Multi-User Resource Allocation For Multi-Carrier DS-CDMA System

In the past few years, the resource allocation optimization problem in communication systems has drew growing attention due to its importance in improving system performance and stability. Previous researcher does usually treat it as an optimization problem to the utility function. There are two kinds of utility function:concave function and non-concave function. The optimization of the concave function has been studied extensively for many years. Unfortunately, most of the systems in reality have non-concave function, which violates the basic assumption of the previous research and makes most of the previous methods unsuitably. For the existing algorithms using the non-concave function, in order to simply the system model, the utility function is often simplified as an S shape function, leading to an over-simplification of the real problem.In this thesis, the real packet throughput has been selected as the system utility function. Based on this model, it has done extensive research on the resource allocation problem in MC DS-CDMA systems, and further introduced some characteristics in military communication systems to make our algorithm more practical, for example, jamming, imperfect channel state information (CSI), complex network topology and spectrum detection. Due to the specialty of uplink in MC DS-CDMA systems, here the resource allocation problem in a distributed MC DS-CDMA network has been considered directly, and proposed a new resource allocation algorithm in MC DS-CDMA systems with low complexity. Besides, in the last chapter, the video transmission, which has been rarely studied in previous work, was studied and a cross-layer resource allocation algorithm has been proposed in a distributed network.In the downlink data transmission, we studied the recourse allocation problem with both perfect and imperfect CSI, and proposed an iterative bisection algorithm with low complexity. Besides, in each iteration, a closed-form expression was derived for the transmit power, subchannel, and modulation assignment, which significantly reduced the computational complexity in the second chapter.Due to the difficulty of sharing information in the distributed data transmission network, the resource management is usually done based on the individual information. Here, a new distributed iterative resource allocation algorithm has been proposed with low complexity, which may not only improve the instant system performance, but also enhance the average system performance in the long term in the third chapter.In the fourth chapter, we derived the analytical expression of the real throughput in MC DS-CDMA systems with imperfect CSI. One mistake in the current research has been found:the error introduced by the imperfect CSI is proportional to the transmit signal power. It is different with existing literatures in the sense of regarding the noise as a simple Gaussian distributed random variable.In the end, we extended our recourse allocation algorithm to the video transmission scenario, and proposed a simple cross-layer design for the resource allocation in a distributed videos transmission system. A new distributed iterative optimization algorithm has been proposed using the decomposition method and the water-filling power algorithm.In this thesis, the proposed resource management algorithms for the data or video transmission in the multi-user MC DS-CDMA network can be easily used in both military and civil communication systems.