Академический Документы
Профессиональный Документы
Культура Документы
DfES
Network Services Project
Videoconferencing
Draft v3.1
UKERNA manages the networking programme on behalf of the higher and further education and
research community in the United Kingdom. JANET, the United Kingdom's education and re-
search network, is funded by the Joint Information Systems Committee (JISC).
For further information please contact:
JANET Customer Service
UKERNA Tel: 0870 850 2212
Atlas Centre, Chilton, Didcot +44 1235 822 212
Oxfordshire, OX11 0QS Fax: 0870 850 2213
+44 1235 822 397
E-mail: service@janet.ac.uk
Copyright:
This document is copyright The JNT Association trading as UKERNA. Parts of it, as appropri-
ate, may be freely copied and incorporated unaltered into another document unless produced for
commercial gain, subject to the source being appropriately acknowledged and the copyright pre-
served. The reproduction of logos without permission is expressly forbidden. Permission should
be sought from JANET Customer Service.
Trademarks:
JANET®, SuperJANET® and UKERNA® are registered trademarks of the Higher Education
Funding Councils for England, Scotland and Wales. The JNT Association is the registered user
of these trademarks.
The term Linux® is a registered trademark of Linus Torvalds, the original author of the Linux®
kernel.
Macintosh is a trademark of Apple Computer, Inc., registered in the U.S. and other countries.
Microsoft®, Windows®, NetMeeting®, and NT® are either registered trademarks or trademarks of
Microsoft Corporation in the United States and/or other countries.
Polycom®, People and Content, ViaVideo®, VSX and ViewStation® are of registered trademarks of
Polycom in the U.S. and various countries.
Disclaimer:
The information contained herein is believed to be correct at the time of issue, but no liability
can be accepted for any inaccuracies.
The reader is reminded that changes may have taken place since issue, particularly in rapidly
changing areas such as internet addressing, and consequently URLs and e-mail addresses should
be used with caution.
The JNT Association cannot accept any responsibility for any loss or damage resulting from the
use of the material contained herein.
Availability:
Further copies of this document may be obtained from JANET Customer Service at the above
address.
Videoconferencing
1 Purpose...............................................................................................................5
1.1 Scope.....................................................................................................5
1.2 Target Audience.....................................................................................6
1.3 Strategic Issues......................................................................................6
1.4 Summary of Responsibilities..................................................................7
1.4.1 Schools.....................................................................................7
1.4.2 Local Authorities/RBCs............................................................8
1.5 National Education Network.................................................................11
1.6 Interoperability and Standards.............................................................12
2 Videoconferencing...........................................................................................13
2.1 Overview..............................................................................................13
2.2 Equipment............................................................................................14
2.2.1 Essential Equipment..............................................................14
2.2.2 Optional Equipment................................................................15
2.3 Videoconferencing Systems................................................................16
2.3.1 Desktop Systems...................................................................16
2.3.2 Roll-about Systems................................................................17
2.3.3 Room-based Systems............................................................18
2.4 Collaboration........................................................................................18
2.4.1 Data and Application Sharing.................................................19
2.4.2 Interactive Whiteboards and Projectors.................................20
2.4.3 Media Integration...................................................................20
2.5 Recording Conferences.......................................................................20
3 Infrastructure....................................................................................................21
3.1 Overview..............................................................................................21
3.2 H.320 ISDN Videoconferencing...........................................................22
3.3 H.323 IP Videoconferencing................................................................22
3.4 Call Admission and Control for IP Videoconferencing.........................23
3.4.1 Gatekeepers and Addressing.................................................23
3.4.2 Directory Gatekeepers...........................................................24
3.4.3 Proxy Issues...........................................................................24
3.4.4 Global Dialling Scheme..........................................................25
3.5 Local Network Configuration for IP Videoconferencing.......................25
3.5.1 Firewalls and Network Address Translation...........................26
3.5.2 Physical Separation of H.323 Traffic......................................26
3.5.3 Virtual Separation of H.323 Traffic.........................................28
3.5.4 Broadcast Traffic....................................................................29
3.6 IP Quality of Service.............................................................................29
3.6.1 Overview................................................................................29
3.6.2 Regional and National Implementation..................................30
3.6.3 Classes of Service ................................................................31
3.7 Multipoint Control Units (MCU) and Gateways....................................31
3.7.1 Overview................................................................................31
3.7.2 Gateways...............................................................................32
Purpose
Educational establishments are increasingly using videoconferencing to enhance, or in
some cases enable, their teaching, learning, research and administrative activities.
Schools, colleges and universities in the UK are finding that videoconferencing provides
an efficient means of delivering education and facilitating communication within their
organisations and beyond.
Implementing reliable videoconferencing services for schools can be complex, with a
number of different organisations involved in providing the essential underlying end-to-
end network services. There are networks on school premises, regional networks, Internet
connectivity and the National Interconnect via JANET. The whole forms the National
Education Network. At least three layers of educational management are involved:
schools, local authorities and national oversight. Suppliers include commercial network
and Internet services suppliers, local authorities, regional broadband consortia and
national agencies such as UKERNA. These agencies must work together to produce a
consistent, functional, secure IP network and interoperable videoconferencing
infrastructure across the various management domains.
This document sets out a number of recommendations for the successful deployment of a
reliable videoconferencing service of a consistently good quality across the educational
network.
A number of other existing documents are referenced. Some of these are examples of
technical design and some highlight best practice.
1.1 Scope
The document builds on the needs of the schools sector, highlighting the essential and
optional equipment; the necessity for a secure and well-engineered local, regional and
national network infrastructure and the importance of management, operational, training
and support services.
Videoconferencing is a real-time application, and the service quality is largely dependent
on the underlying network design. To be reliable, the requirements for videoconferencing
must be built in at all stages of the procurement, configuration and management of a
network. This document therefore contains recommendations that will need to be taken
into account at all these stages.
The main beneficiaries of a reliable videoconferencing service will be schools and their
pupils. The recommendations therefore apply directly to local and central equipment and
network configuration within schools, Local Authorities (LAs) and Regional Broadband
Consortia (RBCs). However, videoconferencing is an end-to-end service and it
requirements must also be considered in all procurements of products and services related
to the network infrastructure, whether bought by individual schools or on a regional or
national basis. The recommendations are therefore directly relevant to suppliers to
schools, including LAs, RBCs and commercial network and service providers.
Much of the discussion draws on the significant experience, which the United Kingdom
Education and Research Networking Association (UKERNA1) has gained over a number
of years in providing and managing high quality videoconferencing services for the UK
Higher and Further Education sectors.
The document also reflects the recent discussions, which UKERNA has had with the
Local Authorities, Regional Broadband Consortia, Becta and other agencies, and
equipment manufacturers and suppliers.
A Glossary of terms is provided in Appendix A.
The report of a recently commissioned Market Survey of Videoconferencing Systems
is included as Appendix B. The reports of subsequent Interoperability Testing of
Videoconferencing Systems are included as Appendix C.
In the areas covered by the remaining sections of this document: Infrastructure (Section
3); Network Security (Section 5.2) and Services (Section 6); schools and Local
Authorities will need to work together to determine the most effective way to deliver a
secure and reliable videoconferencing service that meets the requirements of education.
1
UKERNA manages the Joint Academic Network (JANET) and the JANET Videoconferencing Service
(JVCS), which is described in section 6.4.
