Next Generation Access Technologies Broadband Properties Summit April 28, 2009

High Speed Data: What you’re doing today, but more of it 100Mbps bi-directional already standard in Asia 50Mbps tiers already available in North America Rising popularity of user-generated video Increased resolution of digital photography (photo sharing) Remote services (backup, on-line applications, etc.) Video conferencing and video streaming Why do we need Next Generation Access Technologies?

High Speed Data: What you’re doing today, but more of it

100Mbps bi-directional already standard in Asia

50Mbps tiers already available in North America

Rising popularity of user-generated video

Increased resolution of digital photography (photo sharing)

Remote services (backup, on-line applications, etc.)

Video conferencing and video streaming

Video: IPTV Successfully deployed throughout Europe Gaining acceptance in North America Multiple HD TVs in the home More HD channels, more long tail content More unicast demand (VoD, Network PVR) HD Picture-in-Picture and trick modes create two to three times the bandwidth demand Interactive television applications (e.g. Blu-Ray 2.0) Why do we need Next Generation Access Technologies?

Video: IPTV

Successfully deployed throughout Europe

Gaining acceptance in North America

Multiple HD TVs in the home

More HD channels, more long tail content

More unicast demand (VoD, Network PVR)

HD Picture-in-Picture and trick modes create two to three times the bandwidth demand

Interactive television applications (e.g. Blu-Ray 2.0)

New Applications 3 dimensional television technologies have been demonstrated (double the HD bandwidth requirement - minimum) Multi-channel DVR chipsets have already been demonstrated: Every Channel DVR (EV/DVR) Up to 100+ HD channels simultaneously Applications not yet imagined Why do we need Next Generation Access Technologies?

New Applications

3 dimensional television technologies have been demonstrated (double the HD bandwidth requirement - minimum)

Multi-channel DVR chipsets have already been demonstrated: Every Channel DVR (EV/DVR)

Up to 100+ HD channels simultaneously

Applications not yet imagined

In the future, homes will require several Mbps to Gbps of bandwidth The only medium that can meet this demand is fiber optics Fiber access technologies are mature and are advancing in speed regularly The biggest question is not “Why?”, but “When?” It’s Going to be Fiber

In the future, homes will require several Mbps to Gbps of bandwidth

The only medium that can meet this demand is fiber optics

Fiber access technologies are mature and are advancing in speed regularly

The biggest question is not “Why?”, but “When?”

10G Passive Optical Networks The standards bodies have agreed that 10G PON is the next step in PON evolution Capitalizes on existing Ethernet and SONET technologies Allows bursts in excess of 2.4Gbps More efficient use of multicast/broadcast bandwidth Smaller/lower power potential than multi-wavelength solutions Both major standards bodies are working to ensure compatibility with existing PON New wavelength plans do not overlap with existing plans Current ONTs can already implement wavelength blocking based on early IEEE and FSAN work

The standards bodies have agreed that 10G PON is the next step in PON evolution

Capitalizes on existing Ethernet and SONET technologies

Allows bursts in excess of 2.4Gbps

More efficient use of multicast/broadcast bandwidth

Smaller/lower power potential than multi-wavelength solutions

Both major standards bodies are working to ensure compatibility with existing PON

New wavelength plans do not overlap with existing plans

Current ONTs can already implement wavelength blocking based on early IEEE and FSAN work

PON: Standards Bodies From the folks that brought you EPON Data-centric Generally focus on physical layer and engineering design Produce workable standards quickly Largely an equipment / chip vendor community From the folks that brought you GPON Telecommunications focus Attempt to standardize all of the layers from physical through management Produce thorough standards slowly Driven by the carriers with vendor participation IEEE FSAN / ITU

From the folks that brought you EPON

Data-centric

Generally focus on physical layer and engineering design

Produce workable standards quickly

Largely an equipment / chip vendor community

From the folks that brought you GPON

Telecommunications focus

Attempt to standardize all of the layers from physical through management

Produce thorough standards slowly

Driven by the carriers with vendor participation

Status Physical layer complete (wavelengths, etc.) Remaining sections largely complete Chipsets available in early form now, final form in early 2010 Speed 10Gbps downstream 1Gbps or 10Gbps upstream IEEE NG PON (802.3av)

Status

Physical layer complete (wavelengths, etc.)

