Archive for the ‘IT Tech explained’ Category

Cat5e and Cat6 Comparision

Wednesday, April 7th, 2010

Why do I need all the bandwidth of category 6? As far as I know, there is no application today that requires 200 MHz of bandwidth.

Bandwidth precedes data rates just as highways come before traffic. Doubling the bandwidth is like adding twice the number of lanes on a highway. The trends of the past and the predictions for the future indicate that data rates have been doubling every 18 months. Current applications running at 1 Gb/s are really pushing the limits of category 5e cabling. As streaming media applications such as video and multi-media become commonplace, the demands for faster data rates will increase and spawn new applications that will benefit from the higher bandwidth offered by category 6. This is exactly what happened in the early 90’s when the higher bandwidth of category 5 cabling compared to category 3 caused most LAN applications to choose the better media to allow simpler, cost effective, higher speed LAN applications, such as 100BASE-TX. Note: Bandwidth is defined as the highest frequency up to which positive power sum ACR (Attenuation to Crosstalk Ratio) is greater than zero.

What is the general difference between category 5e and category 6?

The general difference between category 5e and category 6 is in the transmission performance, and extension of the available bandwidth from 100 MHz for category 5e to 200 MHz for category 6. This includes better insertion loss, near end crosstalk (NEXT), return loss, and equal level far end crosstalk (ELFEXT). These improvements provide a higher signal-to-noise ratio, allowing higher reliability for current applications and higher data rates for future applications.

Will category 6 supersede category 5e?

Yes, analyst predictions and independent polls indicate that 80 to 90 percent of all new installations will be cabled with category 6. The fact that category 6 link and channel requirements are backward compatible to category 5e makes it very easy for customers to choose category 6 and supersede category 5e in their networks. Applications that worked over category 5e will work over category 6.

What does category 6 do for my current network vs. category 5e?

Because of its improved transmission performance and superior immunity from external noise, systems operating over category 6 cabling will have fewer errors vs. category 5e for current applications. This means fewer re-transmissions of lost or corrupted data packets under certain conditions, which translates into higher reliability for category 6 networks compared to category 5e networks.

When should I recommend or install category 6 vs. category 5e?

From a future proofing perspective, it is always better to install the best cabling available. This is because it is so difficult to replace cabling inside walls, in ducts under floors and other difficult places to access. The rationale is that cabling will last at least 10 years and will support at least four to five generations of equipment during that time. If future equipment running at much higher data rates requires better cabling, it will be very expensive to pull out category 5e cabling at a later time to install category 6 cabling. So why not do it for a premium of about 20 percent over category 5e on an installed basis?

What is the shortest link that the standard will allow?

There is no short length limit. The standard is intended to work for all lengths up to 100 meters. There is a guideline in ANSI/TIA/EIA-568-B.1 that says the consolidation point should be located at least 15 meters away from the telecommunications room to reduce the effect of connectors in close proximity. This recommendation is based upon worst-case performance calculations for short links with four mated connections in the channel.

What is a “tuned” system between cable and hardware? Is this really needed if product meets the standard?

The word “tuned” has been used by several manufacturers to describe products that deliver headroom to the category 6 standard. This is outside the scope of the category 6 standard. The component requirements of the standard have been carefully designed and analyzed to assure channel compliance and electrical/ mechanical interoperability.

What is impedance matching between cable and hardware? Is this really needed if product meets the standard?

The standard has no impedance matching requirements. These are addressed by having return loss requirements for cables, connectors, and patch cords.

Is there a use for category 6 in the residential market?

Yes, category 6 will be very effective in the residential market to support higher Internet access speeds while facilitating the more stringent Class B EMC requirements (see also the entire FCC Rules and Regulations, Title 47, Part 15). The better balance of category 6 will make it easier to meet the residential EMC requirements compared to category 5e cabling. Also, the growth of streaming media applications to the home will increase the need for higher data rates which are supported more easily and efficiently by category 6 cabling.

Why wouldn’t I skip category 6 and go straight to optical fiber?

You can certainly do that but will find that a fiber system is still very expensive. Ultimately, economics drive customer decisions, and today optical fiber together with optical transceivers is about twice as expensive as an equivalent system built using category 6 and associated copper electronics. Installation of copper cabling is more craft-friendly and can be accomplished with simple tools and techniques. Additionally, copper cabling supports the emerging Data Terminal Equipment (DTE) power standard under development by IEEE (802.3af).