1.4.1 Schools
School managers will normally be responsible for ensuring that:
• Schools and Local Authorities work together to determine the most effective way
to deliver a secure and reliable videoconferencing service that meets the
requirements of education. (Section 1.3)
• Careful thought is given, at the planning stage, to what videoconferencing will be
used for, to ensure the appropriate videoconferencing equipment and peripherals
are purchased. (2.4.3)
• All equipment purchased meets the relevant ITU-T international standards to
ensure satisfactory interoperation with equipment at other organisations. (4.1)
• Systems based on proprietary non-standard solutions are not procured. (1.6)
• Data sharing is separated from the video and audio. (2.4.1)
• For schools that do not have Broadband connectivity, the costs of ISDN provision
for videoconferencing is carefully considered, as these costs vary widely between
suppliers. (3.2)
• Unless ISDN is used through a PABX, the ISDN lines are dedicated to
videoconferencing. (3.2)
• All school IP videoconferencing systems are registered with their local authority
gatekeeper. (3.4.1)
• For IP videoconferencing, a 100Mbps, full-duplex, switched environment is
provided, and that no hubs are deployed. (3.5)
• The local distribution and proliferation of CODECs is well within the capacity of
the upstream links. (3.5)
• The NIC (Network Interface Card) configuration is explicitly set, rather than trust
auto-sensing. (3.5)
• The network path is not in contention with any other IP equipment and avoids the
local area network. (3.5)
• Testing is carried out to maintain agreed quality standards for videoconferencing.
(4.2)
• Room design and layout are given due consideration, as they play a large part in
enhancing the quality of the videoconference experience. (4.3)
• Conferences are not recorded without the explicit agreement of all participants.
(5.1)
• Schools and local authorities work together to determine the most effective way to
deliver a secure and reliable videoconferencing service that meets the
requirements of education. (1.3)
• The requirements for videoconferencing are built in at all stages of the
procurement, configuration and management of a network. (1.1)
• All of those using and operating the network are aware of and trained in their
responsibilities for security. (5.1)
• The network is well engineered with sufficient bandwidth provision. (2.1, 3.3,
3.5)
• The number of devices (switches and routers) on the path between H.323
videoconferencing endpoints is minimised. (3.3)
• A gatekeeper is provided for the registration of videoconferencing endpoints.
(3.4.1)
• All school IP videoconferencing systems are registered with their local authority
gatekeeper. (3.4.1)
• The Global Dialling Scheme, based on E.164 numbers, is used for all H.323
videoconferencing systems. (3.4.4)
• Each local authority deploys an H.323-aware firewall or proxy server as part of
their IP network infrastructure. (3.4.1, 3.4.3, 3.5.1)
• All videoconferencing sessions are routed through the proxy, if one has been
deployed. (3.4.3)
• Each local authority’s gatekeeper is registered with the National Gatekeeper. (3.4)
• Directory Gatekeepers should not be deployed at organisation (LA) level or
below. (3.4.2)
• The network is configured and maintained by appropriately skilled technical staff.
(3.5)
• All technical staff receive appropriate technical support and regular training and
updating. (3.5)
• Broadcast traffic is minimised. (3.5.4)
• An MCU is provided to support conferences involving more than two
videoconferencing systems. (3.7.1)
• Point-to-point IP conferences can be initiated by the MCU for multipoint ease of
use and security. (3.7.1)
• Service-class MCUs and gateway facilities are provided to support all scheduled
multipoint conferences. (3.7.3)
• MCUs are deployed regionally for intra-RBC conferences, as and when demand is
sufficient. [MCUs are deployed in the core of the network for all inter-RBC
conferences including IP-ISDN gatewayed conferences] (3.7.3)
• As videoconferencing becomes ‘mission critical’, MCUs are deployed in a
resilient architecture, with redundant capacity. (3.7.3)
• MCUs are not cascaded, as this is not part of the H.323 standard. (3.7.3)
• Initially service-class MCUs are deployed centrally in order to aggregate and
stimulate demand. (3.7.3)
• All centrally provided videoconferencing equipment conforms to the relevant
ITU-T Standards. (4.1).
• The appropriate security risk analysis is carried out to identify the measures to
maintain a service and the measures are implemented as necessary. (5.1)
• The size of the risk to traffic on an IP network is considered and the impact
gauged. (5.2)
• Security risks are kept to minimum by using an H.323 architecture in which the
centrally managed MCUs call out to videoconferencing systems. (5.2)
• Best practice in software updates and firewall configurations is applied to all
systems. (5.2)
• Filtering is used at a border firewall to block external access to management
interfaces where not required from external locations. (5.2)
• Videoconferencing systems have local passwords set to protect system settings
(5.2)
• All default usernames and passwords removed. (5.2)
• General advice and guidance about videoconferencing is provided to schools.
(6.1.1)
• There is a publicised means of fault reporting and that technical assistance is
available to ensure that the conferences can continue. (6.1.1)
Connecting schools in a geographic area are systems and networks controlled by a Local
Authority network, which may be combined with, or a client of, a more general-purpose
Regional Network. Connecting these regional networks together is the National
Interconnect via JANET.
Connection to the Internet should be provided at the LA/RBC or higher level; Internet
connections lower down the network are likely to cause serious operational, management
and security problems. Internet connection aggregation has clear benefits and it is
recommended that this be considered by Local Authorities.
This structure reflects the management domains within the network: who is responsible
for systems and networks at each level. It is likely that the physical network will have the
same organisation, though the locations of the boundaries may vary between different
regions and schools depending, for instance, on networking technology and management
arrangements.
Where they exist, International standards are to be preferred as they are better understood
and more likely to be supported by readily available products compliance to international
standards maximises interoperation. In this document, such standards are therefore
highlighted when appropriate. Before equipment is purchased, potential systems should
be thoroughly assessed for their adherence to standards to ensure interoperability with
other systems, and their appropriateness to provide the required functionality in the
desired environment. To assist the education community in this respect, UKERNA has
commissioned two reports from the Video Technology Advisory Service (VTAS): a
Market Survey of Videoconferencing Systems and a report on Interoperability Testing of
Videoconferencing Systems. The reports of these studies are included as Appendices B
and C respectively.
that future growth would be impeded, as links could only be made with other
organisations deploying the same proprietary equipment.
There will also be a need for locally agreed standards, particularly regarding the
management and configuration of the network. For example if a school does not allocate
IP addresses to computers in a way agreed with the authority that runs the regional
routers, then the network is unlikely to be able to transfer packets as intended. The Video
Technology Advisory Service publishes a document about videoconferencing standards,
which can be found at: http://www.video.ja.net/stan/
There are numerous bodies defining telecommunications standards, but the principal ones
relevant to videoconferencing are:
Videoconferencing
2.1 Overview
Much of our communication is visual, and being able to see as well as hear each other
significantly enhances the interaction between people. Videoconferencing enables
participants at distributed geographical locations to interact in much the same way as they
would if they were in the same location. Videoconferencing can also be used to develop
communications skills and curriculum links, as currently being explored in some local
authorities including, for instance, Hertfordshire2. Although videoconferencing can be
disconcerting for some initially, with a little experience the equipment becomes
unobtrusive and the focus of a session moves to the human interaction.
Videoconferencing takes place via the ISDN (Integrated Services Digital Network) using
the H.320 Standard, or over the IP (Internet Protocol) network, using the H.323 Standard
or Session Initiation Protocol (SIP). H.320 and H.323 are standards from the ITU series
of recommendations. SIP is a relatively new protocol that is still under development and
is therefore not yet widely deployed. For the purposes of this document, discussion of IP
videoconferencing is restricted to the H.323 Standard. An overview of the relevant
2
Using Video to Develop Communications Skills
Hertfordshire, East of England Broadband Consortium
http://www.e2bn.net/e2bn/web/e2bn_tng/e2bn_main/acon_devcomm.htm
standards is provided in Section 4.1 and details of the Videoconferencing Standards are
provided at the VTAS Web site: http://www.video.ja.net/stan/
Schools that are Broadband-enabled can use the IP network for videoconferencing,
whereas those that are not yet Broadband-enabled will need to continue to use ISDN
videoconferencing. ISDN provides guaranteed digital telephone line speeds in multiples
of 64Kbps and is widely deployed, being readily available almost everywhere in the UK.
