IEEE%20802.15%20<MAC-Guard-Time>

1
Project IEEE P802.15 Working Group for Wireless
Personal Area Networks (WPANs) Submission Title
IEEE802.15.3 Guard Time in the 802.15.3
MAC. Date Submitted 19 September,
2001 Source Dr. William Shvodian Company
XtremeSpectrum Address 8133 Leesburg Pike, Suite
700, Vienna, Virginia 22182 Voice
1.703.269-3047, FAX [phone], E-Mail
bshvodian_at_xtremespectrum.com Re Abstract
This proposal describes the use of Guard Time for
the 802.15.3 MAC. Purpose To provide an
improvement to the current version of the
802.15.3 MAC Notice This 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.
2
Overview

• This presentation describes the use of Guard Time
  for 802.15.3
• It also describes Guard Time modifications for
  static GTS slots

3
Guard Time
Beacon
CAP
GTS0
GTS1
GTS3
GTS4
Beacon
GTS2

• In a TDMA system, Guard Time is needed between
  slots to keep the transmissions from two stations
  colliding.
• Each station synchronizes to the beacon, but
  clock inaccuracies result in drift
• The maximum drift between the ideal time and a
  stations clock is

4
Guard Time Position
Beacon
CAP
GTS0
GTS1
GTS3
GTS4
Beacon
GTS2

• Guard time is time between adjacent slots
• Guard time can be defined as being at
• the beginning of the slot
• the end of the slot
• half at the beginning of the slot and half at the
  end of the slot.
• For Simplification, the basic Guard Time for each
  slot in 802.15.3 will be at the end of the slot.
  The slot time in the CTA will denote the
  beginning of a GTS. Guard time occurs at the end
  of the slot, before the start of the next GTS.

Guard
Guard
G
G
5
Worst Case Timing (Transmission)
drift
drift
Ideal GTS n position
Ideal GTS n1 position
Late Stations Estimate of GTS position
Early Stations Estimate of GTS position
SIFS
Guard Time

• The above picture illustrates the maximum offset
  between two stations. The station with slot n
  has drifted the maximum time to the right (late),
  while the station assigned to slot n1 has
  drifted the maximum amount to the left (early).
• In order to avoid transmission collisions or
  violating SIFS, the guard time must be calculated
  to be two times the maximum drift plus SIFS

6
Worst Case Timing (receive)
Early Station Transmits
Ideal GTS n position
Late Stations Estimate of GTS position
Guard Time
Guard Time
Late Stations Receive Window

• Since a station doesnt know whether it is early
  or late, it must start listening one full Guard
  Time before where it estimates the start of the
  GTS slot to be and it must listen until the time
  that it estimates the start of the next GTS slot
  to be - it must listen during the time that it
  estimates to be Guard Time - on both sides of the
  GTS slot.

7
Constant vs. Variable Guard Time
Beacon
CAP
GTS0
GTS1
GTS3
GTS4
Beacon
GTS2

• The amount of guard time is a function of the
  amount of time from the reference event to the
  event of interest
• More guard time is actually needed at the end of
  a superframe than the beginning because it is
  farther from the beacon, which provides the
  reference timing. For simplicity, the same guard
  time will be used for the entire superframe in
  802.15.3.

8
Guard Time for static GTS slots
Beacon
CAP
GTS0
GTS1
GTS3
GTS4
Beacon
GTS2

• Static GTS slots allow a station to transmit in a
  superframe when it did not receive the beacon
  correctly.
• If a device with static GTS slots receives the
  Beacon header correctly, it can synchronize to
  the beacon header
• A device with static GTS slots does not need to
  correctly process the CTAs in the beacon.
• If a station with static GTS slots does not
  correctly receive the beacon header, additional
  guard time is needed.
• Dynamically changing the guard time to account
  for missed beacons requires adding guard time at
  the start of a GTS and at the end of the GTS and
  increases complexity.
• Fixed guard time can be used to account for the
  worst case maximum missed beacons. This is
  simpler and will be recommended for 802.15.3

9
Guard Time Calculation when Static GTS allowed

• Guard time is calculated by the PNC
• Guard Time is a new field in the Piconet
  Synchronization Parameters Information Element.

10
Guard Time Overhead
11
Guard Time for 802.15.3

• New Field in the synchronization Information
  Element
• GuardTime