What is meant by the term “Electrically Balanced”?

A simple open wire circuit consisting of two wires is considered to be a uniform, balanced transmission line. A uniform transmission line is one which has substantially identical electrical properties throughout its length, while a balanced transmission line is one whose two conductors are electrically alike and symmetrical with respect to ground and other nearby conductors.* “Electrically balanced” relates to the physical geometry and the dielectric properties of a twisted pair of conductors. If two insulated conductors are physically identical to one another in diameter, concentricity, dielectric material and are uniformly twisted with equal length of conductor, then the pair is electrically balanced with respect to its surroundings. The degree of electrical balance depends on the design and manufacturing process. Category 6 cable requires a greater degree of precision in the manufacturing process. Likewise, a category 6 connector requires a more balanced circuit design. For balanced transmission, an equal voltage of opposite polarity is applied on each conductor of a pair. The electromagnetic fields created by one conductor cancel out the electromagnetic fields created by its “balanced” companion conductor, leading to very little radiation from the balanced twisted pair transmission line. The same concept applies to external noise that is induced on each conductor of a twisted pair. A noise signal from an external source, such as radiation from a radio transmitter antenna generates an equal voltage of the same polarity, or “common mode voltage,” on each conductor of a pair. The difference in voltage between conductors of a pair from this radiated signal, the “differential voltage,” is effectively zero. Since the desired signal on the pair is the differential signal, the interference does not affect balanced transmission. The degree of electrical balance is determined by measuring the “differential voltage” and comparing it to the “common mode voltage” expressed in decibels (dB). This measurement is called Longitudinal Conversion Loss “LCL” in the Category 6 standard. * The ABC’s of the telephone Vol. 7

Buildings as Networks: Danger, Opportunity, and Guiding Principles for Energy Efficiency

Wednesday, April 7th, 2010

The coming twenty years will see a dramatic transformation in the patterns of energy consumption in buildings. Each year an increasing proportion of both devices and end uses in buildings will be influenced or dominated by controls that are defined by digital networks. Some of these networks will be established specifically to save energy but more often the controls and networks will be installed for other reasons, and can just as easily increase rather than reduce consumption. The efficiency community needs to be a lead actor in defining these networks’ creation and evolution, to assure that efficiency is a primary goal in their design and deployment. The alternative is to forever try to tame the energy consumption of networked products and technologies after they have been designed and installed.

The past twenty years of increasing networking of electronics shows the danger of a lack of attention to energy minimisation. Apart from niche wireless devices, energy has not often been a concern of the electronics industry in the myriad ways that devices are networked with each other. Consumption of electronics has risen dramatically in this time, partly due to increases in the stock of devices and services delivered, but a significant amount is wasted from lack of considering energy in network design. We should expect a similar outcome for other energy end uses.

Engaging with the industries that create the products — and the standards they will rely on to operate — will require significant investment by the efficiency community. However, in most cases there will be no incremental cost for manufacturing or deploying the more efficient products. Furthermore, there are likely to be few available projects for industrialized countries that rival efficient networks as an energy efficiency resource in terms of size and cost-effectiveness. The most effective and least costly time to address this issue is now.

The electricity delivery system is a vast and extraordinarily complex network — one we have had for over a century. Information networks are also not new; for example the telegraph network arose over 150 years ago. Traditional light switches and thermostats are very simple network examples. The telephone system is also quite old, though originally — and still partially — analogue rather than digital. Computer networks emerged as entirely digital from the beginning. Consumer electronic devices have long been networked — until recently almost entirely through analog connections, though they are now undergoing a rapid shift to digital.

