The handoff in Mobile IP networks causes packet sequence disruption during a packet forwarding procedure and may result in performance degradation in higher layer protocols. We investigate the impact of handoff in the Mobile IPv6 networks, where an optimized routing with the smooth handoff is adopted. The impact on the packet sequence is measured by an 'unstable time period (UTP)' and a 'silence time period (STP).' The UTP explains the time duration of out-of-sequence packets while the STP reflects the blackout duration of a mobile node after the initiation of handoff procedure. With the analysis on the UTP and STP, the total transient time period (denoted as handoff time period or HTP) after the handoff initiation can be estimated. In our previous work, focusing on the UTP, the packet flow sequence under the smooth handoff is analyzed for the Mobile IPv4 networks. The proposed queuing-based analysis is extended in this work for the Mobile IPv6 networks. That is, several modifications are made to conform to Mobile IPv6 and at the same time the queuing analysis itself is improved to better model the handoff procedure. The numerical results show that the queuing delay for the handoff packets (affected by background traffic) and the involved link (or route) capacities affect the estimated UTP, STP, and HTP. In addition, two schemes such as priority queuing and buffered packet forwarding are introduced to reduce the transient period and the improvements are analyzed for comparison.

A workflow is a flow of work carried out by workers, and workflow management is to automate the flow of work. In workflow management, an actual work is carried out based on the workflow, which is called case. In order to effectively meet various requirements, it is necessary to change current workflow dynamically, which is called dynamic workflow change. When the dynamic change is required, there exist cases in the workflow. In order to handle these cases and further to keep the queuing order, the dynamic change takes period of time (called transient time) until the changed workflow becomes steady state again. During the transient time, workers are forced to do irregular work, and therefore it is important to clarify if a change type takes shorter transient time. In this paper, we do the performance evaluation on transient time of dynamic workflow changes. To do so, we first give a definition of transient time, and then propose methods of computing transient time of three change types proposed by Ellis et al. Finally, we do the performance evaluation for 90 dynamic changes by computing the transient times.