This paper proposes novel and generalized expressions to characterize the performance of modern cellular networks under realistic user mobility behavior. The η-µ distribution is employed to derive the received power probability density function, the average bit error rate for different modulation schemes, and the coverage probability assuming a Poisson point process spatial distribution of base stations in downlink. The user is assumed to experience fading with Maximum Ratio Combining (MRC) and move according to a random way-point mobility model. To get more insights on the achivable diversity order, accurate asymptotic expressions for the coverage probability and average bit error rate are derived. The derived expressions are applicable to different widely-used fading environments, such as Rayleigh and Nakagami-m as particular cases, by an appropriate selection of the η-µ parameters. Monte Carlo simulation was used to show the validity of the proposed expressions. In addition, the generalized expressions allow the system designer to quantify the effects of user mobility on the cellular network performance, in different propagation environments, and network topologies as a function of the number of base stations and MRC branches.