DFS Update from the European Union

1
DFS Update from the European Union

• Date 2008-3-18

Authors
2
Dynamic Frequency Selection

• In the EU, EN 301 893 defines regulatory
  requirements for RLAN operation in the 5GHz band
• DFS was devised to protect radar operating in
  that band
• Radar and RLANs are co-equal users of the band
  (EU)
• RLANs detecting radar patterns defined by 301 893
  must move to a clear channel
• Weather radar signatures were not included in the
  initial (or second) version of DFS (v1.3.1 and
  v1.4.1)
• The EU has started to receive interference
  complaints from the meteorological community
  (EUMETNET)

3
EU Weather Radar

• EUMETNET Report to the Wi-Fi Alliance
• Philippe Tristant attended the Prague session
• Government weather radar needed to protect
  populace from serious weather conditions
• Serious interference issues seen
• Petitioned TCAM to help protect weather radar
• TCAM decisions
• Changes to EN 301 893 in phases
• Immediate remedy for the radars banning RLANs
  from 5600 5650MHz band until new patterns,
  pulse detection becomes mandatory

4
EU Regulatory

• The TCAM Committee
• is the Telecommunication Conformity Assessment
  and Market Surveillance Committee
• Role is defined within the Directive 1999/5/EC
  (RTTE)
• See articles 13 to 15 of the Directive 1999/5/EC
• Formal procedures article 5 provides for
  safeguards against shortcomings in standards.
• The Commission (after opinion of TCAM) can
• Withdraw standards
• Interpret standards or give guidelines on the
  conditions under which they give presumption of
  conformity
• Informal procedures try to manage issues through
  consensus
• TCAM issues Opinions based on proposals made by
  the representative of the Commission (TCAM Chair)
• Common understanding on the way forward

5
TCAM Decisions

• Document TCAM24(24) from the EC
• Meteo radars interfered by 5 GHz RLAN
• EC concluded DFS has shortcomings
• Industry need to address these before 5 GHz RLAN
  technology is suitable for mass market
  development
• Staggered PRF
• Narrower Pulses (0,8 µSec / 0,5 µSec)
• Solution for noise calibration of weather radars
  (10 min CAC or equivalent)
• only for the band 5600 - 5650 MHz
• Other radars have comparable characteristics,
  hence Staggered PRF and Narrower Pulses should be
  applied across the whole band.
• EC conclusions/proposals above endorsed by TCAM

6
DFS Changes

• Now RLANs can not operate in the 5600 5650 MHz
  band
• After 3/31/08 must meet EN 301 893 v1.3.1 or
  1.4.1 requirements
• After 6/30/08 must be able to detect staggered
  PRF (3 PRF values) in 5600 - 5650 MHz band or
  block the band (channels 120, 124 and 128)
• After 3/31/09 must also meet EN 301 893 v1.5.1
  requirements
• Detect 0.8uS pulse width across 5250 5350 and
  5470 5725 MHz bands
• Detect staggered PRF across 5250 5350 and 5470
  5725 MHz bands
• Include a solution for weather radar noise
  calibration period (10 min CAC)
• After 6/30/10 must meet EN 301 893 v1.6.1
• All the 1.5.1 requirements
• Detect 0.5uS pulse width across 5250 5350 and
  5470 5725 MHz bands

7
Proposed Relaxation of the CAC Requirement

• Non-Continuous Off-Channel CAC (for all DFS
  channels)
• as an alternative for the continuous CAC
• Scheduled CAC for weather radar channels
• Avoid the weather radar band at start up and do a
  CAC later on or do an Off-Channel CAC while
  operation has started on a channel outside
  5600-5650
• CAC time for weather radar band
• Full 10 min CAC (or equivalent Off-Channel CAC)
  only first time a channel in 5600-5650 MHz is
  being checked.
• If no radar found, subsequent CAC checks should
  only be 60 sec

8
Proposed Relaxation of the CAC Requirement 2

• Relaxation on the Uniform Spreading requirement
• Use 60 of the available spectrum in stead of 80
• Impact is lt 1.5 dB
• Some applications do not require all channels and
  as such they should not be impacted by the 10 min
  CAC
• Relaxation on the Channel Closing Transmission
  Time
• Proposed 1 sec in stead of 260 mSec
• required for mesh networking, other applications
  will not need this
• DFS Threshold level
• Threshold level as a function of the Power Density

9
Some Things We Can (Should) Do

• TCAM has set the DOW for version 1.4.1 to 1 April
  2009
• ETSI should progress the work with the objective
  to finish revision of EN 301 893 latest mid 2008
• National voting processing time ETSI
  processing time EC gt 6 months
• Publication in OJEC should be at least a few
  months prior to the DOW for version 1.4.1
• Once ETSI has sent the new EN 301 893 into
  voting, industry should ask TCAM to revise the 1
  April 2009 date and move it e.g. to end of 2009
  or mid 2010
• The EC is firm with regard to what should be
  included in version 1.5.1
• So once we have done 1.5.1 as they requested, we
  believe we can get a relaxation on the time line
  to give industry more time to implement 1.5.1
  into their products