White Paper:

TDM to IP Migration: Your Network, Your Timeline

Abstract:
Network technology based on the Internet Protocol has now
advanced to the point that data traffic of any kind can be as
reliably transported via Ethernet as it can via TDM. This paper
explores the strengths and weaknesses of both TDM and Ethernet
transport within the context of microwave radio communications.
The paper further describes the operational advantages of Ethernet
and Ethernet radio systems and outlines an approach to migration
from a TDM-based network to an all-IP network.

TDM to IP Migration: Your Network, Your Timeline

Introduction
Just as cellular networks evolved from analog to digital and
television made the same transition, the life of the most
widely used communications transport in history TDM
is drawing to a close, with newer Internet Protocol standing
ready to pick up where TDM left off. But whereas public
network operators and broadcasters numbered in the
hundreds, private networks based on TDM number in the
hundreds of thousands or more. Thus, the transition from
TDM to IP will not be a rapid one as was the case with
television and most cellular networks but, rather will take
place on a timetable determined by the operator of each
and every private TDM network.
Enterprise organizations that operate private data networks
based on TDM technology and use microwave radios to
implement a significant portion of their individual TDM
links face a unique challenge. Because their networks are
primarily owned, not leased, the transition from one
technology to the other will take longer and investment
in new equipment may not generate a return if that
investment fails to take into account the inevitable future
transition to IP. Operators of these systems should begin
migration to an Ethernet/IP-based microwave architecture.
The first step in this migration is upgrading existing
microwave links to new generation microwave radio
systems designed specifically to support TDM and
Ethernet simultaneously.
Initially, manufacturers developed microwave-based
TDM systems for telephony, but today many enterprises also
use these systems for data transmission applications. These
include remote data monitoring, remote control of industrial
apparatus and, increasingly, data-centric applications such
as video conferencing, Storage-Area Networks (SANs), and
private wide-area computer networks.

P1

Data-centric applications specific to industry are as varied

as the industries themselves.
Pipeline and utility companies operate extensive SCADA
(Supervisory Control and Data Acquisition) systems over
the full network of their physical systems, often requiring
data connectivity to remote locations. The trend is to
use these communications systems for real-time control
functions as well as monitoring; as a result, accurate and
timely data communications becomes critical.
Railroads pioneered electrical communications for train
control over 150 years ago, and modern rail systems
continue to use electronic and wireless systems for
communications, train control and asset monitoring.
Public Safety agencies require reliable communications
that remain in operation even in the event of a major
disaster. Consequently, they cannot rely on public or
cellular telephone networks. These agencies must have
access to an independent and fault-tolerant network
that supports their voice communication needs and,
increasingly, bandwidth-intensive applications such as
real-time video surveillance.
In the health-care industry, a wave of consolidation is
creating clusters of hospitals, clinics, urgent-care facilities
and labs operating within one business structure. These
organizations need reliable communications among
all facilities, and increasingly they require substantial
data capacity for medical imaging, records transfers
and related applications.
Universities operate satellite campuses for evening and
graduate students, and the shift to computer-centric
learning means that these satellites need reliable data
connectivity. A similar trend is emerging in many larger
school districts, where magnet schools and other
programs depend on advanced video and computer
networks to make material available to students.

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

Businesses of all types operate intra-company networks

among their facilities, across the nation and in many cases
around the world. Once, such systems were low data rate
networks aimed primarily at order entry, voice tie-line, and
similar applications, but today they often support video
conferencing, real-time monitoring of factory production,
global engineering and support operations. To better
manage costs and energy use, many large enterprises are
moving to fewer, larger data centers with high-capacity
interconnections to all corporate facilities.
All of these new applications share one thing in common
they require far greater capacity than older legacy data
and control systems. Most have a second characteristic
in common they are built on top of Internet packetbased protocols.
In the rest of this paper, we will explore the capabilities of
both legacy TDM and newer IP-based technology and
architectures. Then the discussion will proceed to the
how of TDM to IP migration and the requirements placed
on microwave equipment if it is to support the smooth
transition from one technology to the next.
Legacy Microwave TDM Architectures
Historically, TDM systems have been built up from
Plesiochronous Digital Hierarchy (PDH) or Synchronous
Digital Hierarchy (SDH) microwave radios, channel banks,
multiplexers, cross-connects and leased lines, often using
a layered approach. Alarm and status information usually
low data rate services are collected and multiplexed
together to fill a 56 kbps or 64 kbps channel. These channels,
along with voice tie-line services, are then combined into
a T1 or E1 line, with provision made for older facilities that
impose robbed-bit overhead. This traffic is then placed
onto a T1 facility, often with a second facility as a backup.
Dedicated equipment supports the T1 interface, and the
traffic carrier equipment (microwave or other) delivers the
traffic. Depending on the type of facility, channel sequence
may or may not be guaranteed; if not, the enterprise must
purchase additional equipment to manage this correctly.

P2

When traditional T1/E1 facilities are used to transmit

computer traffic, a Channel Service Unit/Data Service
Unit (CSU/DSU) typically is used to interface from the LAN
at each end to the leased facility. Again, the enterprise may
need to buy additional equipment to manage robbed-bit
overhead and bond across multiple T1/E1 lines, if data
rates require it.

Do You Need a Dedicated Circuit? Not Anymore

Historically, there has been a belief that a dedicated circuit
provided the highest level of assured communications,
albeit at some higher cost. This belief is no longer true.
Clearly, a circuit any circuit can fail. Worse, the failure
of a circuit carrying traffic for many different functions
(e.g. alarms, voice, data) has a more severe impact than
a circuit which serves exactly one function. On the other
hand, todays networks serve hundreds, even thousands of
functions. If all functions are implemented with dedicated
circuits, the probability of some circuit failing is quite high.
Todays network technology offers a better approach to
deliver cant-fail reliability. By combining traffic onto a
small number of high-capacity circuits and using intelligent
equipment at each end, three things are accomplished
easily and cost-effectively:

Critical traffic can be prioritized to insure

prompt delivery.

Route-diverse redundant paths can be implemented

economically.

All equipment can be monitored so that potential

problems can be detected early.

This is cost-effective, too, because the cost of the

redundancy and monitoring is shared over a large
number of circuit functions.

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

TDM Advantages & Drawbacks

Reliability and Availability

TDM will not go away tomorrow, but it will go away.

The advantage of the TDM architecture is that it leverages
the engineering that has gone into developing these
telephony systems, and it can use existing long-haul
facilities from ILECs, CLECs and other carriers.

Reliability is generally defined as the likelihood that a given

system will perform as defined over time. Availability is
generally defined as the portion of time the system is
available for use. Different industries have slightly different
precise definitions of each term, as appropriate for the
particular industry and application. However, a simple
example demonstrates the difference. Your car is very
reliable, but every 90 days (3000 miles) you take it in for
service. Its availability is therefore 89/90, or about 98.9%.

