Dynamic real time distortion management over multimedia downlink CDMA

In CDMA networks it is possible to increase the maximum number of users of real time applications by smoothly increasing each end-to-end distortion. We study a protocol for downlink CDMA that allocates source coding rate, channel coding rate and transmitted power based on each user´s channel state. The allocation of resources is performed with the constraints of users´ maximum distortion and base station´s maximum transmitted power in such a way that users operate under carefully controlled distortions. The optimization goal is to minimize the overall distortion in downlink single-cell systems. When the system is highly loaded, the idea for the proposed solution is to initialize the resource allocation with that corresponding to maximum distortion for all users, and then incrementally allocate the remaining transmitted power quota to the user that, at each step, can reduce the distortion with the least use of power. We show that this user must have a high distortion, experience a good channel condition, or generate small interference. We also extend our study to include speech´s talk spurt-silence nature and to consider a dynamic model with different offered loads. From simulation results, the proposed algorithm fundamentally reduces the distortions and the necessary maximum transmitted power when the number of users is large, compared with a traditional CDMA scheme (with no distortion control).