Transcript IPO-2 - Internet Engineering Task Force

Unique Features and Requirements for
The Optical Layer Control Plane
<draft-chiu-strand-unique-olcp-01.txt>
IETF49 IPO WG, August 2000
Angela L. Chiu
John Strand
AT&T Labs
Bob Tkach
Celion Networks
IETF48-IPO
1
Objectives of Our Draft
• Describe unique features and requirements of the
Optical Layer and services
» Constraints on routing
• Design of network elements
• Impairment constraints in all-optical networks
• Diverse routing
» Business and operational realities
• Point out impacts on Optical Layer control plane
(OLCP) design and routing
IETF48-IPO
2
Design of Network Elements
• Reconfigurable network elements in the Optical Layer include
OLXC’s, OADM’s, tunable lasers, etc.
D
 11
 1k
 N1
 Nk
A
o
o
A
l1
lN
D
W
D
M
E
all-optical subnetwork
Y
X
OADM
la
OADM
lb
A
A
A
D
W
D
M
l1
lN
A
 11
 1k
A
 N1
 Nk
A
 a1 ak
Figure 1: An Optical Transport System (OTS) with OADM's
•
Adaptation: functions at the edges that transform the incoming optical
channel into the physical l to be transported through the subnetwork
•
Connectivity: defines which pairs of edge Adaptation functions can be
interconnected through the subnetwork
IETF48-IPO
3
Impacts of Adaptation and Connectivity
D
 11
 1k
 N1
 Nk
A
o
o
A
l1
lN
D
W
D
M
E
Y
X
OADM
la
OADM
lb
A
A
A
D
W
D
M
l1
lN
A
 11
 1k
A
 N1
 Nk
A
 a1 ak
• Multiplexing: Either electrical or optical TDM may be used to
combine the input channels into a single l .
• Adaptation Grouping: Groups of k (e.g., 4) l’s are managed as a
group within the system and must be added/dropped as a group
• Laser Tunability: The lasers producing the LR l’s may be tunable
over a limited range, or be tunable over the entire range of l’s
supported by the DWDM. Tunability speeds may also vary.
=> constrained connectivity between adaptation functions
IETF48-IPO
4
Impairment Constraints in Transparent Subnetworks
Opaque vs. transparent:
• Opaque: each link is optically isolated by expensive transponders
doing O/E/O conversions from other links
• Transparent: all-optical, bit rate and format independent
»
Advantages: cost saving, upgrade flexibility, etc.
»
Disadvantage: accumulation of physical impairments

Not all paths through a transparent subnetwork is feasible in general
Objective: allow expansion of “domain of transparency” with proper
routing
=>Need to consider physical impairments including Polarization Mode
Dispersion (PMD), Amplifier Spontaneous Emission (ASE), and
others
IETF48-IPO
5
PMD
OADM
OA
OA
OADM
OA
OADM
Nl
Nl
Kl
Jl
Ll
M
PMD < a% bit duration (a=10) => 10 B
2
D
(
k
)
 PMD * L(k )  1
k 1
DPMD(k) & L(k): : PMD parameter & length of the kth span, k=1, .., M
L: total length of a transparent segment
B \ DPMD
0.5 ps/km
0.1 ps/km
10Gb/s
L<400km
L<10000km
40Gb/s
L<25km
L<625km
• 10Gb/s or lower: not an issue except for bad fibers
• 40Gb/s or higher: will be an issue
IETF48-IPO
6
ASE
Bit rate & transmitter-receiver technology (e.g., FEC)=>SNRmin
PL
PL
SNRo 