The disadvantages include reliability, scalability, suitability

for data applications, and maintenance. Microwave TDM
has its own specific set of shortcomings, too.
Reliability & Availability
Todays network technology takes a new approach to
overall service availability, one based on affordable
redundancy across relatively low-cost equipment.
While telco-grade hardware is generally very reliable,
achieving five nines overall system availability is
challenging. The claim of 99.999% availability means a
system can have only about 315 seconds of downtime
per year.
Records show that in the US, AT&T experiences about one
T1 line failure every second, due to cable cuts or cable
damage. In addition, human error adds to the problem
when routine maintenance activity inadvertently
disconnects circuits.
A given individual piece of equipment might have a
demonstrated reliability in excess of five nines and,
potentially, could deliver five nines of availability,
combining many such pieces decreases the system
availability. There are two reasons for this; one
mathematical, and one practical. As a math example,
consider a system with 20 elements in it, each one have
99.999% reliability. (An element is a line, a piece of
hardware, etc.) This system will deliver only 99.98%
roughly four nines of overall system availability.

P3

A larger system, with perhaps 100 elements, is only

99.9% available.
This math is only valid if the odds of failure of any element
are random, that is, not correlated with any other element.
However, in the real world, equipment often shares power
and physical facilities, and it can be difficult to insure that
long-haul lines are truly route-diverse. Thus, the real-world
odds of failure are higher.
Operational costs are also high, because specialized
knowledge is needed for telco systems, and the specific
parts and technology used are often quite expensive.
This high cost affects initial capital outlays as well
most TDM equipment is much more expensive than
equivalent packet-network equipment. For example,
a typical 96-channel channel bank costs US$5,600. A
comparable Ethernet switch is priced under US$2,000.

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

TDM Scalability

Suitability of TDM for Data

Scalability can be a problem with TDM equipment,

microwave and otherwise. In the case of legacy microwave
radio systems, capacity is typically hardwaredefined and
hardwareconstrained. Often procured as PDH systems,
these radios cannot be readily upgraded to SDH, much less
Ethernet. The purchase of a new TDM system can solve a
capacity problem, but without the ability to migrate the
same system to Ethernet over time, the opportunity to
leverage IP efficiencies is lost.

TDM is optimized for voice. Numerous proven techniques

exist to adapt it to data services, but all of them have one or
more drawbacks.

For other TDM hardware such as multiplexing systems, rates

above T1/E1 vary and often represent big steps, e.g. from
T1 to DS3. Wide-area networks can become a patchwork of
T1, E1, SONET, and other legacy standards, often not well
adapted to each other.

The industry developed improved versions of T1 for data

applications. Extended Super Frame (ESF) and Primary-Rate
T1 do not rob bits or mix byte order, thus they simplify the
hardware at each end. However, they use the full data rate
of the line, so it is not feasible to support traditional POTS
voice with the same facility. Multiple T1 lines can be
bonded for higher data rates, but this requires additional
specialized hardware.

Prices for leased TDM circuits are also subject to large steps
associated with the gross transitions from T1/E1 to DS3
to STM-1 / OC-3 and higher. In the low capacity T1 realm,
prices are essentially linear, with eight T1s costing about
eight times as much as a single T1.
The time-slot structure of TDM, while useful for voice,
often leads to poor capacity utilization, as channel slots
sit unused while other traffic is queued up and waiting.
Extensive research going back more than 15 years has
shown that network traffic is bursty, that is, it is characterized
by wide variations in data rate. Reserved-channel systems
such as TDM do not support this well because they cannot
add (and remove) channels rapidly enough to match the
demand. Patchwork networks of T1, E1 and SONET
exacerbate this problem.

T1 was designed originally for voice only. The total data

rate was divided into 24 channels, and a few bits were
dedicated (robbed) for signaling and control purposes.
Across a network, the order of delivery of channels was
not guaranteed. It is possible to use such a line to carry
higher-volume traffic, but the interface is more complex.

Long-Term Maintenance Concerns

Last but not least, many experienced TDM engineers and
technicians are aging out of the workforce, and younger
workers coming in are far more likely to be familiar with
packet technologies than with TDM. The Bureau of Labor
Statistics reports, for example, that almost half of the
utilities workforce will be nearing retirement age within
the next 10 years.

The ideal solution is isochronous (that is, constant-time and

deterministic) but also able to efficiently map packet traffic,
such as Ethernet, into available capacity. By definition, an
isochronous system has a fixed propagation delay through
the entire network path. The delay does not change during
the course of a connection (call), but each connection can
have a different (but fixed) delay.
Figure 1. Energy Utility Industry Workforce Age Distribution
Source: Electric Light & Power, U.S. Bureau of Labor Statistics

P4

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

Ethernet the Universal Solvent for Digital

Today one can make the assertion that any communications
equipment that does not provide support for native Ethernet
represents a dead end. Over the last 35 years, Ethernet has
evolved to be the universal solvent for data applications
Ethernet equipment easily interfaces to itself, usually with
no manual configuration required. Systems which operate
at different speeds and over different physical media share
a common interface the RJ-45 port. Standardized
Ethernet switches provide a universal interconnect and
switching mechanism that requires little or no setup.
Ethernet itself has been shown to be highly reliable;
because it also supports spanning-tree protocols, it is
easy to configure redundant paths. Even better, Ethernet
has spawned a large industry dedicated to developing
improved versions and driving down cost.
Increasingly, network service providers are installing
and offering Ethernet-compatible connections for
medium and long-haul applications, as well as local
area service. This is a positive trend but even today,
only about 13% of office buildings have access to fiber.
Of course, this is one reason why microwave has become
a popular medium for Ethernet transmission. With current
microwave systems, it is feasible to build an Ethernetcentric network using standard Ethernet equipment

at the edge, Ethernet-compatible microwave backhaul and/or

carrier-supplied leased Ethernet services when appropriate
or available. For owned networks, the availability of leased
Ethernet services presents a microwave backhaul redundancy
alternative, as will be discussed.
The Internet Protocol Family
We have seen how Ethernet has become the dominant layer 2
protocol. At layer 3, the Internet Protocol (IP) , is almost as
universal as Ethernet. Its worth remembering that not all
Ethernet traffic is IP, and IP can (and does) run across layer 2
links other than Ethernet . However, the two protocols are so
dominant that people often use one when they really mean
the other, or both.
IP and the related family of Internet protocols such as TCP and
UDP provides sophisticated mechanisms for fault tolerance,
reliable delivery, prioritization and latency control. Together
with Ethernet, IP forms a universal data transmission mechanism
suitable for all applications.
As a result, a wide range of equipment vendors, a large combined
product development budget, and a growing and well-trained
workforce support Ethernet and the IP protocol family.

