The multitier environment of mobile communication systems is an integration of different mobile systems such as cellular, cordless, and satellite systems for mobile users to communicate with anyone, anywhere, and anytime. An effective and efficient scheme is needed for mobile terminals to dynamically select the most appropriate tier of mobile system for optimizing communication cost and performance. This paper presents a model for studying the effect of level switching in a multitier environment and for the performance evaluation of level switching schemes. This model is based on the continuous-time stochastic process, which is parameterized by a set of state-transition rates and costs derived from the characteristics of the mobile systems and terminals. Several level switching schemes based on different triggering events are proposed and the performance is analyzed using the proposed model. Level switching is found to be an effective way to reduce communication costs. The performance gain increases with the frequency of the switching activities and the level switching costs are insignificant in most cases. By using the model, the most cost-effective switching scheme can be selected to match the characteristics of individual mobile users and mobile systems for optimizing overall system performance.

In managing the locations of mobile users in mobile communication systems, the distance-based strategy has been proven to have better performance than other dynamic strategies, but is difficult to implement. A simple approach is introduced to implement the distance-based strategy by using the cell coordinates in calculating the physical distance traveled. This approach has the advantages of being independent of the size, shape, and distribution of cells, as well as catering for the direction of movement in addition to the speed of each mobile terminal. An enhanced distance-based location management strategy is proposed to dynamically adjust the size and shape of location area for each individual mobile terminal according to the current speed and direction of movement. It can reduce the location management signaling traffic of the distance-based strategy by half when mobile terminals have predictable directions of movement. Three types of location updating schemes are discussed, namely, Circular Location Area, Optimal Location Area, and Elliptic Location Area. Paging schemes using searching techniques such as expanding distance search based on the last reported location and based on the predicted location, and expanding direction search are also explored to further reduce paging signal traffic by partitioning location areas into paging areas.

Hardware-realizable learning probabilistic RAMs (pRAMs) which implement local reinforcement rules utilizing synaptic rather than threshold noise in the stochastic search procedure are described. The design allows for both global and local rewards and penalties (in this latter case implementing a modified version of backpropagation). The architecture allows for serial updating of the weights of a pRAM net according to a reward/penalty learning rule. It is possible to generate a new set of pRAM outputs at least every 100 mu s, which is faster than the response time of biological neurons.>