Commonly used data rates for videoconferencing are 128Kbps (ISDN2) and 384Kbps
(ISDN6), which provides good quality. However, ISDN call charges and line rental can
prove expensive. IP videoconferencing within the education sector is currently free at the
point of use and is conducted at various data rates, most commonly 384Kbps, 768Kbps
and 1.5 or 2Mbps. Conferences at call speeds above 768Kbps will generally be of higher
quality than those at ISDN6. ISDN videoconferencing and IP videoconferencing are
discussed in more detail in sections 3.2 and 3.3 respectively.
Videoconferencing may be point-to-point, between just two systems, or multipoint,
connecting three or more systems. Multipoint conferences, and in specific cases point-to-
point conferences, require the use of a Multipoint Control Unit (MCU). Large
conferences, or those that need to be gatewayed (or ‘bridged’) between the IP and ISDN
networks, require significant MCU resources and gateway facilities. Provisioning of
MCUs is addressed in section 3.7.
Videoconferencing for teaching and learning can be implemented at a number of levels of
sophistication: from basic PC-based desktop videoconferencing systems for one-to-one
conferences, through roll-about systems for groups, to elaborate room-based fixed
systems for larger classes, see section 2.3. Each of these may also incorporate extra
functionality, such as collaboration equipment, enabling the sharing of data, video,
applications, paper documents and artefacts, considered in more detail in section 2.4.
Interactive whiteboards, when connected to a networked PC, can also be used for
standards based T.120, interactive data and application sharing during a videoconference.
Each of these three main approaches is discussed in Section 2.4: Videoconferencing
Systems. It is assumed that, in all cases, the network is well engineered with sufficient
bandwidth provision. Networking aspects are discussed in Section 3: Infrastructure.
2.2 Equipment
2.2.1 Essential Equipment
To hold a videoconference, each location must be equipped with a videoconferencing
system or ‘endpoint’. A summary of the minimum facilities required for
videoconferencing is as follows:
These basic components must be present in every system, from desktop to room-based.
Desktop systems are likely to utilise the monitor, microphone and speakers, supplied as
part of the PC. All that is required to complete such a system is a CODEC and a camera,
which may be integrated into a single compact unit. For roll-about and small room-based
systems the speakers may be integrated with the monitor, and the camera and microphone
may be integrated with the CODEC.
Careful consideration should be given to screen and image sizes, to ensure they are ap-
propriate for the type of activity and size of audience. A small image on a PC screen may
be adequate for individuals or small groups to participate together in a conference. If a
large audience is watching a remote presenter, or long frequent conferences are taking
place, then a large, high quality image is recommended. Such an image could be achieved
by projection. In this scenario, the remote presenter would only require a relatively small
image of the remote audience, particularly if a camera with a zoom function were used,
so an individual asking a question would be clearly visible.
If a presenter remains seated and just talks, a smaller image is more acceptable than if the
speaker is moving around or carrying out a demonstration. In these instances, a large,
higher resolution image is required to enable to audience to gain most value from the con-
ference.
A clock and a telephone, with an attention light rather than a bell/buzzer, are also viewed
as essential items of ancillary equipment and are often overlooked.
3
Confidence monitors are extra monitors, usually positioned at the back of the room behind the local
audience, often smaller than the main monitors, used in a teaching environment so that a teacher/presenter
can see the local and remote view. This gives the teacher/presenter the confidence that they are in shot and
enables them to see the remote audience.
Software CODECs, which use PC hardware for coding and decoding video, are being
constantly improved and, coupled with increasing processing power, are gradually
becoming an attractive cost-effective solution for desktop videoconferencing. However,
inexpensive ‘Web-conferencing’ solutions based on software CODEC and a WebCam
costing around £50, tend to be of inferior quality. Such solutions when associated with a
good quality camera, perhaps costing £200-£300, which might include pan, tilt and zoom
functionality, can produce improve quality. Hardware-assisted solutions are generally of
superior quality to software CODECs and currently retail from around £500, although
these don’t always incorporate a good quality camera. This may take the form of a PCI
card or a stand-alone unit.
• Unless a good quality camera is part of the solution, there will be little or no
control of the camera pan, tilt and zoom. Low quality cameras with small
aperture, will result in a poor quality picture;
• May be poor quality audio, if using PC microphone;
• No echo-cancellation built into the system, necessitating the use of headphones;
• Requires a recent high-specification PC with USB support, which may be a
budgetary constraint;
• Environmental requirements are not normally addressed;
• Some types are insecure (USB versions) and can be stolen easily;
• Inadequate technical knowledge on the part of the user can lead to increased
requirement for local support;
• Can be a poor experience if low quality equipment is used.
2.4 Collaboration
Various tools for collaboration can be deployed, to enrich a videoconferencing experience
beyond just ‘talking heads’, which can be monotonous and limit learning opportunities.
Collaboration tools can be as straightforward as a VCR or document camera, where the
recorded video or image is transmitted from one videoconferencing system to all the oth-
er participants in the videoconference. Alternatively more sophisticated alternatives are
available, such as networked PCs and interactive whiteboards, can be deployed, facilitat-
ing the real-time sharing and collaboration of a wider range of materials and information.
Some manufactures have developed their own proprietary solutions to enable both a
video stream of one videoconferencing system and of content, such as a Microsoft®
PowerPoint® presentation to be sent simultaneously, known as ‘in-band data sharing’.
Notable among these solutions are Polycom®’s People and Content™, and Tandberg’s
DuoVideo, which do not interoperate with each other. These solutions can be used in a
point-to-point conference providing both videoconferencing systems are from the same
manufacturers and have support for the solution. FVC™ Click to Meet™, which is
deployed in some RBCs, also provides in-band data sharing. Other equipment deployed
in the path of the videoconferencing traffic, such as gatekeepers and firewalls must also
provide support for the proprietary standard, which is unlikely. Use of these proprietary
solutions in a multipoint conference can be more problematic, as the MCU must also
support the solution being used.
A new standard, H.239, has recently been agreed for in-band data sharing. Manufacturers
will start to make this standard available in videoconferencing systems during the course
of 2004.
Interactive data sharing means the same as above, but with the additional possibility of
either/any location being able to take control of the desktop or application, and thus
update it in real-time, so that both ends see the changes. A practical example would be a
meeting in which a Microsoft® Word document under development was shared so that all
users could enter, delete or review text.
A PC that is attached to a Local Area Network (that is in turn able to access the Internet)
has become so ubiquitous that it provides the simplest and most cost-effective solution to
setting up a separate, parallel connection over the Internet for the data sharing session.
This data-only Internet session can complement an audio/video conference irrespective of
whether the conference is using the Internet or ISDN as its network connection.
The out-of-band method gives a level of flexibility and interactivity. This method helps to
avoid potential problems caused by incompatibilities between videoconferencing systems
or their software. It also allows network engineers to apply appropriate support within the
network to optimise the quality of each conference.
Careful thought should be given, at the planning stage, to what videoconferencing will be
used for, to ensure the appropriate videoconferencing equipment and peripherals are
purchased.
