Fast linearized energy allocation for multimedia loading on multicarrier systems

Efficient near-optimal energy allocation for multimedia data transmission using multicarrier modulation is investigated. The optimal method typically relies on an iterative process that is computationally intensive, and it does not provide insight into the allocation process for higher order M-quadrature amplitude modulation (QAM) constellations. A sequence of two fast energy allocation algorithms are proposed in this work. First, a basic energy allocation algorithm is developed by solving a simple set of linear equations directly. It is shown to be near-optimal for almost the entire signal-to-noise ratio range. Based on this direct method, we can analyze asymptotic limits of the approximation error, carrier energy, and carrier rate conveniently. Then, we present an advanced algorithm that enhances the performance of the basic algorithm via iteration. The convergence of the advanced algorithm is studied, and an estimate of its computational complexity is derived. It is worthwhile to point out that, while limits for the special case with uniform noise and binary phase-shift keying modulation were studied previously, the derived asymptotic performance here works for a more general case with nonuniform noise and M-QAM constellations. Finally, we examine these limits with several types of multimedia and nonmultimedia sources, and show that our algorithms perform well regardless of the source type.