| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
|
To follow-up on our last teleconference, I had promised to look into
some characterization or simple analytical measures that can be used with the handover call flow template. My initial thoughts were to use this handover call flow or event sequence chart in order to get a rough approximation on - how many messages/events are needed to complete the handover. - how long it takes to complete the handover based on some agreed upon time estimates. As an example of how this might be done, I defined three types of messages and events as follows: 1- Lower Layer (LL): triggers and events to and from the MAC and PHY layers that are local to a device 2- Higher Layer (HL): triggers and events to and from the application layers that are local to a device 3- Media Independent Handover Layer (ML): triggers and events that are local to the MIH layer on a device. 4- Remote (R): messages that are sent between devices. By inspection of the flow chart one should be able to compute the number of LL, HL, ML, and R events and messages. Also, by associating a time value - for example in units of time (u.t.) - to send, receive and process events, one can have a rough approximation on the time it takes to perform the handover. Again as an example of how this might work: 1-Tll: the time it takes to process LL triggers in unit of time: 1 u.t. 2-Thl: the time it takes to process HL triggers in unit of time: 5 u.t. 3-Tml: the time it takes to process ML events in unit of time: 2 u.t. 4-Tr: the time it takes to transmit, receive and process remote messages: 10 u.t. Summing over all messages of all types, we should be able to get an estimate for the time (in u.t.) Note that counting events/messages is done per device. A total number of messages/events can be done for all devices, however the time estimate would have to deal with events happening in parallel. Also, be aware that this is a *very rough* characterization and should not be used in absolute terms. But if everyone uses the same template, then it would be easier to compare between different mechanisms. The excel sheet included in this email is simply an example using the call flow charts provided in contribution 21-04-0162, Figures 14, 15, 16 & 17. The time dimension is not there yet, because I am still debating whether (1) an excel spread sheet makes sense and (2) this is a useful excercise. In case we decide to proceed with this, we could create excel templates for people to use, and add the timing information. I look forward to your feedback. Note that I will be away starting December 20, 2004 until January 4, 2005. We can discuss this in more details during our next telecon. Best regards and Happy holidays! -nada -- Nada Golmie, Ph.D. Manager, High Speed Network Technologies Group National Institute of Standards and Technology 100 Bureau Dr. Stop 8920 Gaithersburg, MD 20899 Email: nada@nist.gov Phone: (301) 975-4190 Fax: (301) 590-0932 Web: http://w3.antd.nist.gov |