| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
Thanx Vivek for the document. I could not take part in the latest conf call but could see that you guys have achieved good job with the document already taking pretty good shape.
I will not be able to attend the Portland meeting but would like to make a couple of remarks to the reqs document as well as to the minutes of the 29th of June conf call . Therefore, below is my contribution to the 13th of July discussion for finalizing the document:
Thanx in advance for taking tkem into account
Section 2.1 Seamless Handover
We need to precisely define what we understand by a "seamless handover"
I would like to suggest this definition of seamless mobility to .21
"Seamless handover refers to the ability to provide users with uninterrupted, albeit sometimes transcoded communication service presentation when they move between access networks and terminal interfaces. The change shall not be noticeable to the communication service except if the experienced quality of service can not be maintained. Further, Such a handover shall require minimum interaction by the user, who should nevertheless be informed of the handover decision made for him/her."
Section 2.3 Quality of Service
Agree that "There shall also be means to map QoS requirements between different media access technologies"
But there shall also be means to transfer the associated contexts!
Section 2.4 Network Discovery
The events leading to detection of network can not trigger the handover process by themselves. What will trigger the handover process will be the learned capabilities (via (information discovery).
Section 3.1 layer 2.5 model
Actually PHY and MAC information/event from my point of view are to be delivered to layer 2.5 that will in turn pass them to higher layer entity, rather than having each layer communicating with higher layer protocols. Otherwise what would be the need to define such an interface for information presentation? If information is still passed transparently across it?
Section 3.2 Layer 2 trigger
Layer 2.5 is NOT a client side only feature and should be extended to the network side
Otherwise, if the decision making entity is in the network how could the information availbale from the layer 2.5 of a host be understood by that entity ?
Have a good meeting! and see you in Berlin hopefully
Eric Njedjou