A beginner's guide by Richard Lewis, IP Product Manager, Video Imaging Supplies Ltd
This month we look at IP Networking for CCTV – the part of the IP jigsaw that provides the most exciting opportunities for CCTV.

Although the economics and functionality of IP cameras, servers, and software bring many new capabilities, it is the networking capability of IP that is the greatest single advance over conventional technology. IP networks, and their ability to connect to other networks, enable camera pictures to be viewed remotely, even from the other side of the world if required, easily and cost effectively.

The traditional network for CCTV cameras, usually referred to simply as cabling rather than a network, has the following characteristics:

  • it is known as a "dedicated network", meaning that it is used exclusively for the CCTV system;
  • the layout of the network is designed specifically for the physical design of the CCTV system – cameras, and a control room at the centre. Any changes to the positioning of the cameras or control room require changes to the cable runs;
  • its coverage is almost always local – restricted to the site where the system is installed, with no links to any other sites or systems.

By contrast, IP networks for CCTV are generally:

  • shared networks rather than dedicated networks – the network used is the company's IT network, known as the LAN (more details below), which is also used to carry other data for many other purposes. Responsibility for the management and control of the network lies typically not with the security department but with the IT department;
  • the network reaches most parts of the site, and additions or changes to camera and monitoring locations can often be handled with minimal cabling changes;
  • although the CCTV system is built on the local network, most of today's local networks link into wider networks including the Internet, giving the ability to be viewed on a global scale.

When considering what subjects to cover in an introductory article on Networking, there is a temptation to write "Installing IP cameras, servers and software is simple. All you need is to get an IP address for each device from the IT Manager, and the rest is plug and play using installation wizards".

For basic installation, that is roughly true. But the installer who understands what is happening on an IP network is more likely to be successful in securing the more lucrative contracts while providing the best support to the customer. Therefore, in this article we explain the common terms of IP networking, and how they fit together.

LANs – Local Area Networks
A Local Area Network is a digital network used within companies to link together computers and their associated devices, such as printers and servers. It is often called "the computer network" of an organisation. When a new device is needed, it is a simple matter to plug it into the LAN and thus link it to all the other devices on the LAN. LANs vary in size, from a link between a couple of computers and a single printer to up to hundreds of computers and additional devices.

Most LANs use the Ethernet networking standard, the world's most popular networking system. (The other standard is called Token Ring, but will not be covered here because 90 per cent of LANs use Ethernet.)

Ethernet cabling can be of one of four types: twisted pair (with two pairs required for connection), thin co-ax, thick co-ax, or fibre-optic cable. All these run typically at 10 Mbits per second, (although LANs running at 100 Mb/s are now available). However they differ in the distance over which they can operate and the maximum number of connections that can be made per segment, where a segment is a single run of cable. These figures show the maximum run length per segment in metres and the maximum number of connections per segment:
Twisted pair: 100m – 2
Thin co-ax: 185m – 30
Thick co-ax: 500m – 100
Fibre optic: 1000m – 2

Most LANs will have a mix of cable types according to the level of traffic on the various parts of the network. Although a LAN may have just a single segment, most are made up of a number of segments joined together by devices called hubs. The design and maintenance of the LAN will be the responsibility of the IT department, so the CCTV installer will be unlikely to be involved in such activities.

As their name suggests, LANs are for local communication – an office or a building – and the figures above show the distance limits. They are also private networks, built and controlled by the owners. However, because LAN network standards and protocols can be used over public networks, they can be connected to long distance communications links to form Wide Area Networks for wide access.

WANs – Wide Area Networks
A Wide Area Network is a network that connects LANs together over hundreds or thousands of miles. These links are usually leased line connections – point-to-point high bandwidth links rented from public communications carriers, such as KiloStream or MegaStream from BT.

Companies rent by the month or year and pay a fixed rate independent of the usage made of them. Leased lines can link UK office LANs together, or international lines can link LANs in different countries or continents.

Where links between LANs are temporary rather than fixed – for example a data download between LANs, then dial up links can be used. ISDN is a popular option, giving a higher data rate than standard dial-up telephone (known as the PSTN – Public Switched Telephone Network) Internet links
Another way of opening up a LAN to remote access is to link it to the Internet. Most LANs are now connected so that PCs on the network have access to email and the World Wide Web. Companies usually link their LANs to the Internet with a reasonably high bandwidth leased line, rather than the very low bandwidth connections used for personal Internet access. With such a link already in place for other purposes, the pictures from an IP camera are automatically available from any PC connected to the Internet, provided it is authorised or running the appropriate software.

Bandwidth
When discussing IP networking of CCTV, the issue of bandwidth must always be considered. Bandwidth is the measure of the volume of data. The key volumes are the maximum volume that the LAN can carry (typically 10 Megabits per second), and the volume of data the cameras and servers create. In our first article, we pointed out that CCTV could have quite a high data requirement, like all applications using images.

