In the current example, the duratio

In the current example, the duration of a time interval is set equal to 1 h, which indicates the within-day dynamics of the stochastic learning process, and only one VOTT user class is considered. The maximum real-time information penetration rate is assumed to be I= 25%. Relatively smallθvalues result in larger dispersion in the perceived costs and, hence, larger spread of users among the available paths and vice-versa. Thus, for the users who receive and follow realtime information, a relatively large value of the cost perception (dispersion) parameter is preferred (θ?−1=2), which indicates that they have an accurate perception of the prevailing conditions on network routes and choose the best ones. The remaining users employ historical path cost information, which corresponds to the path costs occurred in the same period of the previous day (k= 24) and denotes the day-to-day dynamics of the stochastic learning process. Due to their imperfect knowledge on the currently prevailing network conditions, these users may choose many alternative routes and a lower value of the cost perception parameter is selected for them (θ?−1=0.5).