The digital nature of current network developments is the key to their power and potential for energy efficiency. Computer networks, in particular the Internet, were not designed with energy use or efficiency in mind. The number of network nodes was small, and consequently, so was the aggregate direct consumption of network hardware; also, the fact of being networked did not change the consumption of devices on the network. So, the lack of attention to energy use was completely understandable. When power management was introduced into personal computers, network connectivity was not considered in its design; it was simply lost when going to sleep. When connectivity was acknowledged, with the introduction of “Wake On LAN”, the energy efficiency community was not involved (while Wake On LAN “works”, it is not widely used and so saves only a modest amount of energy). Meanwhile, most energy used by desktop PCs occurs when no one is present. There is an enormous infrastructure of hardware, software, protocols, applications, users, expectations, and the like which do not support energy efficiency of networked PCs in allowing them to go to sleep without compromising a basic — and often desired — capability to stay connected to the network. Fixing this problem after the fact is possible, but much more difficult and expensive than doing so when the network technologies were originally developed.
For consumer electronics (CE), people have long been accustomed to powering on and off televisions and devices connected to them with remote controls, and manually with power buttons. For devices other than the TV display, this is an annoyance (if consumers are even aware that other devices are on), with the result that devices are often left powered on during times of non-use. As CE devices become cheaper, and can be more easily networked with others (that may be in different rooms), the likelihood of devices being on when not in active use is rising.
TV set-top boxes are typically on continuously, to provide connectivity both upstream and downstream. As with computers, those concerned with energy use and efficiency have not been involved in developing the standards for inter-device control.
Manufacturers who do so are more focused on simply making things work at all, content protection, and with features which appeal to consumers — rather than to any focused effort on power controls (which is unlikely to increase sales).

In both cases, there has been additional confusion sown by poorly-designed and inconsistent user interfaces around power controls. Industry did not address this topic, but energy efficiency motivated work did (IEEE 1621) and has had success in rectifying this problem.
While there is little about networked electronics to indicate that energy efficiency will be a cost burden, the reality is that without specific attention to energy efficiency, it usually doesn’t happen. The only entities likely to bring this specific focus are those whose primary concern is efficiency; however these organizations struggles to deal with the highly technical nature of integrating power management into networks.

The Traditional End Uses Energy use in buildings is largely a matter of assembling and operating many individual and isolated components, with most of these are largely static. Products put into buildings are generally independent of each other in that the efficiency of one won’t affect the consumption of another (aside from internal loads affecting space conditioning). By contrast, digital networks make behavior of one product a factor in the energy consumption of others on the network, possibly driving it up or down.

The Default Future
As with the introduction of electronics, and most high-tech building controls to date, the forces driving building networking will be to improve the quality of the space for the benefit of the occupant, not saving energy. Other lessons from electronics will also likely apply in the absence of efforts to the contrary:

  • Promoters of specific technologies will ignore or resist opportunities for interoperability, as they try to gain maximum market share for their unique technologies.
  • Efficiency will be an afterthought, with other features driving the process (trade publications and trade shows provide overwhelming evidence for this).
  • Standards will be critical to facilitate some degree of interoperability, but aspects of these standards that could aid energy efficiency will generally be absent or ill-formed. Clear opportunities for harmonization across standards (e.g. in terminology) will not be taken.
  • User interfaces will be neglected, with individual manufacturers seeing this as an opportunity to differentiate their products, at the expense of users and of energy efficiency.
  • Little or no coordination will occur across domains. In electronics this manifests itself as the “IT” and “CE” domains, with different physical, application, and device infrastructure. For buildings, this is the end-use domains such as space conditioning, lighting, security, electronics, and others.

Achieving a Better Future
To arrive at a future in which digital networks optimally support energy efficiency, we should place ourselves into that long-run future — perhaps a ten year look ahead — and identify those features of network architecture not widely present that are most important for energy efficiency. We can then begin to describe key details of these features, how to develop them further as ideas, and how to market them to industry (many), standards organizations, and energy policy organizations.

Guiding Principles
Energy efficiency efforts around building-related networks need some “guiding principles” that can be used to evaluate existing and proposed network technologies. These need further development, but an initial list of Guiding Principles are as follows:

A. The existence of one device on a network should not cause another device to stay awake when it might otherwise go to sleep.

B. The network should be designed such that a legacy or incompatible device will not prevent the rest of the network from effectively using power management.

C. Devices should expose their own power state to the rest of the network and be able to report estimated or actual power use levels.

D. Product interfaces — for people or other products — should follow (international) standard principles and designs.

E. Products or devices that influence energy consumption should adhere to (international) standards for behavior and communication appropriate to their function.

F. Products and connections should have the ability to modulate energy use in response to the amount of service required.

G. Energy efficiency efforts should not favor any particular hardware — or even software — technology. All network technologies must be the target for efficiency efforts. Future buildings will include many different technologies; those in any particular building will be diverse, and always changing.

