Link Ratio for Ad Hoc Networks in a Rayleigh Fading Channel

Mobility metrics have been widely recognised as a useful tool in comparing the performance of different mobility models of nodes in ad hoc networks. Most mobility metrics studied to date have been used in conjunction the transmission range model of communication. That is, it is assumed that if two nodes are within a given distance, r, then it can be said that a communications link exists between them. This assumption is based on the model of a wireless channel in which the only impediment to signal strength is attenuation related to, usually the square of, distance. This model is used extensively in both theoretical and simulation work which, naturally then, agree, bringing about a perceived strengthening of the validity of such a model. A more realistic view of the wireless channel is to assume that the signal is affected by multipath, giving a Rayleigh, Ricean or even Nakagami signalling environment. Such a view brings an entirely different perspective of signal transmission in ad hoc networks. The separation distance between a pair of nodes being within transmission range no longer guarantees that a communications link exists. This more practical approach requires a fresh look at mobility metrics. We consider the link duration, find it inappropriate for such a transmission environment and, instead, introduce a new metric, the link ratio.