Remaining sections largely complete

Chipsets available in early form now, final form in early 2010

Speed

10Gbps downstream

1Gbps or 10Gbps upstream

Status: Progress being made on the physical and coding layers Much of the higher layers still being debates Standardization expected to complete in 2010 or early 2011 Speed: 10G downstream 1.25G, 2.5G, and 10G upstream options ITU NG PON (FSAN NGA1)

Status:

Progress being made on the physical and coding layers

Much of the higher layers still being debates

Standardization expected to complete in 2010 or early 2011

Speed:

10G downstream

1.25G, 2.5G, and 10G upstream options

WDM PON Popular in Korea, some initial forays into the U.S. Business case in the U.S. is as a logical point-to-point for enterprise customers WDM PON is not currently being considered by either the IEEE or FSAN/ITU NGA1

Popular in Korea, some initial forays into the U.S.

Business case in the U.S. is as a logical point-to-point for enterprise customers

WDM PON is not currently being considered by either the IEEE or FSAN/ITU NGA1

WDM PON Multiple downstream wavelengths: one to every ONT or to groups of ONTs One or more upstream wavelengths: shared or per ONT A logical point-to-point network OLT Optics Tx Tx Tx Tx Wavelength Splitter Field Optics ONT  D1  U1 ONT ONT  D3  U3 ONT  D32  U32 CO Rx Rx Rx Rx

Multiple downstream wavelengths: one to every ONT or to groups of ONTs

One or more upstream wavelengths: shared or per ONT

A logical point-to-point network

When? The standards bodies are focusing on the access line Other technological advances will be required upstream Interface card capacities increased by a factor of 4 Backplane capacity increased by an order of magnitude Switch chip capacity hugely increased Upstream interfaces migrate from GbE to multiple 10GbEs per CO / switch Faster processing, memory, QoS marking, etc. Upstream network impact: ROADM, wavelengths, interfaces, core routers, etc. Additional external pressures make the job more challenging Lower cost Less power consumption including coming ATIS energy efficiency standards Smaller footprint The point: Don’t expect a massive upgrade to NG PON as soon as the standards complete

The standards bodies are focusing on the access line

Other technological advances will be required upstream

Interface card capacities increased by a factor of 4

Backplane capacity increased by an order of magnitude

Switch chip capacity hugely increased

Upstream interfaces migrate from GbE to multiple 10GbEs per CO / switch

Faster processing, memory, QoS marking, etc.

Upstream network impact: ROADM, wavelengths, interfaces, core routers, etc.

Additional external pressures make the job more challenging

Lower cost

Less power consumption including coming ATIS energy efficiency standards

Smaller footprint

The point: Don’t expect a massive upgrade to NG PON as soon as the standards complete

While waiting for NG PON, what should a developer deploy today? Option 1: Current generation PON (GPON/EPON) Current generation PON has plenty of bandwidth for most applications today The fiber infrastructure deployed today will work with NG PON Option 2: RFoG Deploy fiber with an RF headend Allows a fiber infrastructure to be deployed without spending capital on headend upgrades Conclusion: Deploy fiber today and you will not regret it in the future Interim Steps

While waiting for NG PON, what should a developer deploy today?

Option 1: Current generation PON (GPON/EPON)

Current generation PON has plenty of bandwidth for most applications today

The fiber infrastructure deployed today will work with NG PON

Option 2: RFoG

Deploy fiber with an RF headend

Allows a fiber infrastructure to be deployed without spending capital on headend upgrades

Conclusion: Deploy fiber today and you will not regret it in the future

Thank You Scott T. Wilkinson VP, Product Management and Systems Engineering Hitachi Communication Technologies America, Inc. [email_address]

Thank You

Scott T. Wilkinson

VP, Product Management and Systems Engineering

Hitachi Communication Technologies America, Inc.

[email_address]