IT for Radio Engineers

September 1, 2007


As engineers, part of your role is to find creative methods to achieve a certain result. You do this for the most part because there is not a commercially viable solution available, or maybe there is a solution but the budget doesn't allow it. For myself, and most of the engineers I knew back in the day, this was the stuff we lived for, making the impossible happen and actually having it work. Give us the roll of solder, a few components, wire, a metal box and voilà: a fix for a problem!

We have all been in the situation where you start working for a new station and come across the myriad of home-made devices that seem out of place, but obviously address a particular problem. We have always had a toolbox of tricks and solutions that could be whipped together. Today, many of the clunky boxes that served us so well in the past have been replaced by networked hardware and software systems, which, while providing a far more powerful and compact platform to fill the operational needs of a facility, now take the form of black-boxes and software files. (The tools you use no longer came from a metal cabinet.)

There are libraries of software designed to address many situations, most of it free or for a small charge. You can also write your own applications, but before you take on such a task, it might be helpful to have a better understanding of how data is transported in a networked environment, what to expect in the future, a more reliable way to connect over the network, and maybe throw in a few tips and tools along the way.

Net Tool Series II from Fluke

Ethernet testers, such as the Net Tool Series II from Fluke, are the modern handheld test tool in an IT-based facility.

In many ways, transporting digital audio around the station, or anywhere else for that matter, has been made much simpler through Ethernet networking. I will discuss a new protocol (IPv6) that plays an important role in the transport of streaming data and a relatively old protocol (IP Tunneling) that perhaps could be one of the most useful tools in terms of creating both reliable and secure IP connections between devices. But first, a review of IP.

IP review: The basics

Internet protocol (IP) is the reason we can transport any type of data file through a local network or even the Internet. The IP establishes the method by which data is packaged and identified on the network; Transport Control Protocol (TCP) defines the method used to transport the IP from one point to another. Hence the acronym TCP/IP that describes how data gets packaged and routed through a network. As with most protocols in networking, these (and other) protocols are based on a set of layers that define the physical connection, transport methods, data packaging and error correction aspects of the protocol. These layers each have a specific job and are designed to work with the other layers above or below it. Layering also gives us a great deal of flexibility to upgrade to new standards while preserving compatibility with other protocols.

The IP address is a unique 32-bit sequence divided into four separate four-byte numbers or octets separated by periods. Each four-byte number ranges from 0 to 255 (aaa.bbb.ccc.ddd).

An organization called Internic is responsible for assigning these addresses with the intention of making sure every terminal device on the Internet has a unique address. The available addresses are further categorized into classes based on the amount of users needing a contiguous address, i.e. a cable provider serving thousands of users needs a large block of addresses to assign to users. The three classes are:

Class A - Addresses assigned to the first octet “aaa” in the range of 0-126. For example, an address beginning with 110.xxx.xxx.xxx is considered a class A address. This class can accommodate 126 different networks with up to 16 million separate hosts.

Class B - Utilizes the first two octets for the network ID and the last two for the host ID. The first two octets will always have an address between 128.000.xxx.xxx and 191.255.xxx.xxx yielding 16 thousand possible networks, each with 16 thousand possible hosts.

Class C - Uses the first three octets for the network ID and the last for host ID. The first three octets with always have addresses between 224.000.000.xxx and 239.255.255.xxx yielding 2 million possible networks and up to 254 hosts for each network.

Each class also has reserved addresses used only for establishing connections that do not go to the Internet. For example, the network in your office or home probably uses an address in the range 192.168.0.0 and 192.168.255.255. This is the range reserved for local connections in class C networks. If the local network has more than 255 users, then there are also reserved ranges for class A and B that accommodate larger amounts of users.

However, in most large company networks the smaller groups are usually divided into individual subnetworks that are grouped into a single larger network.

To reduce the need to assign every device on a network its own IP address, most routers use a technique called Network Address Translation (NAT). Let's say you send an e-mail message from your company computer to a computer in another company's office: from an IP perspective, the data leaves your PC on a local IP address. When it goes through the router, the data is sent on the IP address assigned to the specific office. When it arrives at the destination office, the router reassigns the data to the IP address of the intended recipient. NAT is the protocol that gives routers the ability to recognize IP data packets intended for a specific device and route them properly.



IPv6 is coming

Based on current usage levels, it is anticipated that the current blocks of available addresses under the current system, also called IPv4, will be exhausted in approximately three years. To head this off, a new proposal is in the final stages of approval. Known as IPv6, it will probably provide the most benefit to broadcasters and multicasters, due to enhancements to how IP handles audio and video streams.

