Congestion in spread spectrum wireless mobile communication networks

This paper surveys some macroscopic measures of congestion that pertain to spread spectrum networks which are used in network control to decide whether or not to admit a call. Such measures reflect the congestion in the region of the network that the call will interact with during its lifetime as opposed to properties of the user itself at call set up time which are of little value due to temporal variance. By developing network-wide measures of congestion for spread spectrum systems, the task of call admissions is much simplified. Three models of spread spectrum which are studied are network of cells employing CDMA, interference cancellation and a macrodiversity model. The most startling result here is that measures of congestion can be given in terms of eigenvalues of matrix defining the network geometry. Though it is a macroscopic quantity, it is measurable by any user or cell site in a decentralized way.