Layer #

Layer
Name

Key Responsibilities

Physical

Encoding and Signaling; Physical Data Transmission;

Hardware Specifications; Topology and Design

Data Link

Logical Link Control; Media Access Control; Data Framing; Addressing;

Error Detection and Handling; Defining Requirements of Physical Layer

Network

Logical Addressing; Routing; Datagram Encapsulation; Fragmentation

and Reassembly; Error Handling and Diagnostics

Datagrams/Packets

Messages between local or

remote devices

Transport

Process-Level Addressing; Multiplexing/Demultiplexing; Connections;

Segmentation and Reassembly; Acknowledgments and
Retransmissions; Flow Control

Datagrams/
Segments

Communication between
software processes

Session

Sessions

Sessions between local or

remote devices

Presentation

Data Translation; Compression and Encryption

Application

User Application Services

Session Establishment, Management and Termination

Data Type
Handled
Bits
Frames

Encoded User Data

User Data

Scope
Electrical or light signals sent
between local devices
Low-level data messages
between local devices

Application data representations

Application data

Figure 2. OSI Reference Model Layer Summary

P5

Exalt Communications, Inc.

WP-P25-A-0909

TDM to IP Migration: Your Network, Your Timeline

TDM versus Ethernet

TDM architecture was built around the concept of isochronous
data transmission. By definition, an isochronous system has a
fixed propagation delay through the entire network path. The
delay does not change during the course of a connection (call),
but each connection can have a different (but fixed) delay.
TDM was originally designed to carry voice calls, and humans
are sensitive to delay especially to variations in delay in
conversation. TDM achieves its constant delay by reserving a
dedicated portion of the network capacity for each connection.
However, this capacity goes unused if there is no traffic.
Computer data networks, including Ethernet, do not offer a
constant propagation delay. Most computer applications do
not require it. As a result, Ethernet and other packet data
networks make better use of available network resources.

Fixed Latency? In Ethernet?

Ethernet has never had a truly fixed network-delay, or
latency time, but applications have been developed which
simulate it. A VoIP application such as Skype time-stamps
each packet on transmission. At the receiver, packets are
deliberately delayed until the correct time. Thus, a packet
which arrives early may be held for several milliseconds;
a packet which is delayed in transmission will be held briefly
if at all. Millions of trials have shown that it is quite feasible
to simulate latencies under 100 milliseconds with this
technique. One can argue that latency is not 100%
guaranteed with this technique, but the traditional
PDH/SDH hierarchy does not offer 100% guarantee
either frame slips and other events do happen.

While each approach has its adherents, the best real-world

solution lies somewhere in the middle. Some traffic, such as
voice, is delay sensitive, but other traffic is not. As a result, the
industry has converged on an architecture that combines the
best of both. By adding prioritization and timing mechanisms
to Ethernet, it has been shown possible to create large, busy
networks which carry voice, video and other time-critical
traffic rapidly enough to meet real-world requirements, while
at the same time making the best use of bandwidth by not
reserving (and potentially wasting) network resources. A wellengineered TDM network can guarantee a maximum latency
level, but in most applications a well-engineered Ethernet
network can guarantee a fixed, if not quite as low, latency level.
Computer networks address Quality of Service (QoS) at Layer 2
and Layer 3. At Layer 2, two standards work together to insure
fast packet delivery. Ethernet offers prioritized packet delivery
via its 802.1p standard. This provides eight levels of priority,
including one explicitly for voice or other time-critical traffic.
At Layer 3, a mechanism called Multi-Protocol Label Switching,
or MPLS, provides end-to-end network prioritization.
These protocols form the basis for reliable, low-latency packetbased data transmission. The proof of their workability lies in the
widespread adoption of Voice over IP (VoIP) for telephony and
video conferencing. Vonage, Skype, Cisco and many other vendors and carriers offer telephony products that deliver excellent
service over packet networks.

P6

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

Guaranteeing Latency
The Internet community has developed two key
technologies in order to guarantee latency for delaysensitive traffic, such as voice. At Layer 2, the IEEE 802.1p
committee has defined a technique to assign priorities
to Ethernet frames, and all modern Ethernet switches
support it. This provides the fundamental mechanism
for low-delay packet delivery.
At Layer 3, the MPLS standard performs a similar function
for IP packets. Together, they work well, and tests have
shown that they will effectively throttle traffic when
network performance requires it, either because too
many users are on the network, or the network
performance has been degraded due to hardware
problems, such as a failed link.
The Ethernet standard defines a Bridge device. An
Ethernet Bridge is any system which simply transfers
Ethernet frames from point A to point B. It is not a switch;
it is not a router. It is simply a point-to-point connection
that accepts Ethernet frames at one end and reproduces
them at the other. Ethernet Bridges typical transmit the
frame over a fiber-optic or radio medium.
The key point about an Ethernet bridging is that it
exists BELOW Layer 2. In short, it is a dumb device, as
far as layer 2 protocols are concerned. Thus, if a network
implements 802.1p QoS and MPLS at layer 3, the
presence of an Ethernet Bridge will not affect the
quality-of-service characteristics of the network. Any
system which connects 802.1p QoS-aware Ethernet
switches with Ethernet Bridge devices, such as radio
links, will deliver QoS.

P7

Operational and Maintenance Cost

Advantages of Ethernet
The sheer ubiquity of Ethernet has created a large
ecosystem of equipment vendors and well-trained
personnel. Collectively, this economy of scale has lowered
the cost-per-bit dramatically. Ethernet and Internet
together provide lower operational and maintenance costs,
for several reasons. Most enterprises already have made
substantial investments in equipment and personnel to
support traditional LAN and Internet services. Expanding
this incrementally to support wide-area deployments is
easier and less expensive than creating and staffing a
new TDM infrastructure.
In addition to microwave equipment, a complete TDM
infrastructure typically has statistical multiplexers to
combine low-data rate signals together to an aggregate
of 56 or 64 Kbps. These digital streams are then fed into
a channel bank and aggregated to a T1 rate, or higher.
Numerous vendors make equipment suitable to these
roles. Choice is nice, but each vendors equipment has its
own user interface, method of operation, and limitations.
Thus service personnel must be familiar with many kinds of
equipment, and must understand the difference between
robbed-bit and clear-channel systems, SF and ESF, and
other variations.
Ethernet and IP systems are self-similar; from the edges of
the network to the core, the same physical, electrical and
logical standards are used. Service personnel have fewer
interfaces and fewer technical concepts to master.