H. Harmonization of basic principles underlying efficient design for networked devices should cross all end uses and be global.

For the past two decades, we have seen an inexorable increase in the degree and sophistication of digital networking across electronics (both information technology and consumer electronics). This has greatly increased the services they provide, but spawned the creation of devices whose only function is to provide connectivity, increased the power levels drawn by these devices, and critically, driven up the time spent fully on for many of them. Electronic networks have been designed and implemented with little regard for energy consumption, and without the involvement of the energy efficiency community, so the resulting large increases in consumption are no surprise. Many aspects of set-top box energy consumption will apply to emerging networks in appliances and equipment.

Appliances and equipment in buildings are just beginning this transformation, a path which will lead to them becoming highly networked and controllable, across the major traditional end uses such as space conditioning, lighting, and security. As in the past, for the most part this will be done for reasons other than saving energy, such as greater comfort, control, security, productivity, and entertainment. A likely outcome is increased energy use, even aside from the energy needed to power the network itself.

This future is not inevitable. Action now can lay a strong foundation for devices to be interoperable with each other and with people in ways that facilitate maximum energy efficiency. This action will require careful attention from an efficiency perspective to many diverse standards that accomplish this interoperability — a few of these already exist but can be amended; many others are yet to be developed.

The efficiency community is not generally literate or involved in network standards development.

10-Gigabit Ethernet – Explained

Thursday, November 12th, 2009

10-Gigabit Ethernet.

10-Gigabit Ethernet (10-GBE), ratified in June 2002, is a logical extension of previous Ethernet versions. 10-GBE was designed to make the transition from LANs to Wide Area Networks (WANs) and Metropolitan Area Networks (MANs). It offers a cost-effective migration for high-performance and long-haul transmissions at up to 40 kilometres. Its most common application now is as a backbone for high-speed LANs, server farms, and campuses. It also enables you to connect geographically separated LANs to new MANs and WANs via dark fibre, dark wavelengths, or SONET/SDH networks.

10-GBE supports existing Ethernet technologies. It uses the same layers (MAC, PHY, and PMD), and the same frame sizes and formats. But the IEEE 802.3ae spec defines two sets of physical interfaces: LAN (LAN PHY) and WAN (WAN PHY). The most notable difference between 10-GBE and previous Ethernets is that 10-GBE operates in full-duplex only and specifies fibre optic media. The chart (below) notes the differences between Gigabit and 10-Gigabit Ethernet.

At a glance—Gigabit vs. 10-Gigabit Ethernet


  • CSMA/CD + full-duplex
  • Leveraged Fibre Channel PMDs
  • Reused 8B/10B coding
  • Optical/copper media
  • Support LAN to 5 km
  • Carrier extension

10-Gigabit Ethernet

  • Full-duplex only
  • New optical PMDs
  • New coding scheme 64B/66B
  • Optical (developing copper)
  • Support LAN to 40 km
  • Throttle MAC speed for WAN
  • Use SONET/SDH as Layer 1 transport

The alphabetical coding for 10-GBE is as follows:
S = 850 nm
L = 1310 nm
E = 1550 nm
X = 8B/10B signal encoding
R = 66B encoding
W = WIS interface (for use with SONET).

Distance Wavelength Cable
10GBASE-SR 300 m 850 nm Multimode
10GBASE-SW 300 m 850 nm Multimode
10GBASE-LR 10 km 1310 nm Single-Mode
10GBASE-LW 10 km 1310 nm Single-Mode
10GBASE-LX4 Multimode 300 m,
Single-Mode 10 km
Multimode 1310 nm,
Single-Mode WWDM
Multimode or
10GBASE-ER 40 km 1550 nm Single-Mode
10GBASE-EW 40 km 550 nm Single-Mode
10GBASE-CX4* 15 m — 4 x Twinax
10GBASE-T* 25–100 m — Twisted Pair
* Proposed for copper.

IPv6 Addressing Overview – Explained

Thursday, November 12th, 2009

IPv6 Addressing Overview.

Probably most of us have heard of IPv6 by now. And probably most if us know why IPv6 has come about. The continued growth of the internet or rather IP ready devices that, if needed, can connect to the internet has meant that the number of available standard IPv4 addresses is quickly running out. In fact today complex routing protocols are employed to keep the Internet working as a result.

