You are in the process of developing a new line of xDSL modems. The modem has gone from the drawing board to the prototype stage. Before you put the new modem into production, however - before you commit to a massive investment of man-hours and parts - it would help to know whether the unit performs as it was designed to or whether it still has some kinks to be eliminated. You will want to know just how well the modem performs on the local loop.
Consider another scenario. You have just purchased a new DSL modem. It has been touted as the most advanced of its kind on the market. Possibly it represents a new flavor of xDSL modem entirely. In any case, you take a look at the new modem's spec sheet and are skeptical as to whether it can perform as well as its manufacturers claim. You want to put it through its paces and see how it measures up.
In either case, there are three methods by which you can test your DSL modem's performance. All three methods of testing boil down to an examination of the BER performance of the modem over the local loop at a specific distance and in a specific interference environment.
The first method is to use the local loop itself, set up with your local telco. This is a real world test, throwing the modem into the deep end and watching whether it will sink or swim. Unfortunately, it does not provide the most reliable measure of performance. The "live" local loop you use for the test is not necessarily representative of the parameters within which the modem must function "in the field." You will have no control over, and in some cases no knowledge of, the loop length. The level and type of interference on the line will also be outside your capability to gauge. These operating factors apart, a local loop for testing is physically awkward in a manufacturing environment.
A second method is to test your modem on a coil of twisted pair cable, with the cable mimicking a local loop. This testing method has the advantage of being inexpensive. It allows you to simulate the attenuation and delay of the local loop, further allowing you to measure accurately the loop length. What it cannot do is simulate interference. It ignores the effects of crosstalk. And it too is physically awkward in the manufacturing environment. Coiled cable is bulky and difficult to fit on a laboratory bench. Multiple coils can take up considerable workspace.
The third method is to use test equipment specifically designed to measure the capabilities of your DSL modem, a tool known as the "DSL simulator." As its name implies, the simulator can do what the local loop and cable coil tests cannot: accurately represent attenuation, delay and multiple types of interference over a variety of loop lengths. Like the communications media they are meant to mimic, simulators come in both digital and analog. The digital simulators rely on a format known as Digital Signal Processing (DSP). The DSP-based simulator performs its tests via chips and microprocessors, running programs that provide a digital filter approximation.
DSP-based simulation, unfortunately, shares the problem of space constraint with local loop and wire coil testing. The current generation of digital simulators, while suitable for laboratory use, are too large for bench-level use. Furthermore, DSP-based simulators are costly, running into the tens of thousands of dollars, a prohibitive expense in a factory space demanding multiple test devices. As such, the DSP-based simulator tends to be impractical for use in the manufacturing environment.
This brings us to the analog alternative. Analog DSL simulation relies on lumped, linear, bilateral, passive electrical filters composed of resistors, capacitors and inductors. These tried-and-true electrical components provide an analog approximation of loop conditions. Ironically, analog DSL simulation is much cheaper and more compact than digital DSL simulation.
The sleek design, reliability and low cost of the Telebyte Model 459-A makes it the right choice for the production test environment. In addition, loop lengths are programmable from 4 kft to 22 kft, in 2-kft increments (plus 26 kft). The Model 459-A is ideally suited to test DMT DLSAM channels and modems. The Model 459-A simulates 26 AWG PIC and offers the bandwidth requirements for xDSL technologies such as ADSL, G.Lite, HDSL, HDSL2, G.SHDSL (Annex 1) and SDSL.
The Telebyte Model 458 Multi-Channel Local Loop Simulator provides a wide variety of configurations through use of up to 16 plug-in Line Modules. It is ideally suited for testing DSL modems and other bandwidth-compliant telecom devices in a high-volume production line environment. The 458 Control Module interfaces with a controlling PC or terminal via IEEE-488 or RS-232 to control loop-length settings. A user-friendly interface or Common user-command language may be used for control. The Model 458 simulates 26 AWG PIC and offers the bandwidth requirements for xDSL technologies such as ADSL, G.Lite, HDSL, HDSL2, G.SHDSL (Annex 1) and SDSL.
Additional Features:
Figure 4-7: A wide variety of Model 458 Line Modules plug into the 458 and 458-2SL card cages.
Line Module | Technology | Cable | Lengths/Increments | Bandwidth |
458-LM-20 | ADSL, ADSL2, ADSL Lite ISDN, G.lite | 26 AWG PIC Insulation | 0-20.5 kft/500 ft | DC to 1.5 MHz |
458-LM-E20 | ADSL, ISDN, SHDSL/G.SHDSL (Annex A) | 0.4 mm PE Insulation | 0-6.15 km/150 m | DC to 1.5 MHz |
458-LM-HD | ADSL2++ | 26 AWG PIC |
Channel 1: 0 - 31,750 ft/250 ft Channel 2-8: 0-30,000 ft/2,000 ft |
DC to 4.5 MHz |
458-LM-HDE | ADSL2++ | 0.4 mm PE |
Channel 1: 0 - 9,450/150 m Channel 2-8: 0-9,000/600 m |
DC to 4.5 MHz |
458-LM-HDJ | ADSL2++ | 0.4 mm Paper | 0 km to 7.5 km/500-m | DC to 4.0 MHz |
458-LM-A2-18 | ADSL, ADSL2, ADSL2+, ADSL2++, VDSL | 24/26 AWG | 0-16,000 ft/100 ft | DC to 18 MHz |
458-LM-A8-18 | ADSL, ADSL2, ADSL2+, ADSL2++, VDSL | 26 AWG PIC | 0-15,000 ft/1,000 ft | DC to 18 MHz |
Introducing the industry's first VDSL Local Loop Simulator. Ideally suited for the functional production testing of VDSL devices, this compact and reliable unit accurately simulates attenuation and impedance characteristics of 26 AWG and provides up to 24 channels in a 2-U high rack-mountable chassis. Loop lengths are programmable from 0 to 5,500 ft in 500-ft increments, bandwidth ranges from 100 kHz to 12 MHz and the 460-V can be remotely controlled via RS-232.
Other Products Related to DSL Simulation:
Figure 4-9: Model 4801 Universal xDSL Noise/Interference Simulator (8-slot version shown)
The innovative Model 4801 Universal xDSL
Noise/Interference Simulator features the lowest noise floor
of any noise simulator on the market today. This
highly-advanced system allows manufacturers of today’s xDSL
technologies to test signaling devices such as transceivers
and DSLAMs with ease and accuracy. Our design uses unique
technology to achieve real-life noise simulation, allowing
for accurate and repeatable testing. Reduce clutter with
this all-in-one unit that includes a powerful built-in PC,
eliminating the need to add a PC to the test configuration.
The Universal xDSL Noise/Interference Simulator seamlessly
integrates with our outstanding line of wire line (local
loop) simulators for the ultimate in affordability, accuracy
and ease of use. Modular design allows for a system that can
be expanded as needed.
Figure 4-10: Model 4101-J Japanese ADSL+ Test Loop Simulator
The 4101-J supports TCM loops 2 and 5 and is fully compliant with NTT’s proposed amendment to ITU-T for ADSL+. The system also meets NTT Laboratory’s expanded requirements for attenuation accuracy and loop length increments. This device will operate from DC to 4 MHz. The 4101-J can be controlled by a PC or terminal via IEEE-488 or RS-232 to control loop-length settings. Control can be accomplished by using the supplied GUI or through the use of scripts.