You are on page 1of 27

1

PROTECTION SWITCHING

1
2
PROTECTION SWITCHING
Feature

Most Important Feature of SONET Standard is its Resiliency Nature i.e.


Restoration of the Traffic in Case of Failure within a specified time duration.
Protection switch time
A protection switch refers to the action of re-routing the traffic to the protection
reserved bandwidth upon a failure on the working side. The SONET standard
specifies that a protection switch should take place within 60 msec. This time
interval includes:

Failure detection time.


Switching time.
Frame resynchronization.
Propagation delay.

Typically, in the 60 msec time budget, 10 msec is reserved for fault detection
and 50 msec is reserved for traffic switching.

The main motivation behind this specification (60 msec switch over) is related
to the requirement that the lower speed streams, such as DS1 and DS3, that are
multiplexed in the high-speed traffic stream, must not lose frame
synchronization at their receivers. 2
3
Traffic Protection

Automatic Protection Switching (APS) is the capability of a transmission


system to detect a failure on a working facility, and to switch a standby
(protection) facility to recover the traffic.

Bytes K1 and K2 in the Line Overhead of the first STS-1 signal are used to
carry the protocol that coordinates the protection switching. Therefore,
protection switching in SONET is done at the line level by the Line Terminating
Equipment (LTE).

Protection switching occurs as a result of:


Signal failure (SF) – LOS (Loss of Signal), LOF (Loss of Frame), AIS
(Alarm Indication Signal) (AUTOMATIC)
Signal Degradation (SD) (BER – Bit Error Rate – 3: Excessive Error)
(AUTOMATIC)
In response to certain user initiated commands. (MANUAL)

3
4
Types Of Protection Switching
LINEAR PROTECTION

1+1 Linear Protection


1:1 Linear Protection
1:N Linear Protection

RING PROTECTION
UNIDIRECTIONAL
UPSR – (Unidirectional Path Switched Ring)

BIDIRECTIONAL
2F-BLSR – (Bidirectional Line Switched Ring)
4F-BLSR – (Bidirectional Line Switched Ring)
4
5
1 + 1 Linear Protection

Working
One fiber is called the working fiber
and the other the protection fiber.
100 % Traffic is transmitted Bridge Selector

Protection
simultaneously on both the Fibers. Head-end Tail-end
Both fibers are usually diversely
Working
routed.

Bridge Selector

Protection
Head-end Tail-end
In Normal condition the destination selects one of the two fibers based on the
quality of the received signal, this being the Worker.
In Case of Failure in Worker, the Destination Switches to the Protection.
1+1 is a fast restoration mechanism: since no APS signaling is required to
achieve a protection switch (assuming 1+1 unidirectional).
Doesn’t support low priority traffic (extra traffic).

5
6
1:1 Linear Protection
Working

100% Traffic is transmitted only on


one fiber i.e. the Worker. Bridge Selector

Protection
No traffic or Extra Traffic (Low Head-end Tail-end
Priority Traffic) is Transmitted on
Working
Protection Fiber.
Usually both the Fibers are diversely
routed Bridge Selector

Protection
Head-end Tail-end

In Normal condition the TX and RX at Source and Destination are bridged on


working Fiber.
In Event of Failure in Worker, the TX and RX at Source and Destination switch to
Protection. The Extra Traffic Is lost during this Period.

6
7
1:N Linear Protection
Working Channel #1
Working Channel #2
Node Node
Working Channel #N

Protection Channel

Working Channel #1
Working Channel #2
Node Node
Working Channel #N

Protection Channel
Traffic is provisioned only on working lines and NO or Extra Traffic is provisioned
on Protection Line..
N lines share one Protection Line.
During Failure, the protection Line takes care of only the Defective Line.
Economical Scheme.
If more than one working line becomes defective then , protection line will take
care of the working line with more priority
7
8
APS bytes for Linear Protection Switching

Example: K-bytes definitions for linear protection as per Bellcore GR-253.(SONET)

K1 Byte: Indicate the request K2 Byte: Indicate the bridging action


1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8

Switch Reverse
Preemption Channel # 111 Line AIS
Channel #
Priority 110 Line RDI
101 Bidirectional APS
1111 Lockout of Protection 100 Unidirectional APS
1110 Forced Switch
1101 Signal Fail - High (for 1:N only) 0 1+1 Mode
1100 Signal Fail - Low 1 1:N Mode
1011 Signal Degrade - High (for 1:N only)
1010 Signal Degrade - Low
1001
1000 Manual Switch
0111
0110 Wait-to-Restore
0101
0100 Exercisor
0011
0010 Reverse Request (for bidirectional only)
0001 Do Not Revert (for 1+1 non-revertive only)
0000 No Request
8
9
RING PROTECTIONS
UNIDIRECTIONAL
Information is only transmitted in one direction. A connection to the
neighboring NE may require to traverse an entire length of the ring to
complete the Connection.

