We investigate performance limits of a multiple access communication systemwith energy harvesting nodes where the utility function is taken to be thelong-term average sum-throughput. We assume a causal structure for energyarrivals and study the problem in the continuous time regime. For this setting,we first characterize a storage dam model that captures the dynamics of abattery with energy harvesting and variable transmission power. Using thismodel, we next establish an upper bound on the throughput problem as a functionof battery capacity. We also formulate a non-linear optimization problem todetermine optimal achievable power policies for transmitters. Applying acalculus of variation technique, we then derive Euler-Lagrange equations asnecessary conditions for optimum power policies in terms of a system of coupledpartial integro-differential equations (PIDEs). Based on a Gauss-Seidelalgorithm, we devise an iterative algorithm to solve these equations. We alsopropose a fixed-point algorithm for the symmetric multiple access setting inwhich the statistical descriptions of energy harvesters are identical. Alongwith the analysis and to support our iterative algorithms, comprehensivenumerical results are also obtained.
展开▼
