Architectures and Protocols for Data Communications Over ... - MIT

fullgorgedcutΔίκτυα και Επικοινωνίες

24 Οκτ 2013 (πριν από 3 χρόνια και 11 μήνες)

76 εμφανίσεις

LIDS

MIT

Outline


Motivation




Simulation Study




Scheduled OFS




Experimental Results




Discussion

LIDS

MIT

Optical Flow Switching Motivation


OFS reduces the amount of electronic processing by switching
long sessions at the WDM layer


Lower costs, reduced delays, increased switch capacity


Provide specific QoS for advanced services


LIDS

MIT

OFS Motivation (cont)

Flow Size

1KB

1MB

100MB

10MB

Flow Size

1MB

100MB

10MB

1KB

Optical Domain

Elect. Domain

Optical Domain

Elect. Domain

-
Internet displays a “heavy
-
tail” distribution of connections

-
More efficient optics => more transactions in optical domain (red line moves left)

LIDS

MIT

Optical Flow Switching Study


Short
-
duration optical connections


Access area


Wide area


Network architecture issues


Connection setup


Route/wavelength assignment


Goal: efficient use of network resources I.e. high throughput


Previous work: “probabilistic” approaches


Difficulty: high
-
arrival rate leads to high blocking probability


Problem: lack of timely network state information


Our proposed solution: Use of timing information in network


Schedule connections


Gather timely network state information


This demonstration


Demonstrate flow switching


Demonstrate viability of timing and scheduling connections


Investigate key sources of overhead


High efficiency


LIDS

MIT

Connection Setup Investigation


Key issue:


How to learn optical resource availability?


Distribution problem


“Wavelength continuity” problem makes it worse



Previous work


Addresses issues one at a time


Assumes perfect network state information


Will these results be useful for ONRAMP, WAN implementation?



This work


Assesses effects of distributed network state information


Models some current proposals


MP
-
lambda
-
S


ASON




LIDS

MIT

Methodology


Design distributed approaches


Combined routing, wavelength assignment


Connection setup



Baseline flow switching architecture


Requested flows from user to user


Durations on order of seconds


All
-
optical



Simulate approaches on WAN topology


End
-
to
-
end latency (“time of flight” only)


Approaches: Ideal, Tell
-
and
-
Go, Reverse Reservation



Assess performance versus idealized approach


Blocking probability

LIDS

MIT

Ideal Approach Illustration

A

C

B

D

l
-
Changers

l
-
Changers

l
-
Changers

l
-
Changers

A

C

B

D

Bidirectional

Multi
-
fiber Link

Network Infrastructure

“Tell”

cntl packet

LLR Routing, Connection Setup

Optical

Flow

Assume:

Flow Requested from A
-
>B

LIDS

MIT

Tell
-
and
-
Go Approach Illustration

A

C

B

D

Link
-
state

Updates

Available
l
: 1,2,3

Available
l
: 1,2

Available
l
: 2,3

Available
l
: 2,3,4

Link
-
State Protocol

A

C

B

D

Optical

Flow

Connection Setup

“Tell” Packet
-


Single wavelength

Assume:

Flow Requested from A
-
>B

LIDS

MIT

Reverse Reservation Approach Illustration

A

C

B

D

Information

Packets

A

C

B

D

Route Discovery

Route Chosen by B

Reservation

Packet

Assume
: Flow Requested from A
-
>B

Route, Wavelength Reservation

LIDS

MIT

Simulation Description


Results shown as Blocking Probability vs. Traffic Intensity


Uniform, Poisson flow traffic per node



Fixed WAN topology



Parameters:


F = Number of fibers/link


L = Number of channels/link


K = Number of routes considered for routing decisions


U = Update interval (seconds)




= Average service rate for flows (flows/second)


l

= Average arrival rate of flows (flows/second)




= Traffic intensity. Equal to
l
/




not utilization factor

LIDS

MIT

Simulation Topology

LIDS

MIT

Latency
-
free Control Network Results (1sec
flows)

RR
: F=1, L=16, K=10

TG
: F=1, L=16, K=10

LIDS

MIT

Control Network With Latency Results (1sec
flows)

TG, RR
: U=0.1, F=1, L=16, K=10

LIDS

MIT

Interesting Phenomenon



Why is TG performance better than RR?