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

Ethernet Capacity Utilization Advantages

Packet-based technologies i.e. Ethernet inherently
provide a multiplexing method that allocates bandwidth
based on actual demand. When combined with a traffic
prioritization system, it can often reduce overall bandwidth
requirements in comparison to architectures which reserve
dedicated circuits for different traffic classes. In most cases,
an Ethernet network will make more efficient use of overall
capacity; thus it may not be necessary to provision as much
wide-area capacity as a TDM system would require.
The most-critical traffic is the traffic associated with alarms
and control functions. This is not high-volume traffic, just
high-priority. The Quality-of-Service (QoS) capabilities of
Ethernet allow network designers to place critical alarm and
control functions at the highest priority. This ensures that such
traffic will be delivered promptly, even when overall network
capacity has been degraded due to outage. Alarm and control
traffic typically has an aggregate data rate in the thousands
of BPS range; far less than Ethernets 10 Mbps. Thus there will
always be sufficient capacity for alarm and control.
Systems which require voice circuits (such as public safety
networks) can benefit from this as well. By placing voice
traffic second in priority (only to alarm and control) basic
voice connectivity is also assured, even in the presence of
degraded network performance.
In both cases, network designers do not need to dedicate
reserved bandwidth in order to ensure delivery of critical
traffic. By implementing the prioritization features available
with 802.1p/q and MPLS, delay-sensitive traffic can use the
Ethernet infrastructure with low latency, while still allowing
traditional data traffic to share the same lines.
Ethernet offers a second capacity-utilization advantage as
well. The step sizes of data rate increases are more rational,
and intermediate rates are supported. TDM step-sizes jump
from 64 Kbps to 1.544 Mbps to 45 Mbps. Intermediate rates
are possible, but require specialized hardware.
Ethernet moves smoothly from 10 to 100 to 1000 Mbps, with
full- and half-duplex options, and in most cases data rates
can be mixed without any special equipment. Most Ethernet
gear auto-configures as well. Even better, Ethernet will move
across non-standard-speed links, such as 802.11a/b/g/n,
without any special rate-matching equipment. Thus, wired
and wireless links can be combined easily.
P8

What is toll quality?

In the age of cell phones, Vonage and Skype, what is
toll-quality? In short, its better phone service than
you expect.
The term has historically reflected a combination of five
factors: (1) a high probability of completing the call on the
first attempt; (2) a low probability of dropping the call;
(3) relatively low delay; (4) relatively low audio noise;
(5) relatively good audio fidelity.
Cell phones changed the publics perception of toll quality;
perhaps more precisely, it changed the industrys concept
of what the public would accept in exchange for
convenience or lower cost.
VoIP technology has taken things three steps farther, and in
three directions, so to speak. Vendors such as Vonage offer
POTS-like service to consumers, often using the same
station instruments (i.e. phones) and dialing plans, but at
lower cost than offered by the phone company. Vendors
such as Cisco, Microsoft, Agere and others provide unified
messaging to the workplace. Neither of these systems has
sacrificed any aspect of toll quality.
Skype has become dominant in PC-to-PC voice and
video. Skype outperforms POTS in at least two ways: its
free, and it offers substantially better fidelity. Traditional
POTS has an audio bandwidth of about 3 KHz and uses
12-bit-equivalent sampling at 8 KHz. Skype uses 16-bit
sampling and offers 16 KHz and 24 KHz sampling rates
giving audio bandwidths of 8 to 12 KHz. The market has
demonstrated that good and free are popular, even if
Skypes call-placement reliability is not quite up to
POTS standards.
Users will demand a maximum level of delay, but there is a
point of diminishing returns. Once that delay is kept below
a certain threshold, convenience and cost become the
drivers of voice telephony.

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

The Nature of Network Traffic

Traffic on modern packet networks is primarily low-priority:
email, web access and file transfers, as shown in Figure 3
below. Thus it is feasible to assign a high priority to alarm
traffic, control traffic and voice traffic, and be assured of
good performance while still carrying large volumes of bulk
traffic. This means less overall system capacity is required,
which reduces either lease-line costs or the costs of RF links.

The benefits to antenna size and path length are further

enhanced in the event that a network operator chooses to
operate at a lower availability level (e.g. 99.9%) than typically
used for TDM traffic. Furthermore, for the same total budget
as a TDM solution, it is possible to deploy multiple radio and
wired links in a redundant configuration.
Lower Cost Redundancy
Ethernet and IP together can provide multiple levels of
fault-tolerance, and each protocol does so in a way that is
integrated and easy to implement and support.
Ethernet is a best-effort protocol that supports retransmission.
Thus, momentary link interruptions are handled and recovered
quickly. Ethernet can be configured to automatically retransmit
via an alternate route, using spanning tree protocol. This lets
you deploy parallel paths, loops, or other redundant topologies
without creating packet loops. No special switching or failover
hardware is required. Furthermore, the full bandwidth of all
links is available, unlike hot-spare systems, where one link sits
at-the-ready, but idle.

Figure 3: Relative Bandwidth Share by Category,

Daily Average, Global Downstream
Source: Sandvine, 2008 Global Broadband Phenomena

P9

What impact does this traffic distribution have on the

design of a network? It means that individual links can be
lower capacity. Because packet-based protocols are not
strictly tied to specific data rates, the links can be sized
to more closely match traffic needs, rather than to the
standard rates of a traditional TDM architecture. Many
microwave links can be operated at anywhere from
10 Mbps to more than 300 Mbps per radio channel,
as needed for the application.

Using a spanning tree and a meshed topology (see sidebar),

each node in a network has two (or more) connections to the
mesh. This significantly increases overall reliability, but without
the high cost of a full 1+1 redundancy model as is common
in TDM systems.

With the ability to provision less overall system capacity

using Ethernet, as compared to TDM, a number of
additional operational and cost benefits arise. A lower
capacity requirement means that lower channel bandwidth
and modulation levels can be used, thus increasing system
gain. That increase in system gain can translate into smaller
antenna sizes, longer paths, higher availability or a
combination of all three.

IP routing offers another level of fault-tolerance. Industrystandard routers will always find the best available path to
a destination; link failure merely causes a shift in routes.
Again, because this is a standard feature of the protocol,
implementation cost is low, and personnel are already
familiar with its operation.

Because Ethernet is agnostic about the physical layer, a

network can be designed that combines traditional leased
lines with microwave, DSL, cellular and almost any other
medium to build the overall network.

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

Another layer of fault-tolerance is available as an option

packet retransmission. While Ethernet and IP are both
connectionless, best-effort protocols, the industry standard
TCP protocol will retransmit as required to assure 100%
reliable delivery for applications which require it. On the
other hand, protocols which do not require it, such as voice,
are not burdened by the cost or overhead of retransmission.
Together, these technologies have been shown to achieve
99.999% or better reliability of packet transmission. While
individual packets are occasionally dropped, the overall
delivery reliability of the network remains a five nines or better.
Universal Application Support
In todays computing world, there is nothing that cannot
run over Ethernet/IP. Whether voice, alarms, video, remote
data monitoring, telemetry or traditional data applications,
all are supported over IP. New services are migrating to IP as
well witness the trend in storage-area networks to carry
low-level disk IO traffic over IP.
As a result, there is an ecosystem of companies investing
billions of dollars in new and improved equipment and
software. Enterprises which move aggressively to adopt an
Ethernet/IP backbone take advantage of this, in the form
of lower capital costs, lower operating costs, easier overall
maintenance and access to an enormous array of
compatible equipment.
For example, the first generation of Ethernet switches,
delivered in the early 1990s, were priced at over US$100 per
port. Today, Ethernet switches operating at gigabit speeds
cost about $25 per port.

