Bernard and Agnes

After getting a high speed broadband Internet connection at home, many people want to share it between multiple computers. Most cable and DSL modems (or more generically, "broadband modems") support the connection of only a single computer at a time, so an additional piece of equipment is necessary to share the connection. We’ll examine three common home network configurations and their relative strengths and weaknesses.

Option 1: Broadband Router

The most common configuration involves an Broadband Router such as the Linksys BEFSR41 or the Microsoft MN-100. Other popular makers of home networking equipment include NetGear and D-Link. Installation is fairly straightforward in most cases. The router is plugged into the modem. All computers are then plugged into a network port of the router. Most of the time the router will automatically configure itself to use the Internet connection and allow it to be shared amongst the attached computers. In rare cases, the Internet provider will give you a 4-part numerical IP address to type in. The router will have instructions on where to type in this address.

This option is easy to install and pretty inexpensive — usually around $50 for the router and necessary network cables. The option tends to be relatively secure due to the fact that all of these routers provide NAT and almost all have built-in firewalls. The purpose and technology behind NAT and firewalls will be explained in detail in the future on this site, but in short, both technologies make it harder for hackers to get at machines inside our Local Area Network (LAN). This option is somewhat constrained in the number of computers that can be attached at one time, because of the limited number of network ports on the router (typically 4 or 8); however, it is possible to accommodate additional computers by using an inexpensive hub (about $20):

Each computer inside the LAN can communicate with each other at 100 mbps. The connection to the Internet is the slowest part of the whole network with typical speeds of 750 kbps.

Option 2: Internet Connection Sharing

In this option, we avoid the purchase of an Internet router. A computer running Windows 2000, Windows XP or Linux can easily play the same role. Microsoft calls the feature Internet Connection Sharing (ICS). To set up this option, the computer running ICS needs two network cards. One will connect to the modem. The other will be connected to a hub, which is used by all the computers sharing the connection. ICS is enabled for the connection to the cable modem, and the additional computers will all automatically configure themselves to use the ICS computer as the Internet router. If we are sharing the connection with only one other computer, we do not need the hub. Instead, we can connect the two computers directly with a special network cable called a crossover cable.

This option is a tiny bit harder to set up than Option 1 because of the need to install a second network card into the ICS computer, and then enable ICS in Windows. This option is cheaper because we need only a second network card ($5) and probably a hub ($20). The configuration is marginally less secure. While the ICS computer uses NAT and a software firewall to protect the LAN, it is directly accessible to hackers who attempt to intrude. The direct accessibility makes the ICS computer more vulnerable to intrusion than any computer in Option 1. The speed of the network will be identical to Option 1, and the expandability is limited only by the number of ports on the hub. The main disadvantage of this option is that the ICS computer must be on for any of the other computers to use the Internet connection.

Option 3: Wireless Networking

In the first two options, all of the computers had to be close enough to the modem for a network cable to reach. Network cables can be long (about 25 meters) and they can be placed inside walls, but there is a big advantage to setting up a wireless network instead. Computers can be placed anywhere in the house and one can pick up and move a laptop at any time. Many PDAs can also use a wireless network connection.

Instead of buying the router in Option 1, we need to buy a Wireless Access Point (WAP) such as the Linksys BEFW11S4 or the Microsoft MN-500 that also supports the capabilities of a router (not all WAPs do). The WAP is connected to the modem, and then each computer uses its wireless network card to connect through the WAP to the outside world.

This option is easy to set up. It costs more than the other two options because of the WAP ($100) and the wireless network cards for each computer ($50 each). The WAP provides NAT and a firewall, just as in Option 1, but it suffers from a security flaw:it is not difficult for a stranger to use the wireless network by getting close enough to the WAP to make the connection. Even a password-protected wireless network is not difficult to connect to. The speed of the LAN is slower than the other two options (11 mbps), but since the LAN is still much faster than the connection to the outside world, the difference will probably not be noticeable. Newer WAPs that support the wireless network protocol 802.11g are faster (54 mbps) but more expensive. This option easily supports a large number of computers because we don’t need a network port for each computer. A consumer-level WAP will probably support 10-15 computers before the wireless network becomes sluggish. One last point to note is that wireless networks can become flakey from interference from other devices that use the same radio frequency such as 2.4 GHz wireless telephones. For this reason, some people choose to connect at least one computer to a network port on the WAP with a cable.

