(A) study on downlink transmission scheduling in CDMA data networks

Abstract

In this thesis, we consider throughput maximization problem with fairness constraints in CDMA data networks. The throughput is a measurement of the capacity in data transmission hence throughput maximization problem is one important problem in CDMA data networks. In order to maximize throughput, optimal transmission scheme is one by one transmission during a certain time duration. Considering this, we formulate the throughput maximization problem with fairness constraints to a transmission rate scheduling problem and this problem becomes stochastic optimization problem since the interference from other cells to each user in a certain cell is random. Also, to measure the efficiency of proposed scheme, we propose a scheme for hypothetical model where inter-cell interference is assumed to be known. We suggest two heuristic methods to solve this problem. First, we propose a scheme that preferentially assign to the minimum throughput requirements for all users to strictly guarantee the requirements for all users. And next, we suggest another scheme that employs the soft-fairness policy since the inefficiency is occurred in strictly satisfying the requirements for all users. Although some users do not satisfy their requirements, if the unsatisfied amount of each user is too small then it is not serious problem so this amount could be covered in the next scheduling time duration. By numerical simulations, the second proposed scheme is more efficient than the first proposed scheme with respect to throughput, and the unsatisfied amount of each user is shown to be quite small when the second scheme is applied.