This paper introduces a new application of adaptive control theory to power control in a code division multiple access (CDMA) cellular system operating over mobile fading radio channels. Conventional feedback power control algorithms allow the base station to send a power command to either raise or lower each user's transmission power according to a bang-bang-like control policy. In this paper, we present an adaptive minimum-variance power control methodology which can be shown to improve power control performance consistently against a random nature of the near-far effect, shadowing and fast varying fading. Two adaptive implementations are considered: direct and indirect control. In the indirect adaptive control, a minimum-variance controller is combined with a constrained estimation algorithm to ensure the stability of a link gain model. In the direct adaptive control, the controller parameters are obtained directly from a standard estimation algorithm. Our simulations have shown that the proposed adaptive minimum-variance power control schemes provide much smaller error variance than the conventional fixed-step bang-bang control scheme and consequently the reverse channel capacity of the CDMA system can be significantly increased.

The bi-level digital video, because of its simplicity and compactness, can be utilized to provide for a quick and faithful preview of its original content. The proposed bi-level digital video compression technique exploits the context-based probabilistic estimation model towards adaptive pixel prediction which can be used towards generating residual image frames which may then be Run-Length-Rice coded. Towards promoting error-resiliency and random-access, each bi-level digital video frame may be typed into either intra- or inter- picture format. The proposed technique can be seen, in comparison to existing JBIG compression technologies in simulation runs, to provide added temporal redundancy removal.