These three options are summarized here:

Security

One note that relates to security: in our options, we talked about how the router or ICS machine or WAP uses NAT and provides a firewall. If we don’t take further steps to secure our network, we end up with what is known in security circles as a LAN with a "hard crunchy exterior and soft chewy interior". It’s like having one lock on the outer-most door of a building and no doors at all inside the building. It’s a good practice to protect each individual computer in case an intruder gets into the network:

• Use antivirus software and keep its virus signature file up to date. The software should be configured to check all files for viruses.
• Regularly install security patches released by software vendors. Most known security flaws in software have corresponding patches that fix them. Microsoft makes it easy to find all Windows updates on their website http://windowsupdate.microsoft.com.
• Use a password with all user accounts and make a habit of changing the password regularly.
• Require a password for all shared folders. Hide the shared folders so they are not visible to someone browsing the network by giving them a name that ends with a $.
• Consider running a software firewall on every computer.
• Be very restrictive in how any firewall is configured — it should only allow network connections that are absolutely necessary.
• Don’t give administrative privileges to your normal user account. Instead, log in as the administrator only when you need to make a system configuration change.
• Shutdown any extraneous services in Windows. Each of these programs potentially have security holes. If they are not running, it’s less likely they can be used to compromise the security of a computer.
• Notice and investigate any unusual behavior on the computer.

Happy networking!

Pre-recorded music or software CDs have physical pits in the surface. A cd player or cd-rom drive reads the CD by shining a low power laser on the spinning disc and measuring the amount the laser reflects. A pit will reflect less light than a section with no pit. Recordable CDs also have sections that reflect more or less light. Instead of pits, though, these sections are created by burning a dye on the disc with a high power laser, which creates dark and light sections. On this page, we try to describe how a rewriteable CD works, since both pre-recorded and recordable CDs use very permanent mechanisms for writing the disc.

Essentially, rewriteable CDs need to have dark and light sections, but the process of creating the sections need to be reversible. One needs to be able to turn a dark spot into a light spot and vice versa. In addition, a rewriteable CD needs to be recordable using the same physical equipment as a normal recordable CD. The way it works is that CD-RW discs have a layer of alloy on them, instead of the dye that a CD-R has. The alloy has two states: a crystalline state and an amorphous state. These two states are analogous to sand and glass. Both are the same material, but they have different optical properties. On the CD-RW, the alloy’s crystalline state reflects light well, while the amorphous state does not. This alloy has the additional benefit that one can change the state of the alloy by heating it to different temperatures using different intensity lasers. A high power laser writes amorphous spots on the disc. An intermediate power laser writes crystalline spots. A CD-RW’s alloy can be re-written approximately 1,000 times, and normally should store its data for about 30 years.

Some people have floated a theory that there really are no CD-R’s sold these days. The theory says that it must be more economical to create only one kind of CD (the CD-RW), and simply label them as recordable or rewriteable. By making only one kind of CD, the producer can eliminate dual sets of manufacturing equipment. This theory probably does not hold water though. It really is cheaper to make both kinds of discs.

I bought my Toshiba e740 Pocket PC a little after I started flying for work. In addition to using it for standard Personal Information Management (PIM) applications such as tracking appointments and contacts, I also use it for entertainment. Some applications, such as playing music or video, can drain quite a bit of the battery during a six hour flight cross-country, and so I found myself looking for ways to keep the pocket pc charged while I was away from home. At first, I carried my AC power adapter with me to charge the pocket pc in the hotel. Later, I found another option which involves less bulk and more flexibility. The solution I found for keeping my pocket pc charged will probably work for any of a number of other devices out there, including other PDAs, cell phones, and digital cameras.

The AC adapter for Toshiba e740 is a 5 volt, 2 amp power supply. As it turns out, pretty much any 5 volt power source will work to charge the lithium ion battery in the pocket pc, as long as it does not supply too much current. Taking advantage of this fact, a number of companies have produced USB cables to charge a pocket pc. A USB port will supply 5 volts at 0.5 amps. Because the USB port supplies less current, the device will take longer to charge, but it still successfully charges. The USB cable that I bought was the miniSync from BoxWave. I like it because it automatically retracts, and it is very compact and light.

The USB adapter allows me to keep my pocket pc charged while I have my laptop with me, but my laptop isn’t always the most convenient thing to carry around. Luckily, various companies have produced other power sources with a USB port. Belkin produces a car adapter with a USB port and an AC adapter with a USB port. CoolerStar produces a 4 AA battery pack that has a USB port. These power sources give me a lot of flexibility on how to charge my pocket pc. Below, you can see a picture for these three power sources. I have no idea why CoolerStar displays their battery pack with a USB light plugged in.

By getting a USB charging cable for my pocket pc, I have a lot of options for charging my pocket pc. These options should be available to anyone who can find a USB charging cable for their device.