Embed Size (px)
Citation preview
Evaluation of 1588v2 for Deployment.Presenter: Tommy Cook, CEO Calnex Solutions Ltd
COMPANY CONFIDENTIAL 2
Presentation overview
• Items for evaluation.
• Protocol Behaviour
• Performance in Network Congestion
• Evaluation Plan.
• Current methodology.
• Enhanced methodology and Goal for the Future.
Want aspects and
behaviours do I need to
evaluate?
I want to deploy 1588v2 to
transfer timing across my
network.
How do I prove
equipment is fit for
deployment?
COMPANY CONFIDENTIAL 3
Proving 1588v2 clock transfer
Clock Output must comply with the
relavent ITU-T clock specification
(MTIE & TDEV specification);
• G.81x series of specifications
• (G.823/4 for TDM delivery.)
Master
Clock
Slave
Clock
t-sm
t-ms
Follow_Up
Sync
Delay_Req
Delay_Resp
Propagation Delay
Message Exchange
t1
t4
t2
t3
895.477518885894.977768110894.478017335Offset=0.002 ppm
-0.000296349
0.000645879
Time IntervalError (TIE)vs Nominal(seconds)
MKR-1:x=894.874691063, y=0.000360070MKR-2:x=894.874691063, y=0.000360070
Delta: x=0.000000000, y=0.000000000
x=895.477518835
y=0.000444870
T (a)
Clock
Output
COMPANY CONFIDENTIAL 4
Protocol Behaviour
COMPANY CONFIDENTIAL 5
Session Management Problems
•Are the Session
Management protocol
stacks robust to
unexpected events?
•Will they always
recover and continue
as expected?
Master
Clock
Slave
Clock
Session Start-up
.
.
Master
Clock
Slave
Clock
.
.
.
.
Session Status
COMPANY CONFIDENTIAL 6
Handshake Sequence problems.
•Are the protocol stacks
robust to unexpected
events?
•Will they always recover
and continue as expected?
Missing Sync message.
Missing Delay_Response.
COMPANY CONFIDENTIAL 7
Network Delay & PDV
COMPANY CONFIDENTIAL 8
895.477518885894.977768110894.478017335Offset=0.002 ppm
-0.000296349
0.000645879
Time IntervalError (TIE)vs Nominal(seconds)
MKR-1:x=894.874691063, y=0.000360070MKR-2:x=894.874691063, y=0.000360070
Delta: x=0.000000000, y=0.000000000
x=895.477518835
y=0.000444870
T (a)
Network Impact of 1588v2 clock transfer
t1
Sync PDV (1-step or 2-step)
Master
Clock
Slave
Clock
t-ms
Follow_Up
Sync
t-sm
Delay_Req
Delay_Resp
Propagation Delay
Message Exchange
t4
t2
t3
Does Master > Slave PDV impact clock recovery?
895.477518885894.977768110894.478017335Offset=0.002 ppm
-0.000296349
0.000645879
Time IntervalError (TIE)vs Nominal(seconds)
MKR-1:x=894.874691063, y=0.000360070MKR-2:x=894.874691063, y=0.000360070
Delta: x=0.000000000, y=0.000000000
x=895.477518835
y=0.000444870
T (a)
Clock
Output
COMPANY CONFIDENTIAL 9
895.477518885894.977768110894.478017335Offset=0.002 ppm
-0.000296349
0.000645879
Time IntervalError (TIE)vs Nominal(seconds)
MKR-1:x=894.874691063, y=0.000360070MKR-2:x=894.874691063, y=0.000360070
Delta: x=0.000000000, y=0.000000000
x=895.477518835
y=0.000444870
T (a)
Network Impact of 1588v2 clock transfer
t-ms
Follow_Up
Sync
t-sm
Delay_Req
Delay_Resp
Propagation Delay
Message Exchange
t1
t4
t2
t3
Master
Clock
Slave
Clock
Delay_Req PDV
Does Slave > Master PDV impact clock recovery?
