Most existing models of mobile computing systems assume the presence of stationary base stations connected by a high bandwidth wireline network backbone. Such systems are not suitable for digital battlefield situations and emergency response operations. Hence, there is a need for mobile computing systems with mobile base stations. The flexibility and robustness of such systems is at the cost of increased complexity. Some of the problems that arise are: (i) Difficulty in providing cellular coverage to the entire service area all the time. (ii) Changing configuration of cells complicates wireless channel allocation. Also, some wireless channels are needed for communication links between the mobile base stations. (iii) Mobile base stations and the backbone links between them are resource poor compared to their static counterparts. (iv) Changing cellular configuration makes location management and communication handoffs between cells more complicated. In order to solve some of the problems mentioned above, a five step approach will be adopted: (i) Formalization of a taxonomy of mobile computing systems based on the mobility pattern of base stations. (ii) Development of stochastic models with different base station mobility characteristics. (iii) Employment of analytical results in the development ofdistributed algorithms for cell planning, channel allocation, location management, and handoff management. (iv) Simulation experiments to measure the performance of these algorithms. (v) Software implementations of the algorithms. Time and resource permitting, the software solutions will be ported to off-the-shelf computation and wireless communication devices.***