Fault Tolerance in Microwave Networks

Traditionally, TDM architectures in microwave networks
use a hot-standby technique to deal with line failures. A
Monitored Hot Standby (MHS) line protects one or more
in-service links. This is called 1+1 protection. It works, but it
substantially increases the cost per site without increasing
the system capacity.
In contrast to MHS, diversity is used to ensure path
availability in the event of deep path fading. In a TDM
system, a deep fade incident that occurs in a non-diversity
link can cause the TDM circuit to lose clock synchronization,
in which case all data will be lost.
In a wired Ethernet network, a mesh or ring topology
provides redundant paths. Thus the failure of any one line
does not cut off traffic. Furthermore, as long as all lines are
up, the full capacity is available. Loss of one line reduces
capacity but does not cut off the flow altogether. In many
designs, the reduced capacity is not an issue, but in designs
where it is critical, QoS can be used to insure that
high-priority traffic gets delivered.
In the microwave scenario, it might appear that two
Ethernet connections to a single site is no more costeffective than MHS, but it still offers some advantages.
First, data traffic is unpredictable and bursty, so the extra
capacity is useful. Second, because redundant Ethernet
connections are easily implemented with standard
off-the-shelf hardware, beyond the additional radios
themselves, the installation cost and the maintenance cost
are low. Network maintenance personnel do not need
training in 1+1 architectures or other standby solutions.
Although a deep-fade incident is a rare scenario, the
implications of it on an Ethernet microwave system are
likely to be less severe, due to the packet retransmission
characteristics of Ethernet. In other words, Ethernet traffic
can be reliably supported at a lower availability level than
TDM traffic. Of course, it is up to the network operator to
determine whether or not to take advantage of the ability
to operate at lower availability levels.

P10

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

How Can a Network Designer Migrate to IP

Without Service Interruption?
Actually, its quite easy. Begin by deploying microwave radio
equipment that carries TDM (as TDM, not packetized) and
also can add Ethernet frames. This allows customers the
ability to add Ethernet capacity instantly to existing legacy
TDM microwave links without migrating their entire
endpoint infrastructures to IP first.
When the radios are installed, TDM traffic is cut over to the
TDM radio links, and old equipment may be taken out of
service. If there is an immediate need for Ethernet as well,
it can be added, and the equipment configured to provide
the correct amount of capacity and priority for all traffic types.
At the same time, the Ethernet capability provides capacity
for all IP data and emerging IP voice and video services. It
also provides a migration path to growing IP-based services
like VoIP and digital video. If, and when, network operators
migrate all services to full IP infrastructures, connections can
be gracefully migrated from TDM end-to-end circuits
to Ethernet.
The relative ease of migration can be shown by two
migration examples, one for enterprises operating in
a campus-type environment; one for critical
infrastructure enterprises.

Enterprise Campus Scenario

(Fortune 1000, Healthcare, Education)
Step 1. Begin by moving existing leased lines to
microwave radio systems, using equipment which
can carry native TDM as well as Ethernet. Add
additional TDM microwave links as needed to
connect to fiber connection points-of-presence.
Benefits:
(a) Eliminates or minimizes ongoing operational
costs associated with leasing.
(b) Establishes links with built-in flexibility to
expand TDM and Ethernet capacity immediately,
as needed.
Step 2. Move all inter-building LAN traffic from TDM to
Ethernet. Use the same microwave radio(s), but
shift protocols.
Benefits:
Moves traffic to more scalable protocol.
Simplifies network-wide LAN management.
Step 3. Move all voice traffic to VoIP. Shift from legacy
PBX to VoIP PBX; replace tie-lines with Ethernet.
Benefits:
Reduces operational costs by eliminating TDM
equipment. Standardizes network on a networkwide management system. Organization gets full
VoIP feature set, e.g. unified messaging.

P11

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

Enterprise Critical Infrastructure Scenario

(Utilities, Energy, Railroads)

Using Network Topologies to Protect Ethernet

Step 1. Overlay all at-capacity legacy TDM microwave links

with dual-mode TDM/Ethernet equipment.

With packet-based systems, several network architecture

alternatives are feasible:

Daisy-chain, also known as tree.

Benefits:
Increases TDM capacity immediately and provides
easy migration path to an all-IP infrastructure.

Ring

Mesh

Step 2. Upgrade any below-capacity TDM microwave links

with dual-mode TDM/Ethernet equipment.

Bounded mesh

Benefits:
Ethernet capacity is in place to handle future
migration to IP, but current TDM capacity is not
impacted its increased.
Step 3. Move all LAN traffic from TDM transport to
Ethernet; especially microwave Ethernet. This can
be done on the same radio, simply by shifting
from TDM ports to Ethernet ports.
Benefits:
Moves LAN traffic to a more scalable protocol.
Simplifies network-wide management. Enables
QoS and VLAN across the entire network. Reduces
over-provisioned capacity by moving to a shared
resource.

The use of ring, mesh or bounded mesh topologies

provides additional flexibility in the design of individual
microwave links, as compared to links used in a daisy-chain
topology. Specifically, the availability of redundant paths
means that individual link availability requirements can
be relaxed, allowing for longer paths, smaller antennas
or both. Since network path redundancy still translates
into more equipment, it often makes economic sense to
balance this additional expense with such relaxed per-link
availability requirements.
Ethernet can be configured to support Ring, Mesh,
and Bounded Mesh Topologies. As noted previously,
no additional hardware is required; it is simply a
configuration option on the Ethernet switch itself.

Step 4. Move all voice traffic to VoIP, by shifting from PBX

to IP PBX. Replace TDM trunks and tie-lines with
Ethernet. Implement on microwave.
Benefits:
Organization can take advantage of VoIP feature
set, e.g. unified messaging. Reduce OPEX costs
by eliminating TDM equipment. Further reduce
maximum provisioned capacity by moving to a
shared resource relax link design requirements.
Step 5. Move real-time traffic to IP. Prioritize traffic capacity
(using QoS) and move mission critical traffic to it.
Once all legacy TDM traffic has been migrated,
decommission legacy TDM-only equipment.
Benefits:
Real-time application performance is ensured
with QoS. All traffic and systems can now be
monitored using standard management
protocols. Retired equipment can be sold.