There are a number of enhancements coming with IPv6. Perhaps the most obvious is the addressing standard, which has been changed dramatically in order to provide a vast number of IP addresses something in the order of 3×10 38 . This should see us all well into the future, even if every single device with electronics in has an IP address even your toaster.

Most of us recognise an IPv4 address, commonly written as dot separated decimal, for example These addresses are 32 bits long or 4 octets (bytes). Consider for a moment that an IPv6 address is 128 bits, or 16 octets. Now imagine having to remember as your station IP address, and now add to that a gateway address. It soon becomes obvious that you will need a lot of note paper and pens, not to mention the problems with typing errors.

With IPv6 the problem is simplified to some extent. Firstly instead of using decimal as we do today for IPv4 hexadecimal is used in the same way as MAC addresses. The second is to compress the address to remove some zeros. So an IPv6 address in long form could look like this 3ADF:1B4C:0000:0000:0000:0045:2CD2:EFA1. Now since a typical address like this might have a number of zeros this address can be displayed in short form notation, and becomes 3ADF:1B4C::45:2CD2:EFA1. Notice also that the address uses : rather than . as the separator.

The convention here is that leading zeros within the 4 digit groups can be dropped; you will notice that 0045 in the long address becomes simply 45 in the short version. Also a group of consecutive 16 bit numbers with the value of zero can be replaced with a double colon ::. It is only possible to replace one null string with the double colon, which can then be filled out to retrieve the long form address. If there are two null strings, only one can be compressed like this because if both were compressed it wouldn’t be possible to determine how long each one was so you’d end up with an ambiguous address.

Finally there is a slightly modified form of the IPv6 address for use when it’s desirable to express an IPv4 address in IPv6 format. To save having to convert constantly between base 10 and base 16 and to avoid conversion errors this convention uses the original dot separated decimal notation for the last 32 bits of the address, so the original IPv4 address of in IPv6 long format would be 0000:0000:0000:0000:0000:0000: which compresses into the short form as :: Despite the fact that the address space in IPv6 has been quadrupled the old IP number can still be expressed unambiguously in the new format with only 2 additional characters.

IP Addresses – Explained

Thursday, November 12th, 2009

IP Addresses.

IP (Internet Protocol) addresses are numbers that identify Internet hosts. They provide universal addressing across all the networks of the Internet.

IP addresses are placed in the IP packet header and are used to route packets to their destinations. An IP address is a 32-bit value split into four 8-bit pieces (octets) that are separated by dots. An example of an IP address is Each of the 4 numbers within the IP address can be between 1 and 255.

IP addresses are prefix based. The initial prefixes of the IP address can be used for generalised routing decisions. For example, the first 16 bits of an address might identify a corporation, the next 4 bits may identify a branch of that corporation, the following 6 bits may identify a particular LAN in that corporate branch, and the entire 32-bit address might identify a specific host within that LAN.

To simplify packet routing, Internet addresses are divided into five classes: Class A, Class B, Class C, Class D, and Class E. Very large corporations and entities receive Class A addresses, mid-sized companies and universities usually have Class B addresses, and most smaller companies and ISPs have Class C addresses. Class D is a multicast address and Class E is reserved.

Class A addresses are given to large organisations such as major universities and very large corporations. Class A addresses begin with a number between 1 and 126 (127 is reserved) in the first octet, leaving the 3 other octets open to split into local addresses. Although there are only 126 Class A codes, there are more than 16 million individual IP addresses within each Class A. Class B addresses are claimed by mid-sized companies, universities, and other entities that need thousands of IP addresses. Class B IP addresses begin with numbers between 128 and 191 in the first octet and have numbers from 1 through 255 in the second octet, leaving the last 2 octets open to denote local addresses. There are 16,384 Class B addresses with 65,536 individual IP addresses each.

Class C addresses —the most common—are used by most companies and ISPs. A Class C address has a number from 192 through 223 in the first octet and a number from 1 through 255 in the second and third octets, leaving only the fourth octet free for local addresses. There are more than two million Class C addresses and each contains 255 IP addresses.

Subnetting enables a network administrator to further divide the host part of the address into two or more subnets to make them easier to manage. A filter called a subnet mask is used to determine the subnet to which an IP address belongs.

