Improving wireless network capacity by introducing logical discontinuous coverage

Mobile cellular networks are now largely deployed, and mobile customers are used to place and to receive calls at any location. The network coverage is continuous, but radio conditions met by users are not homogeneous. Thus, data rate is varying according to user mobility. In this paper, we propose to discriminate terminals by their radio conditions in order to select which data has to be sent. More precisely, terminals are supplied with data only when they are in zones of the cellular network offering good radio quality. With such an approach, continuous cellular networks are exploited in a logical discontinuous way. In order to hide this logical discontinuity, we propose to introduce a network cache between the mobile terminal and the content server. Its role is to store incoming data sent by the server even when the terminal is not under good radio conditions. Then, as the terminal enters into a zone of the highest radio throughput (a transfer area), it starts storing data in its cache, as fast as the radio conditions allow. Then data are consumed to continue running the service up to the next transfer area. Our approach is based on a smart scheduling solution (1) that favors data transfer when users have a good radio link, (2) and that regulates the flow of data between the network cache and the terminal according to the available bandwidth. Simulation results highlights the improvements with QoS policies guided by radio conditions for streaming services. Results are promising, a cache-based architecture contributes to a good service quality with no interruption. Furthermore, simulations show that this approach can increase significantly users density supported by the network