Planning your way to ASON viability

rufftartAI and Robotics

Oct 29, 2013 (4 years and 9 days ago)

158 views

Planning your way to ASON viability


TextStart

By Fu Bo


Transport networks are becoming more intelligent, but only rational network planning
can effectively leverage the intelligence
advantages of automatically
-
switched optical
network (ASON) technology.


Traditional optical transport technologies, designed for TDM service transmission,
cannot meet the needs of the data surge, which is where generalized multi
-
protocol
label switching (
GMPLS) and ASON come in. ASON technology delivers the
reliability, flexibility, and robustness that next
-
gen optical transport requires, while
enabling service differentiation and reducing network construction and O&M costs.


As the starting point of any
network, planning has the largest single impact on
construction costs, reliability, and ease
-
of
-
maintenance; ASON is no exception.


ASON: Advantages & challenges


SLA diversity


ASON infrastructure enables service level agreement (SLA)
differentiation, w
here different levels of services are granted different protections and
resource allocations. However, their proper selection has proven challenging as the
protection/investment ratio must be carefully considered.


Reliability


ASON infrastructure feature
s automatic rerouting, where routes are
selected based on user
-
predefined policies after a network fault, which ensures service
accessibility as long as an alternative route is available. Fault tolerance is therefore
greatly increased, but this requires am
ple resources that can enable reliably objective
and effective verification.


Reduced investment


ASON architecture is primarily meshed, which endows node
connectivity superior to that for ring/chain. ASON allows significantly more
connectivity through m
aximized bandwidth sharing, making route solution critical
during planning since small differences in this phase make huge differences in cost.


Four questions concerning ASON planning


Reliability, maintainability, scalability, and cost must be considered during network
planning, as it is a cyclical process of planning, verification, and re
-
planning. Huawei
has carried out roughly 100 WDM/SDH ASON deployments worldwide,
accumulating a ri
ch body of experience along the way; it is therefore able to offer a
complete suite of ASON planning methods and relevant software, designed to help
solve the various problems encountered during each phase.


Phase 1: Network structure & SLA selection

A net
work should be meshed whenever possible so that ASON architecture is best
leveraged. The number of optical directions at each node is determined based on
network survival requirements. Each node should include at least three optical
directions so that oper
ation continues after optical
-
direction faults occur at any two
points. Direct optical channels should also be designed between large
-
volume nodes
whenever possible, so that transmission efficiency is guaranteed. SLA selection is
primarily based on service

reliability requirements.


ASON infrastructure enables a variety of SLA choices, including permanent 1+1, 1+1
SNCP, dynamic restoration, and none; their availabilities and restoration times are
shown in Table 1.


Generally speaking, permanent 1+1 protect
ion is recommended for key private line
accounts, while 1+1 SNCP is suggested for general voice services, and dynamic
restoration is typical for data service.


Phase 2: Resource conservation

The planning phase focuses on service route selection and resourc
e allocation.
Traditional ring/chain topology offers few service routes, while ASON enables an
exponential increase, thanks to its meshed architecture. However, this plethora of
choices makes an optimal route difficult to determine.


Huawei offers an adva
nced route and resource allocation algorithm which calculates
service routes and required resources based on user
-
defined conditions for network
survivability, resource utilization, and route constraints. Comparison tests have shown
that Huawei
'
s algorithm

delivers 30% reduced costs over shortest
-
path (the most
prevalent route algorithm). In addition, the transmission for analog
-
based WDM
ASON systems is affected by optical performance. Huawei
'
s planning algorithm takes
optical performance restrictions into

consideration during route planning, as it will
select paths with good optical performance to reduce the number of configured
repeaters employed.


Phase 3: Cost reduction

The design phase focuses on equipment configuration, based on the resource
requireme
nts determined during the planning phase. ASON and traditional networks
are configured quite differently, with REG configuration having the greatest effect on
network cost. Regeneration boards in the electrical layer are expensive; they should
be used as l
ittle as possible. The number of REGs can be reduced through
service
-
restoration path sharing and regeneration board configuration by as many
service restoration paths as possible. Network testing shows that Huawei
'
s algorithm
can reduce the number of REGs

by up to 20% compared with traditional
configuration methods.


Phase 4: Network management

Like IP, ASON infrastructure is dynamic, as the network self
-
allocates resources and
service routes. Operators must know whether or not a fault will lead to a serv
ice
interruption, but testing an operating network is impossible. The complexities of
ASON architecture also make manual calculation unfeasible; network emulation is the
only answer. Huawei has developed an emulation software suite which includes
algorithm
s that port the ASON control plan. The software can gather the network data
from the network management system and simulate various fault scenarios so that
engineers can better understand service interruptions and post
-
fault resource
distribution. The emul
ation software allows prediction of any potential fault, while
enabling full control over the ASON.


Network reliability may worsen over time as new services are continuously launched;
optimization, which can be considered network re
-
planning, is needed. F
irst, the
existing network must be assessed. Huawei offers a comprehensive assessment
system, based on its aforementioned experiences, that determines network bottlenecks
through systematic assessment of network resources/reliability and optical
performanc
e, and provides recommendations for improving network
quality/performance.


ASON: An intelligent future


Huawei

has built more than 70 WDM ASONs as of summer 2011, covering 40
carriers worldwide; this includes the international backbone networks for Telecom
Italia, a national backbone network for Vodafone (Portugal), the backbone network
for Telecom Egypt, the nati
onal backbone network for Bharti (India), and the
backbone WDM ASON for China Telecom (Shaanxi).


These carriers generally face problems with long service provisioning periods, a
limited number of service protection modes, and frequent service interruption
s caused
by aging fiber. Huawei helps through joint planning of the service matrix, network
topology, and SLA, converging small
-
granularity bandwidths whenever possible and
sharing protection resources among as many services as possible. This reduces
netwo
rk building costs, while service reliability reaches 99.999% through protection
paths and preset reply paths which ensure service continuation after any two fiber
connections break down. Huawei
'
s network emulation algorithm simulates both fiber
and node fa
ults, thereby encompassing all possible network abnormalities, which
makes ASON infrastructure predictable and reduces the pressure on operation &
maintenance (O&M) personnel.


All transport networks are smartening up, but only reasonable network planning
can
leverage the intelligence of ASON architecture. Planning algorithms and emulation
technologies are critical for ASON planning, and Huawei is able to provide network
planning and optimizing services based on them, so that the ASON for any operator is
ec
onomical and competitive.



Table 1 Reliability and restoration time for certain SLAs



TextEnd