Title: LMP Command Discussion
1Project IEEE P802.15 Working Group for Wireless
Personal Area Networks (WPANs) Submission Title
LMP Command Discussion Date Submitted November
13, 2001 Source Adrian P Stephens, Mobilian
Corporation Company Mobilian Corporation Address
NW Evergreen Pkwy. Hillsboro, OR. USA. Email
adrian.stephens_at_mobilian.com Abstract This
document contains a framework for discussion of
the LMP commands necessary to manage
AFH. Purpose To stimulate discussion. Notice Thi
s document has been prepared to assist the IEEE
P802.15. It is offered as a basis for discussion
and is not binding on the contributing
individual(s) or organization(s). The material in
this document is subject to change in form and
content after further study. The contributor(s)
reserve(s) the right to add, amend or withdraw
material contained herein. Release The
contributor acknowledges and accepts that this
contribution becomes the property of IEEE and may
be made publicly available by P802.15.
2LMP Command Discussion
- Agenda
- Nature of channel metrics reporting
- LMP Commands
3Open Issues (Raised in meeting)
- Channel measurement
- Overhead of polling doesnt seem too great
- Master can respond to change in its own
environment in polling rate - No objections to polled mechanism
- Use 1 bit of slave reporting per channel, no
benefit currently demonstrated for gt1bit - Re-examination of bad channels? How? How Often?
- Capabilities
- Is channel measurement technology reported?
- Is baseband ACK usable to imply slave has
received AFH update? - Support for Partition Mapping is not included
4Channel Metrics reporting
- Optional behaviour at both Master and Slave
- Slave reports capability to master when asked by
the Master - Slave continuously measures channel metrics
- Slave reports channel metrics when polled by
Master based on current state - Master polls according to its own schedule to
meet applicable regulations - Slave characterises channels as good-or-unknown
vs bad (i.e. reports only definitely bad
channels)
5Capability Exchange
- Additional fields in the existing features
structure (2 or 3 bits needed) - Existing LMP_features req/res commands suffice
- AFH supported when operating in a slave role
capability - Channel metrics measurement reporting supported
capability (?? and classification technology) - Partition-mapping capability iff it becomes
optional - reserved bits for future use (5 bits)
6Directed Signaling
- All LMP commands regarding AFH are directed
- Reliability simplicity
- Reduced overhead over broadcast in typical small
piconets interference levels - Master updates slaves in a sequence of its own
choosing and at a time of its own choosing.
There is a transition period when the different
slaves views of the adaptation are slightly
different. - LMP_accepted LM ACK or baseband ACK of LMP packet
indicates slave is operating the adapted hopset.
(?? why did HOLD LMP command need a timer to be
added to its ACK) - Require a certain amount of overlap between
successive adaptations
7LMP_set_AFH command
- Master -gt Slave
- Specifies a mask of used/unused channels
- Implicitly enables AFH operation by the slave
- TBD Will also need to specify partition-mapping
enabled/disabled state if this is optional and
may need additional parameters. We should avoid
the need for two LMP packets per update.
8LMP_disable_AFH command
9LMP_channel_metrics
- LMP_channel_metrics_req
- Master-gtSlave poll
- LMP_channel_metrics_res
- Slave-gtMaster
- Channel map bad vs not-known-to-be-bad
- Over lt last 10 s measurements
10Behaviour
- Legacy non-AFH
- AFH-disabled (initial state of a link)
- AFH-enabled
11Behaviour
- SNIFF HOLD
- AFH-enabled state unaffected
- Master can consider the slave to be AFH-disabled
if its adaptation and the current adaptation have
no channels in common. - PARK
- Slave is implicitly AFH-disabled on park
12Broadcast
- A Beaconing access windows (park mode slaves)
unadapted hopset - B Broadcast to active slaves restricted slots
(unadapted good channels) - Broadcast to all Both A and B must be used