Research On Location Management In Unstructured Cellular Mobile Computing Environment

With the rapid development and broad application in Internet, wireless mobile communication and computing, novel technologies of mobile computing continuously boom, and now it has been one of the important direction in scientific research of information technology and industrial developing.The core of mobile management—location management, is one of the most significant challenge in mobile computing. To design a logical, efficient and optimized strategy of location management in wireless mobile network, and give a set of effective arithmetic and wise prove in theory will be very important to theory and practice in mobile computing.Currently, most strategies of location management are based on the hypothesis that the cellular network is restricted to structured cell configurations such as the same size, shape and distributing. To the above-mentioned phenomenon, we bend ourselves into the research on the location management based the arbitrary cell topology, which contains different sizes cells: macrocells, microcells and picocells. The terminal will compute the physical distance (meter or kilometer) between the cell which is currently resided in and the last reported cell when registering or updating to supply the gap that it must know the whole cellular network topology to compute the distance based the numbers of cells, by regrouping cells and comparing the CCs (Cell Coordinates) which are broadcast by each base station. Being based on the terminal's character of moving ahead, the thesis proposes an advanced distance and direction angle based dynamic location update algorithm, which will adaptively adjust the fan-shaped LA based on the history, call arrival rate, mean velocity, cell size and so on.Due to adopt the individuated fan-shaped LA to reflect the movement character of terminal, we take the one-step paging strategy with the least delay. We describe how to formulate the total cost of location update and paging between two call arrivals as a semi-Markov decision process. Finally, we analyze the results by numerical values and draw a conclusion that individuated fan-shaped LA is better than circular LA for location tracking problem when the terminal has a personal moving character.