Infrastructure
3.1 Overview
When a videoconferencing system is installed and operational, it does not automatically
mean that the system will be able to join even a simple point-to-point conference. The
appropriate local, regional and national network infrastructures also need to be in place.
Standards compliance and interoperability with local, regional and national resources are
also essential.
ISDN lines are leased from a telecomms provider. ISDN circuits are switched, which
means a dedicated connection is established between the videoconferencing systems for
the duration of the conference, as with a telephone call. Typically ISDN operates at or
ISDN6. The costs of ISDN provision should be carefully considered, as they vary widely
between suppliers.
As well as the call charges associated with ISDN videoconferencing, there are also in-
stallation and line rental charges. Unless ISDN is used through a PABX, the lines have to
be dedicated to videoconferencing, which can make the on-going costs significant. It is
worth noting that some users have reported difficulties when using ISDN videoconferen-
cing through a PABX.
Ideally the delay caused by the local network should remain constant, as jitter (varying
delay) can result in the significant degradation of the quality of a conference.
H.323 is not very robust in the face of traffic loss. Packet loss should be below 1% for
point-to-point conferences and 0.75% for multipoint conferences. From these require-
ments it is clear that H.323 traffic must not be run over heavily congested IP networks,
(heavy congestion is considered to be a switched network running at more than 40% ca-
pacity).
SIP is a relatively new protocol; it is used by Windows® Messenger. In the future SIP
may replace H.323 as the standard for videoconferencing over IP networks, however, it is
not yet widely used in videoconferencing systems and is not addressed in this document.
Calls between videoconferencing systems registered on the same gatekeeper, i.e. within
the same H.323 zone, are resolved within that zone's gatekeeper. Calls to
videoconferencing systems outside the local zone are forwarded as appropriate to the UK
National Gatekeeper or to the World Gatekeeper. The key to the UK implementation of
the hierarchy is the UK National Directory Gatekeeper. This allows organisations within
the UK to dial each other, and to dial into, or be called from, core resources such as
MCUs and gateways.
Organisation-level (local authority and below) Directory Gatekeepers are not needed as
part of this scheme. It has been shown that these could actually hinder the correct
functioning of the addressing scheme.
It is increasingly common for organisations to run firewalls and it is common for these
firewalls to use default deny inbound policies. If a firewall is ‘H.323-aware’, ports can be
determined in real-time by monitoring the call control and setup channel. However, these
firewalls are expensive and complex to configure. If a non-H.323 aware firewall is
deployed, then a proxy should be deployed, often (but not always) supplied together with
a gatekeeper as a single composite product.
Although not formally an entity that is described in the H.323 standard, the proxy can sit
on a local authority’s border network and communicate with all of its registered H.323
CODECs. The proxy can be chosen to be the only source trusted for inbound H.323
sessions. All sessions are routed through the proxy, which translates videoconferencing
system IP addresses, effectively masking the addresses to external networks and act as
intermediaries between videoconferencing systems to provide them with a layer of
security. Proxies can also be used to allow H.323 systems to use IPv6 services on the
LAN. Both firewalls and proxies add minimal latency, but this is not significant and the
benefits outweigh the extra delay.
Reference: http://www.ja.net/schoolsbroadband/E164.pdf
The entire path from terminal to site router should be at least 100Mb/sec over Cat5 (or
later) structured cabling and should include as few hops as possible. It should not be in
contention with any other IP equipment and should avoid the local area network, either
physically or, if this is not possible, by the use of VLANs (as described below, in Sections
3.5.2 and 3.5.3 respectively).
Non-switched networks (i.e. those using hubs), unless they are lightly loaded (10%
loading of a non-switched network is significant and will adversely affect performance),
are not suitable for the demands that real-time videoconferencing traffic places on them.
The major benefit of a purely switched environment over structured cabling is that they
enable the network to transport Ethernet frames efficiently, even when there is significant
traffic.
H.323 services can utilise substantial amounts of bandwidth, especially for video, and for
peaks of data and application sharing. In addition to pure ‘bandwidth’ considerations,
H.323 services are very sensitive to issues such as latency, jitter and loss. Local network
managers will need to take care that the distribution and proliferation of CODECs stays
well within the capacity of upstream links and that no congestion occurs on those
upstream links or within the LAN links feeding the H.323 systems. They may also wish
to consider dedicated links, carefully planned switched or routed networks or prioritising
H.323 traffic, as it is generally accepted that the end-to-end available bandwidth for an IP
videoconferencing session needs to be double the selected call speed.
technical staff. Such staff will require appropriate technical support and regular training
and updating.
The following sections go into more detail concerning the issues surrounding the
planning of a well-engineered network.
It should be noted that H.323-aware firewalls tend to be expensive, and more complex to
configure and manage than a regular firewall. An alternative to the deployment of an
H.323-aware firewall is an H.323 proxy, along side the existing firewall. Proxy issues are
addressed in more detail in Section 3.4.3.
Whilst in general there seem to be increasingly fewer problems in the firewall and NAT
areas, there have been instances where the introduction of one or the other system into an
organisation has had an impact on H.323 traffic. These cases have involved failure of
traffic throughput, rather than occasional instances of loading or packet loss, so are fairly
easy to trace. In some cases the failures have been highly irregular in nature, such as all
videoconferences from an organisation suddenly terminating after 30 minutes or so. In
this particular case, the fault was traced to the interoperation between the newly installed
firewall and the videoconferencing systems, and was quickly fixed with a patch from the
firewall manufacturer.
Where private IP networks are deployed and there is a need to videoconference between
videoconferencing systems using the internal private space and external
videoconferencing systems, then there are two options:
A number of refinements can be made to this topology if the Gatekeeper is also capable
of proxying H.323 traffic. To increase security, and in the cases where schools wish to use
private IP address space for videoconferencing systems, the gatekeeper/proxy can be
moved to be in line between the switch and the Site Access Router. Videoconferencing
calls are established based on E.164 number, which for local purposes are resolved at the
proxy.
In many cases it will not be possible to physically separate H.323 and LAN traffic, and it
will be necessary for all traffic to share the same physical links. In this case there are
some methods that can be used to provide some level of protection to H.323 traffic, above
that provided to other traffic.
However, VLANs are layer 2 constructs and any traffic that needs to exit a VLAN will
require a layer 3 routing decision to be taken, in the same way as if it was a collection of
separate physical LANs connected to a router. The figure below shows the use of a
VLAN to link H.323 videoconferencing systems together.
When a VLAN truck is employed between switches, Class of Service (CoS) can be ap-
plied to the videoconferencing packets, to ensure sufficient bandwidth is allocated to the
traffic. However, in many academic institutions VLANs are used primarily in the control
of broadcast traffic and the building of segmented, secure and sub-netted networks.
Quality of Service (QoS) techniques exist in IP routing technology to mark packets; place
packets into queues that are given weighted priority; signal the network to reserve
network capacity for end-to-end QoS; match traffic to available network capacity (traffic
shaping); and perform network congestion anticipation. However, there is an issue with
scaling with some of these technologies.
Birmingham Grid for Learning (West Midlands RBC), for example, has enabled QoS
using Diffserv and an IP precedence of 4, with 512kbs reserved on the routers for video
traffic. This has been found to improve quality significantly, even when links and routers
appear to be only lightly loaded.
Reference:
http://www.wmnet.org.uk/wmnet/custom/files_uploaded/uploaded_resources/294/VC1.pd
f
JANET's national model for QoS is currently based on Differentiated Services Expedited
Forwarding (Diffserv EF), using custom queuing. A given percentage of each link on the
network is reservable on a hop-by-hop basis, including access links from the regional
networks.