1 sec flows and large rho => small inter
-
arrival times


Smaller than round trip time


Thus, with high probability, successive flows will see
same state (at
least locally)


Increases chance of collision


Effect of distribution (latency)



Why is Rand better than FF?


This is exactly opposite of analytical papers’ claim


Combination of reasons


Nodes have imperfect information


FF makes them compete for
same

wavelengths (false advertisement)


Not seen in analysis because distribution was ignored







LIDS

MIT

Scheduled OFS in ONRAMP


Inaccurate information hurts performance


In this case: Simple speed of light


Biggest problem: Core network resources wasted



Our proposal: Use of timing information to schedule flows


Deliver network information on time to make decisions


Exchange
flow
-
based
information


Maximize utilization of core network


Possible small delay for user



Issues


Can timing be implemented cheaply, scaled?


Can schedules be implemented?


Must make use of current/future optical devices


Low cost



ONRAMP OFS


Demonstration of scheduled OFS in access
-
area network


One example of an implementation



LIDS

MIT

Fixed

l

Xponder

Tunable

l

Xponder

Access Node #2

OXC

Router

GE

GE

IP

FLOW

IP

Control

Xmitter (X)

Fixed

l

Xponder

Tunable

l

Xponder

Access Node #1

Router

OXC

GE

GE

IP

FLOW

IP

Control

Intermediate Node

OXC

OXC

Router

Router



Receiver (R )

Fixed

l

Xponder

Tunable

l

Xponder

Access Node #2

OXC

GE

GE

IP

FLOW

IP

Control

)

Fixed

l

Xponder

Tunable

l

Xponder

Access Node #1

Router

OXC

GE

GE

IP

FLOW

IP

Control

Intermediate Node

OXC

OXC

Router

Router

X
-
a

R
-
a

OXC Sched

OXC Sched

OXC Sched

Scheduling in ONRAMP

LIDS

MIT


Uses timeslotting and schedules for lightpaths


X =>
l
i busy on ou瑰u琠o映node i a琠corresponding slot


OXC Schedule

Slot 1

…..

Slot 2

Slot 3

l 1

l 4

l 2

l 3

X

X

X

X

ONRAMP Connection Setup

LIDS

MIT

-
Overheads includes all timing uncertainty


-
Efficiency of any scheduled algorithm related to

timing uncertainty, and switching/electronic overheads


-
Rough efficiency =
Flow duration / Flow duration + Overhead

Slot 1

Overhead
-

Dependent on timing uncertainty

TIME

Scheduling OH

Cannot

go in next timeslot

Scheduling OH

Can

go in next timeslot

Slot 3

Slot 2

Algorithm Timeline

LIDS

MIT

Utilizing Link Capacity


Sending GigE over transparent optical channel


Clock rate 1.244 Ghz


Rate 8/10 coding results in raw bit rate of 995.2 Mb/s


Payload capacity for UDP


Send MTU
-
sized packets


9000 bytes


Avoid fragmentation


Headers


Ethernet (26 bytes) + IP (20 bytes) + UDP (8 bytes) = 54 bytes


Result: 8946 bytes of payload/packet


Link payload limit


989.2288 Mb/s


Rate
-
limited UDP


Input: desired rate


Timed sends of UDP packets achieve desired rates


Demonstrates transparency of OFS channel

LIDS

MIT

Experimental Setup


OFS implemented in lab



One second timeslots


Timing overhead negligible



Routing/wavelength selection


All available wavelengths (currently 14)


Both directions around ring



Gigabit Ethernet link layer


Flows achieve theoretical maximum link rate ~989 Mb/s




Rate limited UDP


Unidirectional flows


No packet loss (100s of flows)


Variable rate


Demonstrates transparent use of optical connection


LIDS

MIT

OFS Performance

LIDS

MIT

Current Performance Limitations

LIDS

MIT


Current overhead is 0.10 seconds


Efficiency for one
-
second flows is therefore 90%


Analysis of overhead reveals possible overhead of Gigabit Ethernet
frame sync


Still under investigation


Switching overhead and timing uncertainty negligible


I.e. scheduling viable, efficient


Current Performance Limitations(cont.)

time

Scheduling

Command

GBE Sync?

Receiver

Laser

Switching

Algorithm Overhead Timeline

Flow begins…………

10ms

150ms

100ms