895.477518885894.977768110894.478017335Offset=0.002 ppm
-0.000296349
0.000645879
Time IntervalError (TIE)vs Nominal(seconds)
MKR-1:x=894.874691063, y=0.000360070MKR-2:x=894.874691063, y=0.000360070
Delta: x=0.000000000, y=0.000000000
x=895.477518835
y=0.000444870
T (a)
COMPANY CONFIDENTIAL 10
895.477518885894.977768110894.478017335Offset=0.002 ppm
-0.000296349
0.000645879
Time IntervalError (TIE)vs Nominal(seconds)
MKR-1:x=894.874691063, y=0.000360070MKR-2:x=894.874691063, y=0.000360070
Delta: x=0.000000000, y=0.000000000
x=895.477518835
y=0.000444870
T (a)
Network Impact of 1588v2 clock transfer
t-ms
Follow_Up
Sync
t-sm
Delay_Req
Delay_Resp
Propagation Delay
Message Exchange
t1
t4
t2
t3
Round Trip Delay
Master
Clock
Slave
Clock
Will the in RTD variation impact the Slave clock?
895.477518885894.977768110894.478017335Offset=0.002 ppm
-0.000296349
0.000645879
Time IntervalError (TIE)vs Nominal(seconds)
MKR-1:x=894.874691063, y=0.000360070MKR-2:x=894.874691063, y=0.000360070
Delta: x=0.000000000, y=0.000000000
x=895.477518835
y=0.000444870
T (a)
COMPANY CONFIDENTIAL 11
895.477518885894.977768110894.478017335Offset=0.002 ppm
-0.000296349
0.000645879
Time IntervalError (TIE)vs Nominal(seconds)
MKR-1:x=894.874691063, y=0.000360070MKR-2:x=894.874691063, y=0.000360070
Delta: x=0.000000000, y=0.000000000
x=895.477518835
y=0.000444870
T (a)
Network Impact of 1588v2 clock transfer
t-ms
Follow_Up
Sync
t-sm
Delay_Req
Delay_Resp
Propagation Delay
Message Exchange
t1
t4
t2
t3
Asymmetry in Path Delay
Master
Clock
Slave
Clock
Will the Delay Symmetry impact the Slave clock?
895.477518885894.977768110894.478017335Offset=0.002 ppm
-0.000296349
0.000645879
Time IntervalError (TIE)vs Nominal(seconds)
MKR-1:x=894.874691063, y=0.000360070MKR-2:x=894.874691063, y=0.000360070
Delta: x=0.000000000, y=0.000000000
x=895.477518835
y=0.000444870
T (a)
COMPANY CONFIDENTIAL 12
895.477518885894.977768110894.478017335Offset=0.002 ppm
-0.000296349
0.000645879
Time IntervalError (TIE)vs Nominal(seconds)
MKR-1:x=894.874691063, y=0.000360070MKR-2:x=894.874691063, y=0.000360070
Delta: x=0.000000000, y=0.000000000
x=895.477518835
y=0.000444870
T (a)
• Monitor variation in arrival time of Sync
messages at Slave Clock
Network Impact of 1588v2 clock transfer
t-ms
Follow_Up
Sync
t-sm
Delay_Req
Delay_Resp
Propagation Delay
Message Exchange
t1
t4
t2
t3
Master
Clock
Slave
Clock
Sync IPG
Does Slave need regular Sync messages?
895.477518885894.977768110894.478017335Offset=0.002 ppm
-0.000296349
0.000645879
Time IntervalError (TIE)vs Nominal(seconds)
MKR-1:x=894.874691063, y=0.000360070MKR-2:x=894.874691063, y=0.000360070
Delta: x=0.000000000, y=0.000000000
x=895.477518835
y=0.000444870
T (a)
COMPANY CONFIDENTIAL 13
895.477518885894.977768110894.478017335Offset=0.002 ppm
-0.000296349
0.000645879
Time IntervalError (TIE)vs Nominal(seconds)
MKR-1:x=894.874691063, y=0.000360070MKR-2:x=894.874691063, y=0.000360070
Delta: x=0.000000000, y=0.000000000
x=895.477518835
y=0.000444870
T (a)
Network Impact of 1588v2 clock transfer
t-ms
Follow_Up
Sync
t-sm
Delay_Req
Delay_Resp
Propagation Delay
Message Exchange
t1
t4
t2
t3
Master
Clock
Slave
Clock
Follow_Up PDV
Will the variation impact the Slave clock?
