Recent attention in resource allocation algorithms for multiservice CDMA networks has focused on algorithms optimizing the uplink throughput. For a given set of real time (RT) and non real time (NRT) communications services, an optimal uplink throughput can be obtained. In this paper we focus on the optimization of the downlink throughput. New constraints for the allocation policy are detailed. We present the maximal downlink throughput and propose two different algorithms optimizing different criteria and leading to solutions suited to the particular orthogonal variable spreading factor (OVSF) policy of wideband code division multiple access (WCDMA) access technology. For each algorithm the distance between the throughput obtained and the maximal throughput is quantified.

We show how orthogonality constraints for DS-CDMA spreading sequences yield directly to OFDM sequences. Having shown this duality we do a comparison of performances between these two waveforms in a multipath fading channel context. The comparison includes the MC-CDMA approach. We show how the cyclic prefix solution can be used for DS-CDMA in order to propose an MMSE optimal despreading solution. This first part of our study is based on a perfect channel estimation. The second part includes channel estimation uncertainty in the comparison. Parameters of simulations, presented in this paper, correspond typically to an UMTS downlink transmission.

In this paper, a wireless communication network providing data services is considered. First, the total number of possible allocations a resources' administrator of such a network is confronted with at each decision time is computed. The obtained number obviously shows that it is too considerable to allow the administrator choosing, among all potential allocations and in a real-time environment, the allocation estimated as optimal. Therefore a method based on an arranged subset of the whole set of allocations is proposed to decrease the number of necessary computations every period of decision. The gain of computation, increasing with the number of users and resources of the network, is evaluated and the method is illustrated by an example.

The radio spectrum is a unique, ubiquitous natural resource shared by various types of services. Unlike many other natural resources, it can be repeatedly reused. In practice, though, it remains a finite resource, and it can only accommodate a limited number of simultaneous users. This restriction requires careful planning and management for maximizing its value for all services. This is especially true since the demand for communication spectrum worldwide is rapidly increasing. As the pressure on spectrum increases with a generalized image of radio techniques, traditional ways to assess the merits of new technical solutions and to allocate as well as to tax frequencies appear to be adapted less and less. This paper reviews the present situation and stresses the need for more proactive, harmonized, and dynamic approaches linking technical and regulatory aspects in order to introduce broadband wireless communications systems.