 SNRmin w/ unity gain
Noise 2Mnsp h (G  1) Bo


PL
=> M  

 2SNRmin nsp h (G  1) Bo 
Assuming signal power PL=4dBm, amplifier gain G=25dB,
excess noise factor nsp=2.5,, hB0=-58dBm w/ carrier
freq.  and optical BW B0
• With FEC, SNRmin=20dB => M  10
• Without FEC, SNRmin=25dB => M  3
IETF48-IPO
7
Other Polarization Dependent Impairments
• Other polarization-dependent effects besides PMD
influence system performance.
• For example, many components have polarizationdependent loss (PDL) which accumulates in a system
with many components on the transmission path.
• The state of polarization fluctuates with time, and it is
generally required to maintain the total PDL on the path
to be within some acceptable limit.
IETF48-IPO
8
Chromatic Dispersion
• For reasonably linear systems, there are reasons to
believe that this impairment can be adequately (but not
optimally) compensated for on a per-link basis.
IETF48-IPO
9
Nonlinear Impairments
• It seems unlikely that these can be dealt with explicitly in a routing
algorithm because they lead to constraints that can couple routes
together and lead to complex dependencies, e.g. on the order in
which specific fiber types are traversed.
• A full treatment of the nonlinear constraints would likely require very
detailed knowledge of the physical infrastructure, including
measured dispersion values for each span, fiber core area and
composition, as well as knowledge of subsystem details such as
dispersion compensation technology.
• This information would need to be combined with knowledge of the
current loading of optical signals on the links of interest to determine
the level of nonlinear impairment.
IETF48-IPO
10
Nonlinear Impairments (cont.)
• Alternatively, one could assume that nonlinear impairments are
bounded and increase the required OSNR level (SNRmin) by X dB,
where X for performance reasons would be limited to 1 or 2 dB,
consequently setting a limit on route length.
• For the approach described here to be useful, it is desirable for this
length limit to be longer than that imposed by the constraints which
can be treated explicitly.
• It is possible that there could be an advantage in designing systems
which are less aggressive with respect to nonlinearities, and
therefore somewhat sub-optimal, in exchange for improved
scalability, simplicity and flexibility in routing and control plane
design.
IETF48-IPO
11
Implications For Routing and Control
Plane Design
• Additional state information will be required by the
routing algorithm for each type of impairment that has
the potential of being limiting for some routes, e.g.,
DPMD(k), L(k), G(k), nsp(k) and X dB, or in some
aggregated fashion.
• The specific constraints required in a given situation will
depend on the design and engineering of the domain of
transparency; for example it will be important to know
whether chromatic dispersion has been dealt with on
per-link basis, and whether the domain is operating in a
linear or nonlinear regime.
IETF48-IPO
12
Implications (cont.)
• It is likely that the physical layer parameters do not
change value rapidly and could be stored in some
database; however these are physical layer parameters
that today are frequently not known at the granularity
required. If the ingress node of a lightpath does path
selection these parameters would need to be available
at this node.
• In situations where only PMD and/or ASE impairments
are potentially binding the optimal routing problem with
the two constraints, OSPF algorithm enhancements will
be needed. However, it is likely that relatively simple
heuristics could be used in practice.
IETF48-IPO
13
Diverse Routing
• Key requirement for
»
Optical layer protection/restoration
»
Higher layer restoration, e.g., IP reroute for link failure
• "Shared Risk Link Group" (SRLG): relationship between
two non-diverse links, incl. many Types of Compromise:
» Examples: shared fiber cable, shared conduit, shared ROW, shared
optical ring, shared office without power sharing, etc.
• Impacts on control plane: nodes need to propagate information of
» Number of channels available for each channel type (e.g., OC48,
OC192) on each channel group
» Channel group: a set of channels that are routed identically and
should be given unique identification.
» Each channel group can be mapped into a sequence of fiber cables
while each fiber cable can belong to multiple SRLG’s based on their
definitions.
IETF48-IPO
14
Other Unique Features
• Bi-directionality: channel contention issue
• Protection/restoration:
» A pre-established protection path does occupy ports and
wavelengths
=>not optimized for shared mesh protection across different
endpoints
IETF48-IPO
15
Business and Operational Realities
• Expensive & periodically scarce optical network
resources and equipments => critical to control network
access, measure and bill for usage
• Support legacy services => multiple client types besides
IP routers
• Heterogeneity:
»
May across multiple carriers incl. Local Exchange Carriers &
national networks, may across trust boundaries
»
Transparent vs. opaque
»
Ring vs. mesh protection/restoration
»
Introduction of new technology => vendor specific
design/constraints
»
Different max. supportable bit rates
IETF48-IPO
16
Business and Operational Realities (cont.)