Traffic marked for differentiated services is policed inbound from the regional networks;
traffic which exceeds the percentage allowance is dropped. It is therefore important for
regional networks to perform tight admission control to avoid problems due to dropped
packets.
QoS on JANET is still under development, and this model is likely to be amended or
changed based on experience. Reference: http://www.ja.net/development/qos/
At the time of writing, it is anticipated that the following classes will be available:
• IP Premium
• Best Efforts
• Less than Best Efforts
The best efforts service is that which is generally available today, and less than best
efforts supplies a service, which will attempt to use all available bandwidth, until best
efforts or premium traffic requires it.
Typical uses for IP premium are expected to include IP videoconferencing (although this
can, and does, operate well without QoS, where no heavily congested network links are
involved. See Section 3.5 for more information about adequate provisioning of links).
Some videoconferencing systems offer limited MCU capabilities within the videoconfer-
encing system (typically enabling up to 4 videoconferencing systems to participate in a
videoconference). Various types of hardware MCU are available, some of which are small
scale, usually a sealed unit. Others are chassis based and can be scaled to support a large
number of simultaneous large multipoint conferences. In some cases the scalability of
software MCUs is governed by the purchase of licences.
different call speeds, and audio and video protocols to join the same conference. During
the conference the MCU holds the state of the call. This means if the MCU fails all con-
ferences running on it will be lost. As it is an unavoidable single point of failure, reliabil-
ity and vigilant management are paramount.
A management tool, usually supplied with the MCU, is required for the scheduling of
conferences, and to assist with resource allocation and management. This tool should al-
low conferences to be scheduled to take place in the future, and, to be torn down when
they are completed.
The MCU can either allow videoconferencing systems to dial into a scheduled conference
or the MCU can initiate the conference and dial-out to the videoconferencing systems. If
dial-in videoconferencing is used, a means of circulating dial-in numbers has to be found.
Some allow for the set-up of permanent conferences with a Conference ID (Identifier).
Users can dial-in when they want to participate in a conference. This scenario would be
difficult to operate on a heavily used MCU. There are also security issues associated with
this mode of operation as it is difficult to know who is in the conference. When videocon-
ferencing is taking place over IP, initiation by the MCU can facilitate ease of use and ad-
ditional security. When a conference uses ISDN, the organisation managing the equip-
ment that initiates the conferences will be sent the invoice from the telecomms operator
for the ISDN call charges.
3.7.2 Gateways
A gateway is usually only required when one or more H.323 (IP) terminals are in a con-
ference with a non-H.323 terminal. The most common application is a gateway to an
H.320 (ISDN) network, although gateways to other types of network do exist. The gate-
way has an interface to both networks and translates the call signals in both directions,
enabling seamless communication. A gateway can be stand alone or incorporated into
MCU.
It is possible that a requirement will develop for gatewaying between H.323, H.320 and
the Access Grid, used by the e-science community. Information about Access Grid can be
found at: http://www.accessgrid.org/.
wish to deploy MCUs to stimulate local or regional demand for videoconferencing and to
identify the issues relating to booking and scheduling, technical support, operation and
management.
Service-class MCUs and gateway facilities are recommended for supporting all scheduled
multipoint conferences, especially those that are essential to the operation of an
organisation. Bearing in mind the sophistication and cost of service-class MCUs and
gateways, it is recommended that they be deployed in the core of the network for all
inter-RBC conferences including IP-ISDN gatewayed conferences, and regionally for
intra-RBC conferences as and when there is shown to be sufficient demand. As
videoconferencing becomes ‘mission critical’ it is imperative MCUs are deployed in a
resilient architecture, with redundant capacity, to keep downtime to a minimum.
It is envisaged that initially, especially for those regions that do not already have access to
MCU resources, service-class MCUs could be deployed centrally in order to aggregate
and stimulate demand. As regional and eventually local demand increases, additional
MCUs should be deployed appropriately. Throughout this process, issues of security,
resilience and scalability need to be reviewed and addressed. These will also need to be
accompanied by appropriately scaled operational and support services, as well as
management.
The implementation of the H.350 standard is still in its infancy and collaboration between
a number of national research and education networks is currently underway. It is too
early to be prescriptive about the implementation of regional directory services. Further
work needs to be undertaken in collaboration with the Regional Broadband Consortia.
Quality
4.1 Standards
In order to ensure consistent interoperability between different CODECs and between the
CODECs and the MCU and Gatekeepers, conformance to the relevant ITU-T Standards is
an absolute requirement. The detail of these standards is provided at:
http://www.video.ja.net/stan/.
The number of frames per second (FPS) isn’t necessarily an indicator of picture quality,
as the video compression algorithms allow for a trade off between FPS (which gives
smooth movement) and picture definition. Some manufacturers use their own proprietary
encoding algorithms for audio or video compression when in a conference with their own
products, and these often give a higher quality. These systems use the open standards
when inter-working with other manufacturers’ equipment.
It is recognised that such a test, while recommended, may not be appropriate or desirable
in all circumstances, especially ad hoc point-to-point conferences. In these cases, it is
clearly beneficial for those concerned to test the connection and their equipment in
advance of conferencing, in order to ensure that the conference will be of acceptable
quality. This is especially true for conferences involving international participants, where
there may be additional issues concerning equipment interoperability and different time
zones.
Whatever approach is adopted, it is essential that the expectations of those involved in the
conference be met satisfactorily. A strong deterrent to further videoconferencing is
presented if a conference is of unacceptable quality or presents technical challenges when
there is no technical support available.
Videoconferencing should not take place near to sources of low frequency noise, such as
lift shafts. Nor should facilities be sited next to crowded corridors or busy entrances.
The addition of curtains, carpets and softer furnishings than normal, all contribute to
clarifying the audio by reducing the echo. Full-length curtains covering both windows
and walls, associated with additional artificial lighting, can also prevent the wide
fluctuations of natural light, which can adversely affect the video quality.
Room colours should be mid-tone, tending towards lighter shades. Blue is frequently
used. Patterns and strong contrasts should be avoided. Where paint is used it should have
a matt finish.
Building the videoconferencing equipment into a partition wall can hide the rear of the
equipment and most of the cabling, making it more secure and safer for all participants.
Providing ducting for all cables, including floor ducting, minimises the hazards.
Cost is the major constraint when considering room refurbishment, but the environmental
and safety issues should be given serious consideration before purchasing and locating
videoconferencing equipment.
Further information about videoconferencing room design can be found in the following
document: http://www.video.ja.net/rooms/. Please note that this document is intended for
guidance only.
Security
5.1 Overview
As with any equipment that is connected to the Internet, or is available for public use, the
deployment of videoconferencing equipment has security implications. Issues to be
considered include:
• Personal privacy
• Property security
• Network security
The Data Protection Act (1998) covers issues concerning personal data. Conferences
should not be recorded without the explicit agreement of participants. Conference
booking systems or other databases or systems, which contain data of a personal nature
must be registered, and those managing such systems should have sought and obtained
the appropriate permissions, as required and specified by the Data Protection Act.
Appropriate risk analysis should identify which measures would need to be implemented,
in order to maintain an uninterrupted service.
Most networks are protected by firewalls running at the point of entry/egress to protect
PCs and other devices connected to the network. However, H.323 traffic creates problems
for firewalls, as not all call parameters are predictable in advance, as they are negotiated
during the call set-up at the beginning of each conference. There are options available for
allowing H.323 traffic to operate securely, including the use of an H.323 proxy, or an
H.323-aware firewall (see Section 3.6). It is possible to purchase H.323 modules for some
existing firewalls.
centrally managed MCU calls out to conference participants, such a risk is dramatically
reduced.
