Transcript 2-1 ATM MPLS

Virtual-Circuit Switching:
ATM (Asynchronous Transmission Mode)
and
MPLS
(Multiprotocol Label Switching)
2007. 10
Virtual Circuit (VC) Switching
 Hybrid of packets and circuits
Circuits: establish and teardown along end-toend path
 Packets: divide the data into packets with
identifiers
 Packets carry a virtual-circuit identifier
 Associates each packet with the virtual circuit
 Determines the next link along the path
 Intermediate nodes maintain state VC
 Forwarding table entry
 Allocated resources

Timing of Virtual-Circuit Packet
Switching
Host 1
Node 1
Host 2
Node 2
propagation delay
between Host 1
and Node 1
VC
establishment
Packet 1
Packet 2
Packet 1
Data
transfer
Packet 3
Packet 2
Packet 3
Packet 1
Packet 2
Packet 3
VC
termination
Establishing the Circuit
 Signaling
Creating the entries in the forwarding tables
 Reserving resources for the virtual circuit, if
needed
 Two main approaches to signaling
 Network administrator configures each node
 Source sends set-up message along the path
 Set-up latency
 Time for the set-up message to traverse the
path
 … and return back to the source
 Routing
 End-to-end path is selected during circuit set-up

Virtual Circuit Identifier (VC ID)
 Virtual Circuit Identifier (VC ID)
Source set-up: establish path for the VC
 Switch: mapping VC ID to an outgoing link
 Packet: fixed length label in the header

1
2
1: 7
2: 7
link 7
1: 14
2: 8
link 14
link 8
Swapping the Label at Each Hop
 Problem: using VC ID along the whole path
Each virtual circuit consumes a unique ID
 Starts to use up all of the ID space in the
network
 Label swapping
 Map the VC ID to a new value at each hop
 Table has old ID, and next link and new ID

1
2
1: 7, 20
20: 14, 78
link 7
2: 7, 53
53: 8, 42
link 14
link 8
Virtual Circuits Similar to IP
Datagrams
 Data divided in to packets
Sender divides the data into packets
 Packet has address (e.g., IP address or VC ID)
 Store-and-forward transmission
 Multiple packets may arrive at once
 Need buffer space for temporary storage
 Multiplexing on a link
 No reservations: statistical multiplexing
• Packets are interleaved without a fixed
pattern
 Reservations: resources for group of packets
• Guarantees to get a certain number of “slots”

Virtual Circuits Differ from IP
Datagrams
 Forwarding look-up
 Virtual
circuits: fixed-length connection id
 IP datagrams: destination IP address
 Initiating data transmission
 Virtual circuits: must signal along the path
 IP datagrams: just start sending packets
 Router state
 Virtual circuits: routers know about connections
 IP datagrams: no state, easier failure recovery
 Quality of service
 Virtual circuits: resources and scheduling per VC
 IP datagrams: difficult to provide QoS
Asynchronous Transfer Mode: ATM
 1990’s/00 standard for high-speed (155Mbps to 622
Mbps and higher) Broadband Integrated Service
Digital Network architecture
 Goal: integrated, end-end transport of carry voice,
video, data
 meeting timing/QoS requirements of voice, video
(versus Internet best-effort model)
 “next generation” telephony: technical roots in
telephone world
 packet-switching (fixed length packets, called
“cells”) using virtual circuits
ATM reference model
ATM architecture
 adaptation layer: only at edge of ATM network
data segmentation/reassembly
 roughly analagous to Internet transport layer
 ATM layer: “network” layer
 cell switching, routing
 physical layer

