A Cross-Entropy based subcarrier, bit, and power allocation algorithm for multiuser OFDMA

A Cross-Entropy (CE) based subcarrier, bit, and power allocation algorithm is studied in this paper for orthogonal frequency division multiple access systems, where the total transmit power is minimized under the given quality of service requirements such as data rate and bit error rate. Although those power allocation problems can be translated into a non-deterministic polynomial time hard combinatorial optimization problem via coding the assignment information into a binary sequence, the globe optimal solution cannot be found in polynomial time. To address this problem, cross-entropy method is adopted to search a sub-optimum solution of the allocation type. The simulation results show that our proposed CE method can find a lower overall transmit power than genetic algorithm.