 Now, we are going to look at some other specifications, which come in voiceover IP, which in effect determine the mobility scenarios and the handover latencies in broadband networks. The voice specifications, which start off from the overall data rate per user, the minimum is 12.2 kilobits per second Kodak. This can go higher, but it is the typical recommendation in 4G networks. And the voiceover IP connection has a 50% duty cycle or activity factor. It means 50% of the time is an active talk interval where the talk spurts actually take place and then there is some silence interval. Now the ratio of it is around 50%. The overall packets, which carry voice, the outage or the packet loss probability is around 2%. It means within a duration of 50 milliseconds, up to 2% of the total number of packets is allowed to be dropped or corrupted or even the transmission unsuccessful is permissible. Having this flexible model allows us to have a more adaptable and flexible architecture for providing voice. Now, with these specifications and in order to not exceed the 150 milliseconds desirable end-to-end latency, whenever a mobile user leaves the jurisdiction of a certain base station or a certain tower and enters into the other, the handover has to take place quickly and smoothly. So the handover latencies actually refers to the total time that is consumed in making this handover successful. Now, there are different requirements which come for different scenarios in 4G recommendations. The first one is if two different spectrum bands are utilized. For instance, if you're talking about the 4G spectrum in 800 to 900 megahertz, 1800 to 1900 megahertz, the 5 gigahertz band, the 2.4 gigahertz band, then if it is inter-spectrum handover, it is going to take the maximum amount of time, but it should still not exceed 60 milliseconds. If it is within the same spectrum, but the overall choice of frequencies is limited, so the frequency has to switch over. Even within the same spectrum, 40 milliseconds latency is the upper bound. Whereas for intra-frequency, that is the same frequency was available in the cell to which the mobile belonged to earlier and now it is moving to another cell with the same frequency. Now, it means the handover would be rather quickest, but still it should not exceed more than 27.5 milliseconds. It means that for all these scenarios, the latencies have been adjusted in such a way that the overall latencies should not exceed 150 milliseconds and should never exceed 400 milliseconds because if it exceeds 400 milliseconds, the user would have a jittery and unpleasant wipe experience.