ATM Physical Layer
Physical Medium Dependent (PMD) sublayer
 SONET/SDH: transmission frame structure (like
a container carrying bits);
 bit synchronization;
 bandwidth partitions (TDM);
 several speeds: OC3 = 155.52 Mbps; OC12 =
622.08 Mbps; OC48 = 2.45 Gbps, OC192 = 9.6 Gbps
 TI/T3:
transmission frame structure (old
telephone hierarchy): 1.5 Mbps/ 45 Mbps
 unstructured: just cells (busy/idle)
ATM Physical Layer (more)
Two pieces (sublayers) of physical layer:
 Transmission Convergence Sublayer (TCS): adapts
ATM layer above to PMD sublayer below
 Physical Medium Dependent (PMD) : depends on
physical medium being used
TCS Functions:
 Header checksum generation: 8 bits CRC
 Cell delineation
 With “unstructured” PMD sublayer, transmission
of idle cells when no data cells to send
ATM Layer: Virtual Circuits
 analogous to IP network layer
 very different services than IP network layer
 VC transport: cells carried on VC from source to
dest
 call setup, teardown for each call before data can
flow
 each packet carries VC identifier (not destination
ID)
 every switch on source-dest path maintain “state”
for each passing connection
 link,switch resources (bandwidth, buffers) may be
allocated to VC: to get circuit-like perf.
ATM VCs
 Advantages of ATM VC approach:
 QoS
performance guarantee for
connection mapped to VC (bandwidth,
delay, delay jitter)
 Drawbacks of ATM VC approach:
 Inefficient support of datagram
traffic
 VC introduces call setup latency,
processing overhead for short lived
connections
ATM Layer: ATM cell
 5-byte ATM cell header
 48-byte payload
Why?: small payload -> short cell-creation
delay for digitized voice
 halfway between 32 and 64 (compromise!)

Cell header
Cell format
ATM cell header
 VCI: virtual channel ID
will change from link to link thru net
 PT: Payload type (e.g. RM cell versus data cell)
 CLP: Cell Loss Priority bit
 CLP = 1 implies low priority cell, can be
discarded if congestion
 HEC: Header Error Checksum
 cyclic redundancy check