Few installers need go to the depths of knowledge of network experts … but an understanding beyond a blind plug and play approach will pay dividends

In general terms, our second article dealt with compression techniques and how H261 and Wavelet had low bandwidth requirements, Motion-JPEG high requirements, and MPEG was very high. Specifically, the bandwidth required depends not only on the compression method but also on the application and the number of cameras.

Each IP CCTV system needs to be carefully sized for bandwidth and discussed with the IT department to ensure that the existing LAN can handle the additional traffic.

Addressing
A fundamental concept of IP is the "addressing" of hardware devices – cameras, servers, and other equipment – on a network. Instead of cameras being known as 1, 2, and 3, as on a conventional network, IP addresses must always be in the strict fixed format of: nnn.nnn.nnn.nnn where n is any number between 0 and 9.

The first two groups of three digits are fixed for each LAN. They are assigned by a world body, the Internet Corporation for Assigned Names and Numbers (ICANN), and operate like a telephone number. They are the number by which any site in the world that is connected to the Internet can be addressed, and for this reason they cannot be changed. (In a LAN that does not connect to public networks the user can choose their own numbers, but few networks are now designed for private use only).

The third and fourth group of numbers are controlled by the LAN Manager. When an installer adds an IP camera or server to the network, an IP address will be required, and the Manager of the LAN, typically the IT Manager of the company, will specify the number to be used.

TCP/IP and the seven layer model
Just as Ethernet is by far the commonest standard for networking, so TCP/IP is the dominant protocol for communicating on networks.

TCP/IP stands for Transmission Control Protocol/Internet Protocol. That means little to most people, except for the word Internet. TCP/IP is the basic communication language or protocol of the Internet and of most private networks.

At its simplest, a protocol can be considered like a language. All parties involved in the conversation have to know which language is being used, and able to speak it.

TCP/IP is a two-layer program. The higher layer, Transmission Control Protocol, is concerned with managing "packets". When data is sent over the Internet it is not sent as a total file or message. It is split into small packets of data, each of which is sent separately, possibly even by a different route to other packets, and then reassembled at the destination in the correct order. This is called "packet switching", and is the most efficient way of using the capacity of a data network.

The lower layer, Internet Protocol, manages the address part of each packet. For the packets to get to the right destination each needs to have its IP address. As the packets are sent around the world on the Internet, or around the LAN, the addresses tell the network where to route the packet.

The Seven Layer OSI Model

Anyone working with IP networks will soon need a framework in their mind that fits together all the many terms and functions they pick up along the way. The model that the industry uses is called “the Seven Layer OSI Model”, where OSI stands for Open Systems Interconnection – the principles that allow networks to connect and communicate. The OSI Reference model defines seven layers that describe how computers and other devices on a network communicate with each other. The model applies to all network protocols, not just TCP/ IP, and all networking standards, not just Ethernet. The Seven Layer OSI model is a very useful map on which to place a number of potentially confusing terms and concepts:
Layer 7 Application Layer
Layer 6 Presentation Layer
Layer 5 Session Layer
Layer 4 Transport Layer
Layer 3 Network Layer
Layer 2 Data Link Layer
Layer 1 Physical layer

The Layers are usually described from the bottom up:
Layer 1 - Physical Layer – provides the transmission of data. This layer defines electrical and physical properties. Examples include: coaxial cable, twisted pair, CAT 5, fibre optic cable
Layer 2 - Data Link Layer – controls the transmission of blocks of data between network devices over a physical link. It checks for and corrects any errors that occur on the physical layer. Examples include: Ethernet
Layer 3 - Network Layer – routes data from one network device to another. Examples include: IP, network routers
Layer 4 - Transport Layer – ensures that data arrives at the destination correctly and in proper sequence. Examples include: TCP, UDP
Layer 5 - Session Layer – enables applications to synchronise and manage their communication and data exchange. Examples include:, SMTP, FTP, SNMP
Layer 6 - Presentation Layer – This is where application data is either packed or unpacked ready for use by the running application. Protocol conversions, encryption/ decryption and graphics expansion all takes place here. Examples include: MPEG decoding
Layer 7 - Application Layer – This layer directly serves the end user. It supports end applications. Examples include Microsoft Word and Outlook. To relate this to the idea of packets of data – before a packet is sent from device A to device B, it needs to take data from every layer of the network taking on headers and footers that specify the various protocols being used. When it arrives at its destination, each of the headers and footers is read by the appropriate level of the network and stripped off as it is passed to the next level up. Finally on the detail of TCP/IP, it is in fact a suite of protocols that also embraces:
HTTP - the World Wide Web’s Hypertext Transfer Protocol
FTP - File Transfer Protocol
SMTP - Simple Mail Transfer Protocol
UDP - User Datagram Protocol
PPP - Point-to-Point Protocol
IGMP – Internet Group Management Protocol
SNMP – Simple Network Management Protocol