Efficient and fair scheduling for best-effort downlink packet data

A scheduling method to compromise between pure opportunistic (PO) and relative best (RB) scheduling is proposed for downlink packet-based data transmission, where multiple and high rates are supported by multicode CDMA and high-order modulation formats. Here, selection diversity offered by short-term fading is used to increase fairness compared to the PO, while multiuser diversity offered by independent spatial user distribution is partially used to yield higher throughput than the RB. A statistical modeling for tractable analysis is made to account for in-cell and intercell interferences, which leads to evaluation of the average rate capacity in multiple cells. It is shown that the proposed scheduling offers balanced trade-off between throughput and fairness, which overcomes shortcomings with the PO and RB scheduling.