P12

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

Network Topology Examples

Mesh Topology

Daisy-Chain

A mesh is any network where most, if not all, nodes on the

network have at least two connections to other nodes. In
some cases, every node has a connection to every other
node. This is called fully meshed. Networks which are not
fully meshed are known as Bounded Meshes.

The daisy-chain or string topology is common in

circuit-switched networks and thus is often a default
configuration for packet-based networks that evolve from
circuit-switched networks. In this topology, there is no path
redundancy, so additional equipment must be used at
each site to increase path availability. Even with equipment
redundancy, each node is still subject to outage under
catastrophic failure conditions.

For reliability, it isnt necessary to have a fully meshed

network; it is only necessary that each node have two or
more connections to other nodes. For reasons of cost,
fully meshed networks are rarely deployed in practice.

Figure 6: Mesh Topology

Fully Meshed

Figure 4: Daisy-chain topology

Ring

Bounded Mesh

Ring topology is a special case of a string topology in which

the beginning and end nodes are one and the same. In
a ring, every node has exactly two connections, each to
its immediate neighbor in the ring. Ring topologies are
common in TDM systems, especially fiber-optic ones, but
present two challenges:

A bounded or constrained mesh network is a mesh network

in which some degree of redundancy has been compromised
in favor of affordability. A bounded mesh network thus
represents a practical trade-off between protection and
cost; for this reason, it is one of the most common packet
network topologies.

If one link on the ring becomes congested, the entire ring

may be affected.

The bounded mesh network has the additional advantage

of being able to grow organically over time, rather than
requiring a defined network design from the outset. In fact,
bounded mesh networks often grow from existing ring
networks when a pair of nodes on opposite sides of the
ring is connected. Additional nodes can be brought into the
mesh with the choice of the number of redundant paths
dependent upon the relative importance of that node in the
network. Similarly, existing nodes that are not connected can
be connected at some point in the future in order to add
redundancy and/or capacity to the network.

Large rings typically have a large number of hops, which

may increase latency in some cases.
Rings must incorporate extra equipment, layers and/or
capacity to safeguard against congestion or node
failure, and this approach carries with it an obvious
additional expense.

Figure 5: Ring topology

P13

Exalt Communications, Inc.

Figure 7: Bounded mesh topology

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

Requirements for Ethernet Microwave Links

IP Functionality

While there will always be a role for wired connections in

an Ethernet system, private wireless LAN and WAN-like
connections are becoming increasingly common in
campus enterprise environments not only for applications
where a wired connection is difficult to implement, but
simply as a cost-effective, reliable alternative to leased lines.
Of course, in critical infrastructure environments,
microwave Ethernet links are a natural evolution from
widely used TDM-based microwave backhaul.

VLAN tagging. Support for 802.1q is a prerequisite for

extending VLAN assignments across a geographically
distributed enterprise that is connected by microwave links.

In either environment, an Ethernet-capable microwave

system must meet certain requirements in order to
fully leverage the cost and reliability advantages inherent
to Ethernet operation. Additional capabilities will
further reduce deployment costs and enhance
deployment flexibility.
Ethernet System Performance
Latency. Latency and jitter must be low, on the order of
100 sec or lower; latency should not be affected by the
packet rate. Many IP applications, notably VoIP, use short
packets; it is important to be able to process packets
at wire speed in order to ensure quality of service for
latency-sensitive applications, as well as meet the
overall data rate.
Latency takes on added importance in the context of
multi-hop networks, as is common-place in long-haul
backbone deployments. Delay budgets may be as small as
1 to 2 ms; the number of hops in a daisy-chain topology
can easily number in excess of five or even ten.
Loss-of-link detection. There should be robust loss-of-link
detection and reporting.
Packet size. The system should support frame aggregation
on higher-speed links, typically above 100 Mbps. This
increases overall throughput.

P14

QoS. In order to accommodate the full range of traffic types

and apply robust traffic classification, systems should allow
quality of service to be established on the basis of multiple
metrics. At minimum, QoS should be able to be applied
on the basis of 802.1p bits, VLAN ID and source or destination
MAC address. The ability to support up to eight priority
levels and eight queues will ensure maximum flexibility in
assigning priorities.
Rate limiting. Ethernet rate limiting is a valuable tool for
ensuring that a given user group or application does not
consume too much of the available capacity.
Layer 2 switching. The availability of Layer 2 switching in the
radio both reduces cost and simplifies deployment when
compared to the alternative of using an external switch.
Interfaces
Gigabit Ethernet. In order to accommodate future capacity
scenarios, the system should support multiple GbE ports,
including 10/100/1000BaseT copper ports (RJ-45) and
fiber-optic interfaces for longer distance applications and
greater resistance to interference. To simplify deployment
and reconfiguration, all interfaces should be softwareconfigurable.
PDH. For any enterprise migrating from TDM to IP, a new
system must support the T1, E1 and DS3 requirements of
the current network as well as the Ethernet requirements
of the future network. In order to minimize sparing and
inventory requirements, the fewer models required to
support the range of PDH requirements in the network, the
better. As with Ethernet ports, TDM ports should be able
to be configured remotely via software to minimize the
expense of site visits.

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

SDH. The migration of high capacity backbone networks

will often entail a move from SDH to Ethernet, thus the
system must support multi-DS3, OC-3 or STM-1 interfaces
in addition to GbE. The same configurability requirements
mentioned above apply.
RF System Performance
High modulation rates. Whatever the business, future
throughput requirements will arise. Systems that offer
256QAM modulation will be able to wring the most
capacity out of a given channel bandwidth. This is especially
important for situations in which the selection of available
channel bandwidths is limited.
Adaptive modulation. In licensed band systems, adaptive
modulation capabilities are a desirable option, enabling the
system to deliver high data rates, yet scale down gracefully
under adverse conditions. In a network that utilizes network
path diversity, the availability of adaptive modulation allows
designers significant flexibility in path design, enabling
lower but acceptable availability levels, longer paths,
smaller antennas or any combination of the three.
Prioritization mechanisms must accommodate such
adaptive modulation and ensure the delivery of the
highest priority traffic.
System gain throughput tradeoff. Maximum peak
throughput is not the driving factor for every link. In
many cases, it would be desirable to trade off maximum
throughput for additional system gain in order to achieve
higher link availability or enable the use of smaller antennas.
Thus, a mechanism such as trellis code modulation that
enables this trade-off is a useful option.