There has been a growing number of denial-of-service security incidents on the Internet
recently, well-known examples being the Blaster and Nachi worms. Such infections may
cause severe network congestion, leading to systems failure (e.g. if a firewall cannot
handle the volume of new connections being made) or degradation in service (e.g. packet
loss leading to poor audio-visual quality in H.323 sessions). The risks and threats are no
greater for H.323 videoconferencing systems than any other desktop systems in a
network. Best practice in software updates and firewall configurations should be applied
to all systems. Many H.323 systems are dedicated units with less potential for
compromise, but the quality of ‘desktop’ systems that may run a more complete operating
system (e.g. Microsoft® Windows® XP) is rapidly improving.
The use of Telnet or HTTP sessions, as opposed to SSH or Secure Hypertext Transfer
Protocol (HTTPS) sessions, means that usernames and passwords for equipment access
are sent in plain text over the network. An attacker may thus be able to snoop passwords
in transit and gain access to critical H.323 components. It is thus important that
management of the device(s) in question is performed locally, over some dedicated
infrastructure, or at least over a fully switched network where the chances of access
information being snooped are minimised. An organisation should use filtering at a
border firewall to block external access to management interfaces where not required
from external locations.
Services
6.1 Support Services
6.1.1 Advice and Guidance
Setting up a videoconferencing environment and system to ensure all the requirements
are taken into account can be complex. Therefore advice and guidance services are
required to ensure all requirements are taken into account and an appropriate solution is
deployed. Advice and guidance should cover all aspects of:
Such advice and guidance should take the form of paper/web documents, an advisory
helpdesk and on-site consultancy. There is a requirement for this type of technical support
at all levels: locally, regionally and nationally. Local Authorities are best placed to
provide advice and guidance to schools.
It is well documented that it only takes one or two bad experiences to deter users from
videoconferencing. On occasions, when participating in conferences, users will
experience faults. Past records demonstrate that faults occur mainly at the start of a
conference, especially when connecting to a conference using an ISDN system. However,
other faults with conferences do occur infrequently. In all instances, it is essential that
there is a publicised mean of fault reporting and that technical assistance is available to
ensure that the conferences can continue. This can be provided locally, regionally and/or
nationally. Those providing such support should be given (remote) access to the
videoconferencing systems and/or the MCUs participating in the conference, in order to
provide fault diagnosis and re-start the conference if necessary.
6.1.2 Training
Both technical and user training are necessary to ensure the effective operation and use of
videoconferencing systems.
Users also require appropriate training in the use of videoconferencing systems, to ensure
their confidence with the operation of the equipment. Training in effective strategies for
the use of videoconferencing as a tool for teaching and learning, is also recommended for
teaching staff, as there are pedagogical issues, which may need to be highlighted. Local
Authorities should provide such training to schools.
The service should allow for the management and deployment of the infrastructure
necessary to allow seamless connection across both IP and ISDN networks between
systems.
Operations staff with appropriate expertise will be required at regional and national levels
to maintain the network infrastructure, videoconferencing equipment and services.
Infrastructure may include VLANs, switches and routers. Equipment comprises
gatekeepers, gateways, MCUs, and servers for booking and directory services. User
services should encompass support, guidance, fault reporting and resolution, and
management of maintenance contracts. Management and operations staff within different
management domains will be required to co-operate to ensure the interoperation of
infrastructure, equipment and services.
user or system. The booking service should allow for the modification and searching of
bookings; however, this should be limited to the conferences that the school is
participating in. The booking service should provide reports of service usage.
It is likely that if many heavily used MCUs are deployed in different management
domains, multiple booking services will be needed as well. Such a scenario would be
more complex for users, as they would need to understand how to use more that one
booking service and what type of conferences should by hosted on the various MCUs (for
example, local or international conferences.
6.2.3 Resilience
MCUs and other operationally critical equipment should be deployed with resilience in
mind, keeping single points of failure to a minimum or removing them entirely. Single
points of failure should be identified, with effective workarounds and recovery plans
drawn up.
ISDN billing information should be generated from either the booking service or the
MCUs, so that invoices for ISDN call charges can be generated and sent either to LAs or
RBCs.
such bugs can cause equipment to fail). It is therefore essential that all upgrades are
thoroughly tested in a non-service environment and all issues resolved, prior to
deployment in a service environment.
However, there is also scope for the deployment of regional videoconferencing services,
as and where regional demand is identified. A hierarchical approach to the deployment of
national and regional services leads directly to a reduction in duplication of effort. Global
recommendations about hardware and software are possible, and national or regional
purchasing arrangements can be put in place, which allows equipment to keep pace with
technological developments. This maintains standardisation, which in turn ensures the
interoperation of all hardware and software.
how the service will continue operations and resilience be provided when equipment fails
and is being repaired.
It should be noted that some organisations that typically use H.323/IP videoconferencing,
view it as business critical and have ISDN lines installed to provide a backup, in the event
of a failure of their network. In such cases ISDN lines have to be regularly tested and
brought into use, to ensure their continued availability.
As demand and use grows, more videoconferencing systems will be procured locally, and
support services and central equipment must be scaled to cope with increased demand.
Registration of all systems and booking of conferences provides valuable statistical data
that can be used to predict growth and changes in usage trends. As the speed of such
growth is unknown, the capacity of the central equipment should be kept under continual
review, to ensure scaling of the service to meet demand.
Large-scale regional procurements are likely to take place through the European
procurement process, or a purchasing agreement that is put in place as an outcome of a
procurement exercise. Time and resources to undertake such procurements should be
factored into capacity upgrades.
H.323 videoconferencing, in particular, is an area where new standards still being agreed
and manufacturers are developing new hardware for MCUs. Provision for MCU
hardware upgrades should be factored into the cost of service operation. It is not usually
possible to freeze MCU hardware, as this may lead to interoperability issues between
hardware and software, and between MCUs and videoconferencing systems. It would
also mean that new, potentially useful features could not be offered to users.
6.3.3 Liaison
Aggregation of supply can bring about closer, more productive, working relationships
with manufacturers of all videoconferencing equipment. One outcome of which is an
improved response to feature requests. Liaison with the manufacturer and supplier
organisations ensures roadmaps are made available in a timely manner, which allows for
the planning of upgrade paths and timescales. Effective communication facilitates
appropriate engineering support during upgrades and timely bug fixes.
6.3.4 Reporting
Reporting requirements are dependent on the Service Level Agreements and monitoring
that have been put in place, and also on what the data is to be used for (e.g. capacity
planning, trend prediction). Statistical information can be provided by the MCU,
providing the software and licensing are in place, by a booking service and by a helpdesk
database. Information that can easily be made available, on a monthly basis, is:
6.3.5 Review
The overall service provision should be reviewed regularly, to ensure it meets the needs
and requirement of the users. The means of collecting feedback will vary according to the
aspect of the service under review.
All contracts, including those with manufacturers and suppliers of both goods and
services, should have a review period incorporated into them. The review period will vary
according to the nature of provision and size of contract.
• Monitoring;
• Reporting.
It is anticipated that a Conferencing on Demand Service would not require sites to pass a
JVCS Quality Assurance Test, and would allow any H.323 compliant videoconferencing
system to use resources. The service would require sites to undertake a minimal initial re-
gistration, which will be limited to appropriate JANET connected sites. It is expected
that a Conferencing on Demand pilot will be launched in Q2 2004.
The UKERNA Automated QA Test Project is assessing the feasibility of providing a ded-
icated on-line audio and video analysis tool for use within the JANET community. The
system will provide feedback to users on their current audio and video levels. The Auto-
mated QA Test system will complement (not replace) the current JVCS Quality Assur-
ance procedure. The Automated QA Test system would be enabled via the Conferencing
on Demand scheduling tool.
