Computationally Efficient Optimal Discrete Bit Allocation for Medium and High Target Bit Rate DMT Transmissions

Abstract

A computationally efficient optimal discrete bit allocation algorithm is proposed for medium and high target bit rate discrete multitone (DMT) transmissions. Unlike conventional greedy bit-loading algorithms, which compare the incremental transmission power of multicarriers bit-by-bit and allocate one bit at a time, the proposed algorithm is based on parallel bit-loading with reference to a virtual equal power assignment bit allocation. After maximum bit rate allocation, the algorithm achieves the target bit rate by two types of parallel bit- loading. If the target bit rate allocation is not optimal, bit swapping is performed to obtain the optimal solution. Simulation results using the standard asymmetric digital subscriber line (ADSL) test loops demonstrate the reduced computational load of the proposed algorithm compared to the existing optimal discrete bit allocation algorithm when the target rate is between 60% and 99% of the loop's maximum bit rate.