P15

Security
Data payload encryption. Because it is a radio system and
inherently subject to eavesdropping, the system must
support strong encryption, to the National Institute of
Standards (NIST) FIPS 197 standard or better. Typically,
this means 256-bit AES encryption.
Management traffic encryption. As with the data traffic,
management must also be secure, using SNMPv3. The
management interface must also be easy to use as well
as secure.
Scalability & Evolvability
Defined latency. The system must use an over-the-air
technology that is inherently isochronous, so that TDM
support is truly low-jitter.
Adaptable scaling. The system must easily, yet correctly,
support TDM and Ethernet traffic. This makes the system
evolvable that is, it can drop into existing TDM networks,
and support service migration on a timeline set by the
network architects and engineers. There need be no rush,
no forklift upgrade.
At the same time, the system should be scalable, so that
capacity can be increased in reasonable increments as
required, without performing a forklift upgrade. Scalability
also applies to legacy TDM traffic; an ideal system is able to
carry existing TDM traffic while supporting new services
on the Ethernet side, such that traffic can be migrated
gradually, until there is no more need for TDM support.
Management
Configuration and management options must be flexible,
so that the microwave system can talk to both legacy TDM
EMS and newer network-centric NMS systems.

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

Conclusion

About Exalt

Packet-based networks are here to stay and traffic of all

types is moving over them. The trend toward unified
packet-based data communications is equally clear.

Exalt Communications designs, manufactures and markets

next generation licensed and license-exempt microwave
systems for enterprises, government organizations and
service providers worldwide. Exalt products are designed to
solve the network bottlenecks associated with the growing
demand for IP- based voice, data and video applications
and the resulting migration from TDM to IP-based networks.
With a flexible architecture and universal product platform,
Exalt provides a full range of carrier-class microwave radio
systems that meet the demand for a cost-effective and
flexible alternative to fiber and leased lines.

While an organization may have no immediate requirement

to move all of its TDM-based services to IP, it is clear that
such a move is technically feasible. Moreover, packet-based
systems can deliver the low latency and high availability
that even the most mission critical alarm and control
systems demand.
Whether the need to transition to IP is immediate or a
few years out, organizations should begin planning and
implementing this transition today by installing microwave
and other equipment that is IP-ready while still handling
legacy (and native) TDM. Fortunately, network planners
can deploy a single radio platform that supports both and
allows full migration to packet technology when the time
is right.

P16

The companys product portfolio includes a wide range

of both licensed and license-exempt radios covering all
popular frequency bands ranging from 2 - 40 GHz, capacities
ranging from 20 Mbps to 1 Gbps and multiple configurations
including all-indoor, all-outdoor and split-mount (indoor/
outdoor). Exalt systems deliver:

The economic arguments for migration to IP are just as

compelling as the technical ones. From system scalability
to equipment cost to maintenance to simplified
management, IP-based systems deliver better value
than their TDM-based counterparts.

Carrier-class availability, performance and features

So for organizations of all types, the question is clearly not

whether but how quickly to begin migrating to the all-IP
network. The good news is that whatever migration path
and timeline private network operators choose to follow,
there is microwave radio equipment designed to support
a smooth, trouble-free transition from TDM to IP.

A rich set of data networking features

Native Ethernet, native TDM or any combination of the

two to support TDM to IP migration
A high degree of flexibility, made possible by diverse
configuration options and full software configurability

The highest level of data and management security

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

EX-i Series (All-indoor)

ExtendAir

The EX-i Series of all-indoor radio systems includes licensed

and license-exempt systems offering flexible, upgradeable
native Ethernet and native TDM for low, medium and high
capacity applications with support for up to 16xT1, 1xDS3
and up to 250 Mbps full-duplex Ethernet throughput and
available Gigabit Ethernet.

ExtendAir is a complete line of entry-level licensed and

license-exempt point-to-point microwave radios systems
that, for the first time, brings high-end performance, range
and features including native TDM to the sub-$5,000
per-link category. ExtendAir is the first entry-level microwave
radio system to offer advanced features such as built-in
Layer 2 switching, 3-port Ethernet, full-speed 256-bit
AES encryption, Ethernet rate limiting and SNMPv3
management. ExtendAir native TDM capability is a first
for license-exempt systems. All ExtendAir radios support
multi-level QoS and VLAN tagging for complete traffic
management flexibility.

EX-r Series (All-outdoor)

The EX-r Series of all-outdoor radio systems includes
licensed and license-exempt systems offering flexible
native Ethernet and native TDM transport for low, medium
and high capacity applications, delivering up to 440 Mbps
sustained aggregate user throughput and available
Gigabit Ethernet.

For more information about Exalt, visit www.exaltcom.com.

EX-s Series (Split-mount)

The EX-s Series split-mount radios are frequency-division
duplexing (FDD) systems operating in ANSI/FCC and
ITU/ETSI licensed frequency bands from 6 to 40 GHz.
EX-s series radios are fully software configurable and,
with native Ethernet and native TDM, designed to support
any TDM to IP migration plan. Available interfaces include
4-16xT1/E1, 1-4xDS3, 2xOC3 and 2xSTM-1, with per carrier
Ethernet capacity up to 366 Mbps full-duplex and available
Gigabit Ethernet.

P17

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

Glossary
General Terms
Asynchronous

A term used for data transmission when there is NOT a strict
relationship between the timing of the send and the timing of the
receiver. The receiver is expected to figure out how exactly how
fast the sender is sending. Old-fashioned dial-up modems used
asynchronous transmission.
Synchronous

A data transmission where the receiver is able to receive or
otherwise recover the exact timing clock used by the sender.
Although more complex to build, it is capable of higher speeds.
Isochronous

A special case of synchronous where not only the clock is
recovered, but the exact delay in transmission is determined
and held constant. This is a requirement for voice traffic. All
isochronous circuits are synchronous, but not all synchronous
circuits are isochronous.
Circuit
Specifically, a complete electrical path; but used idiomatically to
refer to almost any type of communications channel or connection.
Latency
Loosely speaking, transmission time. More precisely, the time it
takes for a given bit in a message to travel from the transmitter
to the receiver. While most electrical signals (and radio) travels
near the speed of light, the travel time is not zero, and can be
significant. In two-way voice communications, humans begin to
notice the delay when it exceeds about 100 milliseconds, and will
find the delay very annoying when it exceeds 500 milliseconds.
Note that latency is not the same thing as transmission rate. It is
possible to have a transmission link with a capacity of hundreds of
megabits per second, but high latency. Likewise, a slow link can
be low latency.