References
DfES Standards Fund Guidance
ICT in Schools Standards Fund Grant 2004-05
Guidance for Schools and LEAs
http://www.dfes.gov.uk/ictinschools/funding/
Videoconferencing Standards
JANET Video Technology Advisory Service
http://www.video.ja.net/stan/
What the research says about video conferencing in Teaching and Learning
Becta Research Report
http://www.becta.org.uk/research/reports/docs/wtrs_vidconf.pdf
Videoconferencing
Learning and Teaching Scotland
http://www.ltscotland.org.uk/services/videoconferencing.asp
What is a firewall?
Becta ICT Advice
http://www.ictadvice.org.uk/index.php?section=te&cat=007000&rid=625
H.264
http://www.lsilogic.com/products/islands/h264/H.264_MPEG4_Tutorial.pdf
http://www.wipro.com/insights/mpeg4videocoding.htm
MPEG-4
http://www.chiariglione.org/mpeg/standards/mpeg-4/mpeg-4.htm
Cisco Glossary
http://whatis.techtarget.com/
Webopedia
http://www.webopedia.com/
Network Design
DfES ICT in Schools Network Services Project
UKERNA, March 2004
Network Security
DfES ICT in Schools Network Services Project
UKERNA, March 2004
Appendix A: Glossary
This glossary explains the terms used in this document. An extensive videoconferencing
glossary can be found at the JANET Video Technology Advisory Web site:
http://www.video.ja.net/resources/glos.html. An extensive general networking glossary
can be found at the JANET National User Group Web site:
http://www.jnug.ac.uk/netglossary.html
Access Router
A router, commonly located on a customer site, which provides the connection
point to the service provider's network.
Address
In this document refers to an IP address. An IP address is the unique layer
identifier for a host on the local IP network.
Address Resolution
A means of converting the address of a videoconferencing system into the
corresponding IP address.
Admission Control
Controls configured on equipment in the videoconferencing network enabling a
videoconferencing system to make a call/receive calls or blocking these features.
Aperture
The camera lens opening, the size of which is either fixed or adjustable.
Auto-sensing
This procedure involves probing the capability of the network using low-level
signalling techniques to select compatible Ethernet speeds.
Auto-tracking
This describes a feature of video cameras that track and focus on the current
speaker, usually either by sensing movement or voice position.
Bandwidth
The capacity of a communication link (normally measured in bits per second).
Best Efforts
A network service where no packet is treated any differently from any other; no
guarantees about speed of delivery or even of actual delivery are made. Often
used in an IP QoS context to refer to the default network service.
Border
The boundary between one network management domain and another; the
connection between the RBC and the National Interconnect is the border between
the RBC and JANET.
Bridge
A connection between three or more conference sites, enabling traffic to pass
simultaneously between all the sites. Videoconferencing bridges are MCUs
(Multipoint Control Units) and Gateways.
Broadband
A transmission medium capable of supporting a wide range of frequencies. It can
carry multiple signals by dividing the total capacity of the medium into multiple,
independent bandwidth channels, where each channel operates only on a specific
range of frequencies. [Source: RFC1392]
The term has been adopted in common usage to refer to connections to the
Internet at speeds above 128Kbps.
Broadcast
A transmission to multiple, unspecified recipients. On Ethernet, a broadcast
packet is a special type of multicast packet which all nodes on the network are
always willing to receive.
Call Control
Refers to the signalling involved in setting up, monitoring, transferring and
disconnecting a videoconferencing session.
Cascading
Usually used to describe the cascading of MCUs, to support a large multipoint
videoconference, involving more participants than a single MCU will support.
Cat5
Category 5 cabling is an American National Standards Institute (ANSI) standard
for unshielded twisted pair cables.
Chair-controlled
In this mode of videoconference operation, one conference participant controls
who is allowed to speak by means of either hardware or software control.
Collaboration
Various tools for collaboration can be deployed in a videoconferencing
environment to facilitating conference participants to work jointly on documents
that are shared electronically.
Composite Video
A type of video signal in which all information - the red, blue, and green signals
(and sometimes audio signals as well) - are mixed together. This is the type of
signal used by televisions in the United States.
CP (Continuous Presence)
Continuous presence allows a conference participant to view multiple participants
on one screen at the same time.
Data Sharing
Data sharing is the simultaneous display of the desktop (screen) of one computer
(or of an application running on that computer), on another display at a different
location, with changes updated on both displays in real time.
Decoding
The conversion of the digital code carrying the sound and vision, back into
analogue mode.
Delay
The time taken for a signal to pass through a network, from the sending
videoconferencing system to the receiving one.
Directory
The implementation of directory services can support the association of
individuals with videoconferencing systems, searchable pages, and clickable
dialling. Directory services can also assist in the configuration of
videoconferencing systems, and user authentication based on authoritative data
sources. H.350 defines a directory-services architecture for multimedia
conferencing for H.323, H.320, SIP and generic protocols has recently been
ratified.
Directory Gatekeeper
The deployment of directory gatekeepers in the videoconferencing network
enables videoconferencing systems’ addresses to be resolved and calls routed
without every gatekeeper being aware of all other gatekeepers. Gatekeepers are
deployed hierarchically, with the National Gatekeeper providing directory
services for international videoconferencing.
Domain
On the Internet, a domain consists of a set of network addresses. This domain is
organized in levels. The top level identifies geographic or purpose commonality.
The second level identifies a unique place within the top-level domain. Lower
levels of domain may also be used.
Doubletalk
When a number of participants speak simultaneously in a videoconference.
Videoconferencing systems should support doubletalk, to allow natural
conversation.
Echo Cancellation
The CODEC delays the vision signal by approximately 200 milliseconds. To
maintain sound/vision coincidence the audio signals are delayed by a similar
amount. This time delay produces unacceptable echo into the conference. Echo
cancellation is introduced electronically to reduce this echo to a workable level.
The conference environment also influences the amount of echo, so echo
cancellers need to adjust to the acoustics of the conference room in use: this
process is termed ‘training’.
Encoding
The conversion of the analogue sound and vision signals into digital code.
Endpoint
The network device at the end of a network connection, such as a
videoconferencing system. The local endpoint is the videoconferencing system
sending your video and audio signals, and receiving video and audio signals from
other, remote, endpoints involved in the videoconference.
End-to-end
A phrase used when reviewing the provision of a reliable service between two
hosts, usually across multiple management domains. All aspects of network
provision must be considered to ensure traffic passes between the hosts with as
little delay as possible.
Ethernet Frames
Used on a LAN to determine how a packet is put on the network. Ethernet
provides four different frame types. These are not compatible with each other. For
two hosts to communicate with each other they must use the same frame type.
Firewall
Router or access server, designated as a buffer between any connected public
networks and a private network. A firewall router uses access lists and other
methods to ensure the security of the private network.
Frames
A logical unit of information, sent by the Data Link layer over a transmission
medium. The term often refers to the header and trailer, employed for
synchronisation and error control that surround the data contained in the unit.
Full-duplex
A type of duplex communications channel which carries data in both directions at
once.
G.711
ITU.T standard for encoding audio signals for voice grade transmission (300-3400
Hz) encoded at data rates of 56 or 64 Kbps. This is the main encoding algorithm
for digital telephony.
G.723.1
ITU.T standard for the narrow-band encoding of audio signals with data rates of
5.3 Kbps or 6.4 Kbps.
G.728
ITU.T standard for the low bit rate coding of audio signals producing 3.4KHz
upper frequency limit but occupying only 16Kbps of bandwidth.