Because IP addresses are difficult to remember, many also have text equivalents such as These text-based addresses are called domain names. A database program called Domain Name Service (DNS) keeps track of the names and translates them into their numeric equivalents.

The Internet is expected to outgrow the number of available IP addresses eventually. A new system of IP addresses called IPv6 has been designed to extend the capacity of the Internet. To date, the uptake of IPv6 has been limited. Most people are still using IPv4 and NAT (Network Address Translation) which allows multiple devices to connect to the Internet using only one ‘real’ IP address.

First Octet Second Octet Third Octet Fourth Octet

Convergence Solutions – Explained

Thursday, November 12th, 2009

Convergence Solutions.

There is a lot of discussion about the merits of Voice over IP—or IP Telephony—at a technical level, but less about the business issues associated with a converged solution.

Black Box is a major supplier of infrastructure systems and is well placed to help you decide what is the best option for your business.

Black Box has experience from the largest to the most basic network. Whether this is the infrastructure for the latest generation network for a mobile operator or Motorola, or a budget network for a small business or even a complete voice installation for one of America’s biggest retailers, we have a solution.

Convergence means having one common system to carry all forms of information (namely voice, data, video, etc.). Traditionally, each service has had its own network and associated technology. Voice utilised expensive PBX or exchange equipment with dedicated wiring to telephones. Data used a completely different set of cables and equipment, most commonly based on Ethernet. With technological advances and the growth of the Internet though, it is now possible to have one network carrying all the services.

So is it right for you, either now or for the future? Black Box can help you decide, plan your network and install the system; however, the solution will depend on a wide variety of factors. What you have now will influence the path you take.

It is a fact that a converged solution can save you money. The real savings, costs and benefits, however, are far less clear because the choices will be dictated by the need to move seamlessly to new systems without interruption of service or loss of quality. Depending on the age, facilities and performance of the existing voice equipment, you may want to consider conversion rather than replacement. Simply use a converter or special interface to allow some or all of the existing connections to utilise low-cost IP call facilities. Having one system simplifies management and support, reducing costs. Modern systems can be accessed and managed remotely, allowing the number of specialist staff on site to be minimised.

Any new system needs to integrate and offer compatibility both with the internal and external networks. There are international standards for IP telephony, so selecting the right equipment becomes easier. The two major techniques used are H323 and SIP. H323 is more established, especially when ISDN is used for the communication network. SIP is newer, but it is supported by major players such as Cisco and Microsoft®. Both will co-exist for some while, and work is under way to allow interoperation between them.

Data packets can be delayed and, if necessary, retransmitted in the event of errors or problems. This is not a possibility with voice communication. Overlaying voice traffic on an old data network can create bottlenecks and delays, resulting in poor quality voice and slow data traffic. The network design needs to be checked and, if required, upgraded to support the extra traffic and ensure priority to the most critical types of traffic. If video is part of the system, this will place even greater demands on capacity.

Starting with a clean sheet and scrapping all existing systems is an ideal unavailable in all but a few special cases. The need to provide paths to migrate from old legacy systems and procedures toward convergence is therefore paramount. Black Box has long been acknowledged as a specialist provider of products and technology that solves the problem of mismatch. Whether it’s a special cable, an interface converter or a complete change of protocol, Black Box is renowned for coming up with a solution.

If all communication is taking place over a single converged network, then this could be considered a single point of failure. In reality, networks can be designed to be fully resilient; however, this has to be planned carefully and tested to make sure that everything continues to work smoothly in the event of problems. Having a converged network will make it considerably easier to establish disaster recovery procedures for a major critical event, such as a fire destroying an entire building. The requirements of new legislation on corporate governance also has an impact on communications, security and integrity of networks. Consolidating all systems under one management enables better control and eases problems of compliance with the requirements of legislation such as Sarbains Oxley, etc.

Any system that involves access from the Internet has to be secure. Users have to be sure that they will not be overheard and that confidential information is not available to the outside world. Building in firewalls and other security procedures should be part of the design process, not a late fix to a leaky system. Support should be rapidly available when required and delivered in the most suitable and flexible manner to match the needs of the situation. Black Box provides on-site and on-line Technical Support on a local and global basis 24 hours a day, 365 days a year.

Building a small network – Explained

Thursday, November 12th, 2009

Building a small network.

