Cellular network with continuum priority set

We consider the following problem of spatial downlink prioritization. Mobiles arrive at a cell at locations that are determined according to some probability distribution. The farther a mobile is from the base station, the weaker is its received power and thus the lower is the transmission rate to it. Beside this uncontrollable phenomenon that differentiates between mobiles according to their location, one can design other controlled mechanisms that differentiate between them. We analyse various priority policies where the assigned priority is given in terms of the distance of the mobiles from the base station. This gives rise to a whole continuum of priority levels. We study the influence that the combined location density and priority policy have on the quality of service of the mobiles and on the network overall performance. Applying our model to a HSDPA system, we calculate a quality of service indicator, the sojourn time, using a priority scheduling strategy, a processor sharing one and a first come first served one. Considering three types of arrival flow, a uniform one, a non uniform one and a flow which generates a constant load in the cell, we show the sojourn time depends on the adopted strategy, but also on the location of the mobile and on the arrival flow type. In particular, a numerical study based on our model shows that the expected sojourn time can be improved by a hybrid policy that defines two zones in the cell and uses maximum SIR priority in one and minimum SIR priority in the other.