Resource allocation in multi-relay multi-user OFDM systems for heterogeneous traffic

HighlightsResource allocation with subcarrier-pairing under coexistence of RT and NRT traffic.Optimization of relay-user selection, subcarrier-pair assignment and power allocation.Flexible QoS guarantee of real-time traffic by system parameter adjustment adaptively.Total power constraint uses power resource better than individual power constraints. Resource allocation is an important issue that has not been well resolved in multi-relay multi-user Decode-and-Forward (DF) OFDM systems with subcarrier-pairing at relays in the presence of heterogeneous flows, i.e., simultaneous real-time (RT) and non-real-time (NRT) traffic. In this paper, we address the issue by first formulating it as a joint optimization problem of relay-user pair selection, subcarrier-pair assignment and power allocation and then solving it through dual decomposition and subgradient methods. Asymptotic optimal iterative algorithms with polynomial complexity are proposed with the objective of maximizing sum-transmission-rate of NRT traffic while providing quality-of-service (QoS) guarantee for RT traffic, in two cases: (i) with total network power constraint and (ii) with individual power constraints. Effective selections of initial dual variables and stepsize for the subgradient method are also presented. Simulation results are provided in multiple scenarios and they show that: (i) the proposed algorithms outperform the existing schemes in terms of providing QoS requirements of RT users and maximizing sum network transmission rates; (ii) algorithms designed under total power constraint exploit the power resource much better than those under individual power constraints.