P18

Jitter
The variation in latency. Very small amounts of jitter, on the order
of 50 ppm, usually have no effect on a system. Larger amounts of
jitter (or latency variation) must be compensated for, especially in
voice applications
Microwave

Exact definitions vary, but generally microwave refers to radio
systems operating at a frequency of 1 GHz or above.
Modulation, QAM
Modulation refers to the method by which an electrical or radio
signal is varied, or modified, in order to represent the information
being transmitted. Simple systems include the AM and FM used
by consumer radios. Modern digital system often use Quadrature
Amplitude Modulation, a technique where multiple signals are
created exactly one-quarter cycle out of phase (hence the term
quadrature) and then varied in amplitude to represent information.
Because multiple signals are being used, lots of information can
be transmitted in parallel. The number of signals is indicated by a
number, e.g. 16QAM or 64QAM. These refer to 16 and 64 bits-at-atime systems.
1+1 protection
A technique for providing a backup for a communications circuit
by having a second identical circuit operating in parallel.
MHS (Microwave Hot Standby)
Similar to 1+1 protection, as applied to radio.
SCADA (Supervisory Control & Data Acquisition)
It refers to an industrial control system that monitors, and
may control, an industrial process. It is commonly used along
pipelines and other physically-large infrastructure to monitor
performance. To the extent it is used for control, rapid response
time becomes critical.

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

Telephony Terms
Many of the terms originating in the telephony industry are
inter-related to other terms. Thus, this glossary presents them
in a logical, rather than alphabetical, order

Primary Rate

To further improve service, ESF was enhanced to created
Primary Rate.

Acronym
Meaning Definition

Digital Hierarchy
Over the years, several systems have evolved to permit the
combination of multiple digital streams into single, faster digital
streams. Collectively, such systems are called Digital Hierarchy.

POTS (Plain Old Telephone Service)

POTS is a retronym coined after the developed of more
advanced digital phone systems to refer to the traditional
analog telephone and service. POTS was the only type of
phone service until the 1970s.
PCM (Pulse Code Modulation)
A technique for converting analog signals to digital ones, used to
convert POTS calls to digital. In the US, PCM is a 64,000 bps digital
data stream.
TDM (Time Division Multiplexing)
A technique for combining (multiplexing) several signals onto the
same transmission facility. In a TDM system, each signal takes turn
using the transmission facility, in a round-robin fashion. Careful
synchronization assures the signals can be unscrambled at the
receive end.
Channel Bank

A device that combines multiple individual channels (typically
voice circuits) into a single higher-speed transmission facility.
In other words, a type of multiplexer. In the US, a channel bank
combines 24 64Kbps circuits into one T-1 circuit. In Europe,
30 channels are combined to make an E-1 circuit.
T1, E1
High-speed digital transmission systems originally developed to
carry telephone calls over longer distances without degradation. A
T1 lines operates at 1.544 Mbps, an E1 line operates at 2.048 Mbps.
CSU/DSU Channel (Data) Service Unit
The two components that act as the interface between the digital
signal and the actual physical transmission facility, e.g., the wire.
ESF (Extended Super Frame)
With the growing desire to send computer data instead of voice
over long-distance telephone wires, the telephone industry
developed an improved T1 called ESF.

P19

PDH (Plesiochronous DH)

The system originally developed by the phone company,
consisting of T1, T1C, T3, and other standards.
SDH (Synchronous DH)
Also known as SONET, this system was developed specifically to
take advantage of fiber-optic technologies.
SONET (Synchronous Optical Network)
See above. The SONET hierarchy is OC-1, OC-3, OC-12, etc, where
each number refers to the number of multiplexed 51.84 Mbps
signals. This, OC-40 is approximately 2.4 Gbps.
DSL (Digital Subscriber Loop)
When the phone company was finally compelled to offer digital
service directly to consumers at reasonable prices, DSL technology
was used to transmit digital data, at 1 to 3 Mbps, over the same
wiring used for POTS service.
ILEC (Incumbent Local Exchange Carrier)
Term used for the legacy local phone company after the
AT&T breakup.
CLEC (Competitive Local Exchange Carrier)
Term used for the other companies which are supposed to offer
phone service.
PBX (Private Branch Exchange)
A small telephone switch that supports all of the phone extensions
with an organization.
EMS/NMS (Element/Network Management System)
EMS systems were developed by the phone company to monitor
and managed telephone transmission equipment. Network
Management Systems were developed by the computer industry
to manage networks. The two systems are slowly converging.

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

Networking Terms
Ethernet
Formally known as 802.11, it is the basis of all modern
computer communications. Its key feature is that it is
completely decentralized; there is no primary controller.
Ethernet sends data in packets (called frames) of varying
sizes, and uses a probabilistic method, based on random
numbers, to determine who gets to talk and when. This
technique is called CSMA/CD.
802.11a/b/g/n
A family of standards for transmitting Ethernet traffic
over radio.
IP (Internet Protocol)
The foundation layer of the Internet. IP provides a global
system for naming endpoints and routing traffic between
them. Hence it is often called a routing layer. In the IP
system, every endpoint has an IP address, a 32-bit number
conventionally written as four decimal values (1 to 255)
separated by dots; for example, 204.50.56.113.
LAN (Local Area Network)
A term used for a collection of computers connected
over a short distance. Short is usually defined as any
distance short enough that the cost of the data
communication facility is not significant compared
to other costs in the system.
WAN (Wide-Area Network)
A term used for a collection of computers connected
over a distance long enough that the cost of the data
communication facility is significant compared to other
costs in the system.

RJ-45
Part of a family of low-cost but reliable connectors
developed by the phone company for telephone wires.
It has become the de-facto standard for modern
Ethernet systems based on cat-5 wiring.
CAT-5 (Category 5)
Originally used by Belden Wire referred to a particular
grade of light-gauge, 8-conductor cable intended for data
communications. It has since become the common term
for Ethernet cable. Cat-5 wires are terminated with
RJ-45 plugs.
QoS (Quality of Service)
In networking, it refers to the ability of a network to
guarantee a maximum transmission time, even when busy.
802.1p
A specific standard for defining QoS.
MPLS (Multi-Protocol Label Switching)
A protocol within the IP family designed to insure QoS
in a large network.
VoIP (Voice over IP)
The use of packet networks specifically Ethernet and IP
to carry telephone calls. Among the best-known examples
are Vonage and Skype.
VLAN / 802.1q (Virtual LAN)
VLANs are logically-isolated sections of a network, used
to manage traffic flows and provide enhanced security.
802.1q is a specific technique for managing VLANs
over Ethernet.

SAN (Storage-Area-Network)
Applications of networking technology directly to storage
media; that is, disks. The goal is to make the storage
appear seamlessly available regardless of physical location.

P20

Exalt Communications, Inc.

WP-TDM-A-0909

TDM to IP Migration: Your Network, Your Timeline

Exalt Communications, Inc.

580 Division Street
Campbell, CA 95008

Produced in the United States of America.

2009 Exalt Communications, Inc.
All rights reserved.
More details on Exalt point-to-point microwave
radio systems may be found at:
www.exaltcom.com

Exalt and the Exalt logo are trademarks of

Exalt Communications, Inc. Other company and
product names may be trademarks of others.
Information contained in this document may
be subject to change without notice.

P21

Exalt Communications, Inc.

WP-TDM-A-0909