• Arrival time of the Follow_Up message
with respect to the Sync message
895.477518885894.977768110894.478017335Offset=0.002 ppm
-0.000296349
0.000645879
Time IntervalError (TIE)vs Nominal(seconds)
MKR-1:x=894.874691063, y=0.000360070MKR-2:x=894.874691063, y=0.000360070
Delta: x=0.000000000, y=0.000000000
x=895.477518835
y=0.000444870
T (a)
COMPANY CONFIDENTIAL 14
895.477518885894.977768110894.478017335Offset=0.002 ppm
-0.000296349
0.000645879
Time IntervalError (TIE)vs Nominal(seconds)
MKR-1:x=894.874691063, y=0.000360070MKR-2:x=894.874691063, y=0.000360070
Delta: x=0.000000000, y=0.000000000
x=895.477518835
y=0.000444870
T (a)
Network Impact of 1588v2 clock transfer
t-ms
Follow_Up
Sync
t-sm
Delay_Req
Delay_Resp
Propagation Delay
Message Exchange
t1
t4
t2
t3
Master
Clock
Slave
Clock
Delay_Resp Response Time
• Launch time of the Delay_Req to the
arrival of the Delay_Resp message
Will the variation impact the Slave clock?
895.477518885894.977768110894.478017335Offset=0.002 ppm
-0.000296349
0.000645879
Time IntervalError (TIE)vs Nominal(seconds)
MKR-1:x=894.874691063, y=0.000360070MKR-2:x=894.874691063, y=0.000360070
Delta: x=0.000000000, y=0.000000000
x=895.477518835
y=0.000444870
T (a)
COMPANY CONFIDENTIAL 15
Network Impact of 1588v2 clock transfer
t-ms
Follow_Up
Sync
t-sm
Delay_Req
Delay_Resp
Propagation Delay
Message Exchange
t1
t4
t2
t3
Master
Clock
Slave
Clock
To fully understand the impact of the
Network on the Slave clock recovery,
the impact of variation and congestion
on the following parameters needs to
be understood:
• Sync PDV
• Delay_Req PDV
• Round Trip Delay
• Path Asymmetry
• Sync IPG
• Follow-up PDV
• Delay_Resp Response Time
COMPANY CONFIDENTIAL 16
Evaluation Plan
COMPANY CONFIDENTIAL 17
Evaluation Plan – Approach to date
1. Evaluate Session Management protocol.
2. Evaluate Time Transfer protocol.
3. Stress Test impact of Network Congestion and PDV.
• Introduce PDV into Sync & Del_Req.
•G.8261 Test Cases.
•Replay Live captures from active
network.
• Vary the Round-trip-delay & path
asymmetry.
•Vary the latency in each path.
• Introduce asymmetry between up-
stream & down-stream path latency.
• Emulate IPG variation on Sync, Follow-
up & Delay-response.
• Emulate varying Master Response times
• Evaluate Time Transfer message
sequence.
•Sync, Follow-up, Del_req, Del_resp
• Introduce Network Impairments to each
type of message;
•Drop, Repeat, Delay, Mis-ordered.
• Evaluate Session Start-up and Status
Management.
•Signalling and Announce messages.
• Introduce Network Impairments to each
type of message;
•Drop, Repeat, Delay, Mis-ordered.
COMPANY CONFIDENTIAL 18
Evaluation Plan – Enhanced
Characterise & compare
Slave performance with
controlled behaviours.
• Emulate recognised
network PDV behaviours.
• Step Noisefloor.
• Ramp, 24hr varying
Noisefloor.
• PDV amplitude.
• Deplete Noisefloor.
COMPANY CONFIDENTIAL 19
Estimate Performance Limits of Slave
Follow_Up
Sync
Delay_Req
Delay_Resp
t1
t4
t2
t3
Master
Clock
Slave
Clock
Pass
Pass
Fail
Clock
Output
COMPANY CONFIDENTIAL 20
Evaluation for mass deployment- The Goal
• Test network prior to deployment against
Performance Limits.
• Manage network to Performance Limits.
Base
StationEthernet
TDM
Base
Station
Future:
Ethernet-only
Today:
Hybrid Network
Core
Network
Engineer Network to agreed
Performance Limits.
What’s required to achieve this Goal;
• Agreement on recognised PDV behaviours.
• Establish PDV Limits.
• Tools to emulate and measure network to
not exceed Performance Limits.
COMPANY CONFIDENTIAL 21
www.calnexsol.com
Calnex Paragon
• IEEE 1588v2
• CES
• Sync-E
• NTP
• Ethernet OAM
21
Tommy Cook
Tel.: +44 (0) 1506-671-416