The major improvement IPv6 provides is its ability to serve 3.4 × 1038 addresses using 128-bit addressing contrasted to the measly 4 billion currently supported by IPv4 using 32-bit addressing. In theory, IPv6 will provide enough addresses for every person on the planet and still have room to grow almost infinitely. Routing IPv6 data will be made simpler due to the elimination of the NAT protocol since each device will have a unique address and address translation will no longer be necessary. The virtually limitless availability of unique IP addresses opens up the possibility of a wide variety of devices, not just PCs or printers. It is expected that even common items, such as appliances or automobiles, can now have their own IP addresses.

There are interesting possibilities for our industry as well. Imagine transmitters, processing equipment and other peripheral equipment all having IP addresses that could be interconnected from anywhere in the world. Take this a step further and consider a transmitter that takes all of its control and audio streaming information through an Ethernet connection connected to a remote control and audio processor in a different state. Do you see where this could go?

IPv6 addresses are 128 bits long. There are 64 bits for the network address and 64 bits for the host address. The host address is derived from the unique MAC address given to the network interface device. Optionally, they can also be generated sequentially. In practice, the IPv6 address is written in a hexadecimal formal consisting of eight groups with four hex digits.

IPv6 is ideal for streaming because the protocol defines three specific types of addresses:

  • Unicast — Basically point-to-point, typical of most applications.

  • Multicast — One-to-many recipients.

  • Anycast — A variation of multicast, only delivered to a single node than routed to other nodes until it reaches the recipient(s).

A detailed discussion of IPv6 would fill volumes, but you need to be aware that many governments and large organizations are starting the transition process. Newer equipment might be software upgradeable but some will need to be changed. Start identifying compatibility issues now. Many manufacturers have information on their websites. Be aware that only Windows Vista and Mac OS X 10.3 and above have native support for IPv6.

IP Tunneling

When it comes to establishing a hard connection between one or more points, IP tunneling is the ultimate tool. Why should you care about tunneling? Consider for example that you are trying to set up a remote broadcast in another county, state or country. Using a dedicated IP address might create problems with latency or possible loss of connection. Utilizing an IP tunnel would establish a dedicated virtual connection between the remote user and host, similar to any local device and typically with improved performance.

The concept of tunneling is based on a process that packages the primary IP packet inside another packet. The purpose of the outer packet is to create a virtual physical connection between two networks that encapsulates the real data packet.

Like everything in the IT world, there is a protocol (or set of protocols) that make implementation possible. There are two primary protocols that can be used to create an IP tunnel:

Point-to-point tunneling protocol (PPTP) may seem familiar from back in the days of dial-up connections. Originally developed by Cisco and later licensed to Microsoft, this became a standard communications protocol to be included standard with later versions of MS Windows and as such, became a popular protocol with dial-up hosting providers.

PPTP requires two separate connections — one connection maintains the data path using another protocol called Generic Routing Encapsulation (GRE). GRE manages the encapsulation process and subsequently strips the encapsulation at the other end. The second connection is used to initiate and maintain the GRE session.

Layer 2 Tunneling Protocol (L2TP) is a newer and more feature-packed version created from a combination of PPTP and another old protocol called Layer 2 Forwarding (L2F). The use of Layer 2 here is deceiving in that it actually operates at the application layer (Layer 5). It works by encapsulating the packet, payload (original data) and header within a UDP (Universal Data Protocol) datagram. Datagrams are basically a package containing a short message. For example, a series of datagrams form the basis for streaming audio and video technologies. The current version is L2TPv3 which provides improved performance and compatibly with other transport services.

One more protocol you should know is called IPsec, which is short for IP security. This is typically used in conjunction with a tunneling protocol, particularly L2TP as a means to provide secure authentication and encryption services. When the two protocols are used together the protocol is called L2TP/IPsec.

In practice, the tunneling process begins by establishing the virtual connection between the client and host systems. The implementation of creating the VPN is typically done through software and/or hardware applications. A popular example of IP tunneling is the Virtual Private Network (VPN), which provides remote users the ability to gain access to their internal network (Intranet) and network resources (printers, etc.) as if they were connected in their offices. The remote user initiates a typical VPN session with a software client that requires you log in with your username, password, etc. The host end of the VPN manages the task of accepting requests to establish the connection and verifying the user login and security information. Once the user information has been accepted the tunneled connected is established and remains connected until either the user (or administrator) chooses to stop it or the host system has provisions to drop the connection when not in use for a preset amount of time.

While tunneling provides a better experience for the remote user, it also offers a much higher level of security for the host network since the connection can't be established without going through some level of security check.

Creating a tunneled connection is easy or cheap (even free) with the use of software applications easily downloaded from the Web. Some of the more popular ones include Zebedee, Nest and Barracuda. A Web search for “IP tunneling software” will direct you to these and a host of other solutions you can use to create your own custom IP tunnel. Most of these also have detailed user guides to help you achieve your goal and make a handy addition to your new toolbox.


McNamara is president of Applied Wireless, Cape Coral, FL.


Want to read more stories like this?
Get our Free Newsletter Here!

Comments