Building a small network.
Never fear-it’s easy to build a twisted-pair Ethernet network. In fact, it’s the simplest and most inexpensive network you can build, and it’s worth installing for even just two or three PCs. Your small network doesn’t have to be slow either-most of today’s Ethernet devices support 100-Mbps Ethernet as well as legacy 10baseT. These dual-speed devices sense and adjust automatically to the speed of connected devices.

Build a basic Ethernet network.
The most basic Ethernet network uses an Ethernet switch to enable two or more PCs to communicate directly with each other. This very simple network, which operates without a network server is called a peer-to-peer network. See the diagram below.

All you need are an Ethernet adaptor card for each connected PC, an Ethernet switch, and some CAT5e unshielded twisted-pair (UTP) cable. If your PCs have built-in Ethernet like many of today’s PCs do, you don’t even need Ethernet adaptors.

To build your network, connect the Ethernet port on each PC to a port on your Ethernet switch using the CAT5e cable. Snap-in, modular connectors make connecting the cable to the PCs and the switch as simple as plugging in your phone. If you need more ports, just connect another switch to the first.

And don’t worry about software—if you have Windows® 95 or later, you have all the software you need for a small peer-to-peer network.

Add a print server.
Even a very small network can benefit from the convenience of a print server, a specialised network device that enables network users to share one or more printers. It accepts print jobs from users and manages these jobs on each printer. See the diagram below.

Typically, a print server is a freestanding device that’s connected between the network and the printer. A freestanding print server is very easy to install—just connect it to your Ethernet switch using CAT5e cable, then connect the printer using a parallel printer cable.

A print server for a small network may also be built into another device, such as a switch or a broadband router.

Your print server will probably come with software utilities to install on your PC, and you’ll need to do some configuration to set it up. But, once installed, it’s virtually transparent to network users.

Connect your network to the Internet.
A remote access router enables your entire network to share a single Internet connection. The small remote access routers used in small and home office networks are usually referred to as broadband routers because they connect your network to broadband DSL or cable modem Internet services. See the diagram below.

A remote access router enables two or more computers to share an Internet connection by using a technology called Network Address Translation (NAT), which enables all the computers on your network to share a single IP address.

Although the primary reason to install a remote access router is the convenience of having all network users share an Internet connection, a router also helps keep your system safe from hackers. NAT masks your true IP address, providing firewall protection between your network and the Internet.

You install the remote access router between your Ethernet switch and your DSL or cable modem. The DSL or cable modem is usually provided by your Internet service provider and has an Ethernet port, which may be a regular LAN port that can be connected by straight-pinned CAT5e cable or may be a WAN port that requires a special cross-pinned CAT5e cable for connection to the Ethernet switch.

Remote access routers normally require extensive setup and configuration but, once installed, operate transparently.

Broadband routers for small networks also often feature a built-in Ethernet switch and print server. This means you only need the broadband router plus some cable to turn a few unconnected PCs into a secure, multifeatured Ethernet network.

Layer 3 switching – explained

Thursday, November 12th, 2009

Layer 3 switching.

In the last decade, network topologies have typically featured routers along with hubs or switches. The hub or switch acts as a central wiring point for LAN segments while the router takes care of higher-level functions such as protocol translation, traffic between LAN segments, and wide-area access.

Layer 3 switching, which combines Layer 2 switching and Layer 3 IP routing, provides a more cost-effective way of configuring LANs by incorporating switching and routing into one device. Although a traditional Layer 2 switch simply sends data along without examining it, a Layer 3 switch incorporates some features of a router in that it examines data packets before sending them on their way. The integration of switching and routing in a Layer 3 switch takes advantage of the speed of a switch and the intelligence of a router in one economical package.

There are two basic types of Layer 3 switching:

  • Packet-by-packet Layer 3 (PPL3)
    PPL3 switches are technically routers in that they examine all packets before forwarding them to their destinations. They achieve top speed by running protocols such as OSPF (Open Shortest Path First) and by using cache routing tables. Because these switches understand and take advantage of network topology, they can blow the doors off traditional routers with speeds of more than 7,000,000 (that’s seven million) packets per second.
  • Cut-through Layer 3
    This method of Layer 3 switching relies on a shortcut for top speed. Cut-through Layer 3 switches, rather than examining every packet, examine only the first in each series to determine destination. Once the destination is known, the data flow is switched at Layer 2 to achieve high speeds.