonal frequency-division multiplexing (OFDM) networks. In homogeneous networks, the multi-user diversity gain is achievable through exploiting the independent fading among users. However, in heterogeneous networks, the different priorities and QoS constraints of heterogeneous services, especially delay sensitive services, bring to inferior system performance. To solve this problem, a utility-based adaptive multi-frame iterative (AMI) algorithm aware of both channel and queue state information is proposed. The AMI algorithm contains two parts: the core and the iterative shell. In the core, the problem of dynamic subcarrier assignment (DSA) is modeled as the weighted independent set (WIS) problem in graph theory. Through joint optimization among subcarriers, a higher spectral efficiency is achieved. In the iterative shell, intra-frame and inter-frame iterations exploit the time diversity, thus a lower packet loss is achieved.