Drawback :

Significant Propagation time variation can occur in both the Transmission


Direction.

BIDIRECTIONAL
Connections are Bidirectional i.e. both the Transmit and receive traverse the
same path and Distance in the Ring.

Shorter Path : Working Path Longer Path : Protection Path


We will look at the three most popular ring topologies;
Unidirectional Path Switched Ring (UPSR),
4 fibers Bidirectional Line Switched Ring (4F-BLSR) and
2 fibers Bidirectional Line Switched Ring (2F-BLSR). 9
10
RING PROTECTIONS

10
11
Unidirectional Path Switched Ring (UPSR)
In UPSR, one of the fibers is considered the working fiber and the other the protection
fiber.
Traffic is transmitted simultaneously on the working fiber in the clockwise direction and
on the protection fiber in the counter-clockwise direction. I.e. B and E transmit in both
Directions simultaneously as in 1+1 protection.
In event of failure between C and D the receiver is E is switched over to Fiber 2(Red -
Protection) where it finds a connection immediately.
The advantages of UPSR relies in its simplicity and low cost.No knowledge of ring
configuration is required.
No Extra Traffic can be Provisioned In UPSR RING

11
12
4-Fiber Bidirectional Line Switched Ring (4F-BLSR)
In 4F- BLSR, two fibers are used as working
fibers and two are used for protection.
Working Traffic is routed to the shortest path
between two nodes in the ring. This maximizes
Working
the amount of spatial reuse obtained.
Using 4F-BLSR 100% Protection can be
achieved.
Extra Traffic can be provisioned on the
Protection Fiber Pair Protection
A BLSR ring can support up to 16 nodes
(limited by the node identifier in the K-bytes).
Nortel OC-192 BLSR supports up to 24 nodes.
Maximum ring length is limited to 1200km (6
ms propagation delay).
Protection switch can be achieved in 60ms.
However, for longer rings (undersea
applications), the 60 ms restoration time has
been relaxed.
BLSR rings are widely deployed in long-haul
network, where the traffic pattern is more
distributed than in access network.
12
13
4F-BLSR: Span switch
A 4F-BLSR ring supports two types of protection mechanisms:
Span switching and Ring switching.

SPAN SWITCHING : Working

Span
In a span switch, if a fiber is cut on
Switch
the working side, the traffic is routed
onto the protection fiber between the
two same nodes, as shown in the Protection

picture.

No switching takes place on the


Other NEs in the ring.

A span switch is similar to a linear


1:1 protection switch. It is always
revertive and both directions are
always switched simultaneously
(similar to linear bidirectional switch).

13
14
4F-BLSR: Span switch

14
15
4F-BLSR: Ring switch

RING SWITCHING :
Working

If both the Worker and Protection


fiber get cut, service is restored by a
ring switch.
Protection
In a ring switch, the traffic on the
failed link is then rerouted around the
ring on the protection fibers.i.e. on the
Long Path

This requires coordination between


the two nodes that terminates the
failed fibers. APS protocol has to take
place around the long path of the ring.

During Ring Switch the Extra Traffic


Is lost.
15
16
4F-BLSR: Ring switch

16
17
2F-BLSR
In 2F-BLSR, both the fibers are used
to carry 50% working traffic, and rest
50% of the capacity on each fiber is
reserved for protection purposes.
Working

Therefore the total Capacity of the


Ring reduces to 50%.
Protection
Unlike a 4F-BLSR, span switch is
not possible in a 2F-BLSR. In the
event of a link failure, the traffic is
rerouted on the long path of the ring
using the protection capacity available
on the two fibers.

A 2F-BLSR ring supports only one


type of protection mechanism; Ring
switching.

17
18
2F-BLSR

18
19

APS K-Byte (K1, K2) – Ring protection


Example: K-bytes definitions for BLSR protection as per GR-1230, issue 4.