• Requirement: design a control plane that is flexible and
extendable to support end-to-end fast provisioning,
reconfiguration, and protection/restoration across
heterogeneous subnetworks
• Task: design scalable information exchange between
subnetworks using extension of routing protocols in the
control plane
• Service requirements being defined in OIF by the Carrier
Subgroup (chaired by John Strand of AT&T) -> IETF
IETF48-IPO
17
Discussions on Control Plane
Architecture
• Today’s OTS is a simple “domain of transparency”
consisting of WDM Mux/Demuxers and Optical
Amplifiers. Because an OTS is not easily reconfigurable
today, these constraints are dealt with at the time of
installation and don’t complicate routing and the control
plane.
• As domains of transparency become both larger and
software reconfigurable, these physical constraints on
connectivity and transmission quality become
increasingly of concern to the control plane.
• The evolution is largely technology driven =>
heterogeneous technologies => different in their routing
implications.
IETF48-IPO
18
Control Plane Architecture Alternatives
Per-domain routing:
• Each domain could have its own tuned approach to
routing.
• Inter-domain routing would be handled by a multidomain or hierarchical protocol that allowed the hiding of
local complexity.
• Single vendor domains might have proprietary intradomain routing strategies.
IETF48-IPO
19
Control Plane Architecture (cont.)
Enforced Homogeneity:
• The capabilities of the control plane would impose
constraints on system design and network engineering.
• As examples: if control plane protocols did not deal with
nonlinear impairments carriers would require their
vendors to provide transport systems where these
constraints were never binding.
• Transmission engineers could be required to only deploy
domains where every possible route met all constraints
not handled explicitly by the control plane even if the
cost penalties were severe.
IETF48-IPO
20
Control Plane Architecture (cont.)
Additional Regeneration:
• At (selected) OLXC’s within a domain of transparency,
the control plane could insert O/E/O regeneration into
routes with transmission problems.
• This might make all routes feasible again, but at the cost
of additional cost and complexity and with some loss of
rate and format transparency.
IETF48-IPO
21
Control Plane Architecture (cont.)
Standardized Intra-Domain Routing Protocol:
• A single standardized protocol which tries to deal with
the full range of possible topological and transmission
constraints will be extremely complex and will require a
lot of state information.
• However when combined with limited application of the
two previous approaches it might be more plausible.
IETF48-IPO
22
Centralized vs Distributed Routing
• A centralized routing model, where routing is done
centrally using a centralized database with a global
network view would be suitable for a domain where
DWDM transmission systems have reconfigurable
OADMs in between terminating points and tunable
lasers on the drop ports.
• A distributed model appear to be an excellent candidate
for a purely "opaque" domain where impairment
constraints play no role in routing.
IETF48-IPO
23
Advantages of Centralized Routing
• Information such as SRLG’s and performance
parameters which change infrequently and are unlikely
to be amenable to self-discovery could reside in a
central database and would not need to be advertised.
• Routing dependencies among circuits (to ensure
diversity) is more easily handled centrally when the
circuits do not share terminals since the necessary state
information should be more easily accessible in a
centralized model.
• Pre-computation of restoration paths and other
computations that can benefit from the use of global
state information may also benefit from centralization.
IETF48-IPO
24
Disadvantages of Centralized Routing
• If rapid restoration is required, it is not possible to rely on
a centralized routing system to compute a recovery path
for each failed lightpath on demand after a failure has
been detected. The distributed model arguably will not
have this problem.
• The centralized approach is not consistent with the
distributed routing philosophy prevalent in the Internet.
The reasons which drove the Internet’s architecture –
scalability, the inherent problems with hard state
information, etc. – are largely relevant to optical
networking.
• A centralized approach would seem to preclude
integrated routing across the IP and optical boundary.
IETF48-IPO
25
Pre-Computing All Possible Routes
Advantage: allowing more sophisticated algorithms to be
used to filter out the routes violating transmission
constraints.
Disadvantages:
• In a large national network there are just too many
routes that might be needed, by orders of magnitude.
This is particularly true when diversity constraints and
restoration routing may force weird routings.
• Every time any parameter changes anywhere in the
network all routes using the impacted resource will need
to be reexamined.
IETF48-IPO
26
Next Steps
• Propose to make this a WG document
• Work with vendors and other carriers to propose
standardized solutions for those that have not been
properly addressed by existing protocols
IETF48-IPO
27