Selective relative best scheduling for best-effort downlink packet data

A scheduling scheme to compromise between pure opportunistic (PO) and relative best (RB) is proposed for downlink packet-based data transmission. For this, an instantaneous channel gain is factorized into two channel gain components such as short-term and long-term channel gains, in order to exploit individual characteristics in designing the scheduling scheme. Here, selection diversity offered by short-term gains is used to improve fairness compared to the PO, while multiuser diversity by independent spatial user distribution, resulting in distinct long-term gains, is partially used to yield higher throughput than the RB. The proposed scheme is referred to as selective relative best (SRB) and is shown to provide a balance between fairness and throughput of the system.