K1 Byte: Indicate the request K2 Byte: Indicate the bridging action


1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8

Switch Destination Source


Preemption Node ID Source
Node ID
Priority Node ID 111 Line AIS
110 Line RDI
1111 Lockout of Protection [LP-s] or 101 Reserved
Signal Fail – protection [SF-P] 100 Reserved
1110 Forced Switch – Span [FS-S] 011 Extra Traffic
1101 Forced Switch – Ring [FS-R] 010 Bridged & Switched
1100 Signal Fail – Span [SF-S] 001 Bridged
1011 Signal Fail – Ring [SF-R]
1010 Signal Degrade – Protection [SD-P]
000 Idle
1001 Signal Degrade – Span [SD-S] 0 Short Path Request
1000 Signal Degrade – Ring [SD-R] 1 Long Path Request
0111 Manual Switch – Span [MS-S]
0110 Manual Switch – Ring [MS-R]
0101 Wait-to-Restore [WTR]
0100 Exerciser – Span [EXER-S]
0011 Exerciser – Ring [EXER-R]
0010 Reverse Request – Span [RR-S]
0001 Reverse Request – Ring [RR-R]
0000 No Request
19
20
Revertive mode and WTR (Wait to Restore)
period
• Once protection switching has been employed, what happens when the original
working fiber cut has been repaired?
• In theory, when using 1+1 protection scheme, no action is needed since the
repaired fiber now becomes the protection fiber. This type of protection
switching is termed non-revertive.
• In 1:N protection scheme, however, it is desirable to have the traffic switched
back onto the original working fiber so that new failures can be handled and
low-priority traffic may continue to use the protection fiber. This type of
protection switching is termed revertive.
• 1:N always uses revertive mode with a user-provisionable WTR period.
• 1+1 can use revertive (with WTR) or non-revertive mode.

20
21
protection Switching Priority

21
22

• protection switch requests can occur automatically by the


system or by user initiated actions.
— Automatic switch
— User-initiated switches
– Manual switch
– Forced switch
– Lockout of Working
– Lockout of Protection

22
23

• Manual Switch :-
• Initiates a switch from either a working line – to the protection
line or vice – versa. This command has the lowest priority

Forced Switch:-
Forces a switch from either a working line to protection line or
vice –versa without regard to the state of either line.

23
24
Wait to Restore (WTR)

• Rapid protection switching and reversion can wreak havoc


on signal quality.
• System must remain stable (good signal on working channel)
for the duration of WTR time before protection channels will
revert back to working channels.
• WTR is 2nd lowest in Protection hierarchy (all user-initiated
and auto switches will override WTR).

24
25
Lockout commands

General
Lockout: Prevent a working channel to use the protection channel, even in
case of line failure.

The Lockout of Working commands (LOW-span or LOW-ring) are not


signalled on the APS channel [the k-byte protocol] but are implemented
locally using the Data Communication Channels (DCC), while the
Lockout of Protection (span) is signaled on the APS with the K-byte
protocol.

LOW-span and LOW-ring are applied against the working line, whereas
LP-span is applied against the Protection line.

Lockout of Protection - span (LP-S) has the same priority as Lockout of


Working - Span (LOW-S), Lockout of Working - Ring (LOW-R), and
Lockout of Protection - all spans (LOP-all spans). Each of these
commands can coexist and must be cleared separately.

25
26
Lockout of Working commands
Lockout of Working -Span (must be applied at
both ends of the fiber span) prevents the
addressed working line from using its span
protection partner for the purposes of span
switching. If a working failure were to occur, a
ring switch would be allowed. The command has no
impact on the use of protection for any other
span. It remains active until it is released.
W
In other words, LOW-span forces traffic to
travel on the working line along the span. P

Lockout of Working -Ring (must be applied at both


ends of the fiber span) prevents the addressed
working line from using its ring protection partner
for the purposes of ring switching, by disabling the
node’s capability to request a ring switch. If a
working failure were to occur, a span switch to the
protection span would be allowed. It remains active
until it is released.
W
When active, no ring switch can occur along the
span. P
When a Global lockout of working is applied to the ring (from PMEM), no ring switch
can occur anywhere around the ring. 26
27
Lockout of Protection

Lockout of Protection -Span is applied


against the protection line ; it prevents the
use of the span for any protection activity
and prevents ring switching anywhere in the
ring. If a working failure were to occur on the
addressed span, neither a span switch (for
this particular span) nor a ring switch would
be allowed. Unlike lockout of working, lockout
of protection span is signaled via the APS
channel. It remains active until it is released.
W
When active, traffic is prevented from
traveling on the protection line along the
span.

When a Global lockout of protection is applied to the ring (from PMEM), traffic cannot
travel on the protection lines around the ring, preventing any span or ring switches.

27