ATM Service
 very different services than IP network layer
Network
Architecture
Internet
Service
Model
Guarantees ?
Congestion
Bandwidth Loss Order Timing feedback
best effort none
ATM
CBR
ATM
VBR
ATM
ABR
ATM
UBR
no
constant yes
rate
guaranteed yes
rate
guaranteed no
minimum
none
no
no
no
yes
yes
yes
yes
yes
no
no (inferred
via loss)
no
congestion
no
congestion
yes
yes
no
no
ATM Adaptation Layer (AAL)
 ATM Adaptation Layer (AAL): “adapts” upper
layers (IP or native ATM applications) to ATM
layer below
 AAL present only in end systems, not in switches
 AAL layer segment (header/trailer fields, data)
fragmented across multiple ATM cells
 analogy: TCP segment in many IP packets
ATM Adaptation Layer (AAL) [more]
Different versions of AAL layers, depending on ATM
service class:
 AAL1: for CBR (Constant Bit Rate) services, e.g. circuit
emulation (phone)
 AAL2: for VBR (Variable Bit Rate) services, e.g., MPEG video
 AAL5: for data (eg, IP datagrams)
User data
AAL PDU
ATM cell
IP-Over-ATM
app
transport
IP
Eth
phy
IP
AAL
Eth
ATM
phy phy
ATM
phy
ATM
phy
app
transport
IP
AAL
ATM
phy
How far along are we?
 Standardization bodies - ATM Forum, ITU-T
 We may never see end-to-end ATM (1997)
 Backbone:
- 1995 vBNS (ATM)
- 1998 Abilene (SONET) - 2000 IP over
DWDM
 ATM - too complex - too expansive
<IP>
 Internet technology + ATM philosophy
 but ATM ideas continue to powerfully
influence design of next-generation Internet
 ex: MPLS, admission ctl., resource
reservation, …...
Multiprotocol label switching (MPLS)
 initial goal: speed up IP forwarding by using
fixed length label (instead of IP address) to
do forwarding
 borrowing ideas from Virtual Circuit (VC)
 but IP datagram still keeps IP address!
PPP or Ethernet
MPLS header
header
label
20
IP header
Exp S TTL
3
1
5
remainder of link-layer frame
Label Substitution
 Have a friend go to B ahead of you using one of
the previous two techniques. At every road they
reserve a lane just for you. At every intersection
they post a big sign that says for a given lane
which way to turn and what new lane to take.
Label Encapsulation
 MPLS Encapsulation is specified over
various media types. Top labels may use
existing format, lower label(s) use a new
“shim” label format.
MPLS Link Layers
 MPLS -- run over multiple link layers
 Following link layers currently exist:
• ATM: label -- in VCI/VPI field of ATM header
• Frame Relay: label -- in DLCI field in FR header
• PPP/LAN: uses ‘shim’ header inserted
between L2 and L3 headers
 Translation between link layers types must be
supported
 MPLS is between L2 and L3
It intended to be “multi-protocol” below and
above
MPLS capable routers
 a.k.a. label-switched router
 forwards packets to outgoing interface based only
on label value (don’t inspect IP address)
 MPLS forwarding table distinct from IP
forwarding tables
 signaling protocol needed to set up forwarding
 Hop-by-hop or source routing to establish
labels
 forwarding possible along paths that IP alone
would not allow (e.g., source-specific routing) !!
 use MPLS for traffic engineering
 RSVP-TE
 must co-exist with IP-only routers
MPLS forwarding tables
in
label
out
label
out
interface
dest
10
A
0
12
8
D
A
0
1
in
label
out
label
10
6
A
12
9
D
dest
out
interface
1
0
R6
0
0
D
1
1
R3
R4
R5
0
0
R2
in
label
out
label
dest
8
6
A
out
interface
0
in
label
6
out R1
label
dest
-
A
A
out
interface
0
Best of Both Worlds
 MPLS + IP form a middle ground that
combines the best of IP and the best of
virtual circuit switching technologies
 ATM and Frame Relay cannot easily come
to the middle so IP has!
Multi-Protocol Label Switching
 Key ideas of MPLS
 Label-switched
path spans group of
routers
 Explicit path set-up, including backup
paths
 Flexible mapping of data traffic to paths
 Motivating applications
 Small routing tables and fast look-ups
 Virtual Private Networks
 Traffic engineering
 Path protection and fast reroute
Status of MPLS
 Deployed in practice
Small control and data plane overhead in core
 Virtual Private Networks
 Traffic engineering and fast reroute
 Challenges
 Protocol complexity
 Configuration complexity
 Difficulty of collecting measurement data
 Continuing evolution
 Standards
 Operational practices and tools

Optical Networks
1 st Generation: optical fibers substitute
copper as physical layer
 Submarine Systems
 SONET (synchronous optical) in TDM
 FDDI for LAN, Gbit Ethernet etc.
 2 nd Generation: optical switching and
multiplexing/ WDM
 broadcast-and-select networks
 WDM rings
 wavelength routing networks

3 th Generation: optical packet
switching???

Optical Switch
 1-input 2-outoput illustration with four wavelengths
Input & Output fiber
array
Wavelength
Dispersive Element
1-D MEMS
Micro-mirror
Array
Input Fiber
Output Fiber 1
1011
Digital Mirror
Control
Electronics
Output Fiber 2
 1-D MEMS (micro-electromechanical system) with
dispersive optics
 Dispersive element separates the ’s from inputs
 MEMS independently switches each 
 Dispersive element recombines the switched ’s
into outputs
All-Optical Switching
 Optical Cross-Connects (OXC)
 Wavelength
Routing Switches (WRS)
 route a channel from any I/P port to any O/P port
 Natively switch s while they are still multiplexed
 Eliminate redundant optical-electronic-optical
conversions
DWDM
Demux
DWDM
Fibers
in
DWDM
Mux
DWDM
Fibers
out
All-optical
DWDM
Demux
OXC
DWDM
Mux
MPS

MPS = Multi-Protocol Lambda Switching
MPLS + OXC
 Combining MPLS traffic eng control with OXC
 All packets with one label are sent on one wavelength
 Next Hop Forwarding Label Entry (NHFLE)
 <Input port,  > to <output port,  > mapping
