Network and single user performance evaluation of a mobile data system over flat fading transmission channels

In this paper the contrasting effects of transmission impairments and capture on both the network and single user performance of a slotted Aloha system are investigated in a mobile radio environment, accounting for frequency non-selective random propagation phenomena, and employing the packet error probability in order to define packet losses and capture. With this study we demonstrate that it is possible to generalize in a real propagation context a method previously proposed in literature for evaluating the network behavior in terms of steady-state throughput, backlog and stability in conventional transmission conditions, i.e., when all the transmission channels were error-free and the collisions caused the loss of all the packets involved. Moreover, we indicate under which specific constraints on the terminal mobility we can apply this method to analytically predict the single user performance, which we show being an important design parameter. Through the numerical results reported we quantitatively point out that in the absence of coding capture increases system stability and moderately improves the overall system throughput and backlog. We also outline the trade-off between an increased capture gain obtained by means of coding and the corresponding system cost in terms of complexity and bandwidth occupancy. Furthermore, we demonstrate the unfairness which affects the single user performance and the consequent need for countermeasures in order not to discriminate among users differently located within the network; yet, this last solution is detrimental as regards the positive capture effects.This work was supported by M.U.R.S.T. and C.N.R. (Italy).