Gatekeeper
The component of an H.323 videoconferencing system that performs call address
resolution, admission control and bandwidth management. A gatekeeper maintains
a registry of devices in the multimedia network. The devices register with the
gatekeeper at start up and request admission to a call from the gatekeeper.
Gateway
A component of a videoconferencing network that provides conversion between
different videoconferencing standards, typically H.320 ISDN videoconferencing
and H.323 IP videoconferencing, enabling ISDN and IP systems to participate in
the same conference.
H.261
For audio visual services; this defines the way in which the picture information is
compressed and coded to enable transmission over low bandwidth networks. It is
the baseline coding which is mandatory for most videoconferencing systems to
ensure interoperability at a basic level.
H.263
For audio visual services, a variation of the H.261 CODEC but specifically
designed for low bit rate transmission, i.e. H.324 (GSTN) and H.323 (IP)
networks at 64-128Kbps.
H.264
Also known as MPEG-4 (AVC) Advanced Video Coding. The latest video
CODEC developed jointly between the ITU-T and ISO/IEC. It uses more
sophisticated compression techniques than H.263 coding and is designed to
require less bandwidth for an equivalent quality signal using other compression
algorithms.
H.320
The umbrella ITU-T standard for narrow band videoconferencing interoperability
over ISDN networks.
H.323
The umbrella ITU-T standard for narrow band videoconferencing interoperability
over IP networks.
H.350
H.350 defines a directory-services architecture for multimedia conferencing for
H.323, H.320, SIP and generic protocols.
Hop
Passage of a data packet between two network nodes (for example, between two
routers).
In-band
A connection made to a device, such as a router, which uses the network of which
the device itself is a component. Failure of parts of the network may make in-
band connections to some devices impossible; out-of-band facilities are generally
provided to address this problem.
Internet
The global public network comprising many interconnected, but independently
operated, service provider networks.
IP (Internet Protocol)
The basic protocol for the transmission of information in the Internet. IP specifies
how data is fitted into packets and provides features for addressing, type-of-
service specification, fragmentation and reassembly, and security. Defined in
RFC 791.
IP Premium
IP Quality of Service (QoS) term for a network service offering priority treatment
of particular traffic over others.
Jitter
The inter-packet delay variance; the difference between inter-packet arrival and
departure. Jitter is an important QoS metric for voice and video applications. Jitter
can cause data loss, particularly at high speeds.
Kbps
Kilobits per second, a unit of transmission speed equivalent to 1024 bits per
second (bps).
LA (Local Authority)
A UK regional body which may operate its own local network providing service
directly to schools.
Latency
The time delay introduced into a videoconference by the time taken to code and
compress (and decode and decompress) the video and audio signals in the
CODEC.
Layer 2
A networking term referring to the network protocol operated immediately over
the layer 1 infrastructure. Ethernet or PPP are examples of layer two
technologies.
Layer 3
The protocols operated over the layer 2 network, such as IP.
Loss
This describes an error condition in which data packets appear to be transmitted
correctly at one end of a connection, but never arrive at the other. This might be
due to network conditions being poor and the packet becoming damaged in transit
or the packet being deliberately dropped at a router because of internet
congestion.
Management Domain
A discrete network, which may be connected to other networks, over which a
network administration team have responsibility for setting, implementing and
maintaining policy.
MPEG4
A low-bit-rate compression algorithm intended for 64Kbps connections.
Multipoint
Communication configuration in which several videoconferencing systems are
connected.
Out-of-band
Method of connecting to devices providing a network service without using the
network itself. For videoconferencing out-of-band usually refers to data sharing,
so the videoconference and data sharing session establish separate connections.
Packet
The unit of data sent across a network. ‘Packet’ a generic term used to describe
unit of data at all layers of the network, but it is most correctly used to describe
application data units. [Source: RFC1392]
Packet Loss
The discarding of data packets in a network when a device is overloaded and
cannot accept any incoming data at a given moment.
Packet Filters
A tool provided on many routers and switches to control the flow of packets in a
router. Often used to implement a simple firewall by restricting access over an
interface to particular IP address ranges and/or services only.
Point-to-point
The simplest method of videoconferencing, where one site communicates with
one other.
Policing
Process used to measure the actual traffic flow across a given connection and
compare it to the total admissible traffic flow for that connection. Traffic outside
of the agreed upon flow can be tagged and can be discarded en route if congestion
develops. Also known as admission control.
Port
A logical channel or channel endpoint in a communications system. The
Transmission Control Protocol and User Datagram Protocol use port numbers to
distinguish between different logical channels on the same network interface on
the same computer. Some port numbers are defined in RFC 1700, divided into
well-known ports and registered ports.
Private Address
An IP address that is used within a network and is translated using NAT at some
boundary to give a level of anonymity to the source. Typically a Private Address
should be from a list defined in RFC 1918.
Private Peering
Interconnection between two service provider networks using a dedicated link;
for example, the connection between an RBC and JANET for the National
Schools Interconnect is a private peering.
Proxy
Intermediary program that acts as both a server and a client for the purpose of
making requests on behalf of other clients. Requests are serviced internally or by
passing them on, possibly after translation, to other servers. A proxy interprets,
and, if necessary, rewrites a request message before forwarding it.
Public Address
An IP address that is unique on the global Internet. Public addresses are usually
obtained from a network's Internet Service Provider, which in turn obtains blocks
of addresses from regional Internet registries.
Real time
Describes an application which requires a program to respond to stimuli within
some small upper limit of response time (typically milli- or microseconds).
Router
Network layer device that uses one or more metrics to determine the optimal path
along which network traffic should be forwarded. Routers forward packets from
one network to another based on network layer information. Occasionally called a
gateway (although this usage is becoming outdated). Often used as a generic term
for an IP router, however the term may be used to refer to a device that is routing
other protocols in addition to IP.
RS-232
A standard interface for connecting serial devices.
Streaming
Playing sound or video in real time as it is downloaded over the Internet as
opposed to storing it in a local file first. A plug-in to a web browser decompresses
and plays the data as it is transferred to your computer over the World-Wide Web.
Streaming audio or video avoids the delay entailed in downloading an entire file
and then playing it. Streaming requires a fast connection and a computer powerful
enough to execute the decompression algorithm in real time.
Structured Cabling
A telecommunications cabling system that can support virtually any voice, video,
imaging, or data application that an end-user chooses.
SuperJANET
The national high speed backbone of the JANET network.
S-Video
S-Video, sometimes referred to as Y/C Video, or component video is a video
signal transmission in which the luminance signal and the chrominance signal are
transmitted separately to achieve superior picture clarity.
Switch
A device that channels incoming data from any of multiple input ports to the
specific output port that will take the data toward its intended destination.
T.120
An ITU-T standard that describes data conferencing.
Telnet
The Internet protocol for remote login. Runs on top of TCP/IP. Defined in RFC
854 and extended with options by many other RFCs.
Terminal
Any device that terminates one end (sender or receiver) of a communicated signal.
Traffic
The term used to describe data as it traverses the network.
ViDe
The Video Development Initiative in North America. See http://www.vide.net/
Voice-switched
This mode of conference control relies on the MCU switching to transmit the
video and audio streams of the current speaker to all conference participants (with
the exception that the speaker sees the video of the previous speaker).
WebCam
A video camera connected to a video capture card in a computer.
Zone
An H.323 zone is a logical collection of videoconferencing systems, gateways and
MCUs, which are managed by a single gatekeeper. A zone must include at least
one videoconferencing endpoint and may include several LAN segments
connected by routers.