SETUP FOR THE EXPERIMENTS


For all the experiments you will be using Linux PCs in LWSN B148 or B158 that are connected to each other and the global IP Internet. TDR equipment is provided by the TA.
 


TIME DOMAIN REFLECTOMETER (TDR) MEASUREMENTS (60 pts)


Purpose of the Experiment

In this exercise, you will observe signal propagation in copper cables by experimenting with a measuring device known as a Time Domain Reflectometer (TDR) that can be used to detect a broken cable.
 

Overview

A TDR attaches to one end of a copper cable and can be used to estimate the length of a cable up to its break point. A similar approach also applies to fiber optic cables using an optical TDR. This is done by transmitting an electrical pulse of a suitable shape. Irregularities in the cable's conductance cause reflection of the electrical signal which is detected at the TDR. Assuming the propagation speed of the signal on the cable is known (depends on the physical properties of the specific cable used) and an accurate timer measures the time lag until the reflection arrives (i.e., round-trip time), the length of the unbroken cable segment may be estimated.
 

Procedure and Details (30 pts)

To see a TDR in action, you will be provided with a TDR and some cables. You will use the TDR to measure the length and related properties of the provided cables. Record the various measurements. Note: some cables are damaged, others are not (i.e., terminator absorbs signals to prevent reflection).

Submit a write-up of what experiments you carried out, the measurements you observed, and your interpretation of the measurements.
 

More Wired Copper Communication (30 pts)

Use the WWW to research what grades (i.e., quality) of copper cable there are and the approximate propagation speeds of electrical signals over these wires. Determine what kind of copper wires are commonly used for residential phone lines, home electrical lines, FastEthernet (100 Mbps) cables, home theater audio systems, and cable TVs. Are they just different in their thickness or is it more complicated? We know that data communication can be performed over phone lines and cable TV coax cables through DSL/cable modems. Research if a wired home data network can be built using home electrical lines and what advantages/disadvantages there may be in doing so. How is the above different from power-over-ethernet (POE), a technology we use in LWSN to connect network devices such as wireless access points? Would you recommend to your friend, who lives in the basement of her parent's house where WiFi reception is poor, to use power lines to get high-speed Internet connectivity? Submit a write-up of your findings and thoughts.
 



PING AND TRACEROUTE NETWORK UTILITIES (50 pts)

Ping (20 pts):

The ping (Packet Internet Groper) program may be used to probe a networked device (connected through a point-to-point link, multi-access link, or internetwork) that speaks IP. The ping utility sends a special IP probe packet (called ICMP echo request) to an IP destination address that replies with a response (called ICMP echo reply) if this feature is enabled. It may be disabled for security and overhead reasons. Ping measures the time taken for a response to arrive (i.e., round-trip time) which, when divided in half, can give an estimate of how long it takes to reach the destination.

Run ping as

$ ping dest-IP -i 1 -c 10 -s 250

from a machine in LWSN 148 or 158. The "-i" flag specifies that packets be emitted once every second, "-s" specifies the size of the packet (in bytes), and "-c" specifies the number of packets that should be emitted. Carry out ping experiments using different dest-IP values. They should include hosts in: LWSN 148/158, one of the CS labs in HAAS, a host in an ITaP lab in a different part of the campus, web server at UI at Chicago, and web servers of three additional universities on the way to the west coast. Record the terminal output using the script utility. Do not send excessive pings to the same destination IP address since an admin at a remote location may view it unkindly (possibly mistake for intrusion or other network security related attack).

Get rough estimates of the physical distance to the target destinations (universities tend to run their own web servers which is different from companies that tend to oursource their web services to service providers operating proxies that may be physically located anywhere). Using the speed-of-light (SOL), estimate the round-trip time latencies based on physical distances and the results from ping. How significant are the differences? In your view, what may be the biggest factors (ranked by magnitude) that contribute to the observed discrepancy?

Find Internet sites in Africa, Asia, Europe, and South America (one each), to which the ping values are especially high. How do the ping round-trip times (RTTs) compare against SOL based latency estimations? How do the estimation errors (as a percentage of ping RTT) compare to that of the U.S. university ping measurements?
 

Traceroute (30 pts):

The traceroute utility tracks the route that a packet from a source machine follows to the destination machine over an IP internetwork. It outputs the intermediate hops that a packet goes through and displays their corresponding IP address and delay (similar to ping). It discovers intermediate hops along a route to the destination by utilizing a field in the IP packet header called time-to-live (TTL). Every IP router, mandated by Internet standard, decrements the TTL value (a counter that is set to some positive value at the originating sender) of incoming IP packets. If the TTL value (after decrementing) becomes zero, the router discards the packet (i.e., it is not further forwarded) and sends a special IP packet (called ICMP time exceeded) to the originating sender informing it that its packet has been dropped. When the originating sender sends an IP packet with TTL value 1, it will elicit a response from the first router on the path to the destination machine thus discovering its identify. When the originating sender sends an IP packet with TTL value 2, it will discover the second hop, and so forth. As with ping, for security and overhead related reasons, a router may have its ICMP time exceeded response disabled in which case it becomes invisible to traceroute probing.

Run traceroute with the destinations from the ping experiment. Use script to record the output. For those destinations where route discovery is successful, use the domain names of the intermediate routers to guess what ISPs are traversed to reach the final destination. Determine who Purdue's ISPs may be and if our ISPs change based on destination. From traceroute responses within Purdue's own network, determine if the path taken by a packet sent out from LWSN 148/158 changes as it makes its way toward Purdue's gateway. Is there anything special about some of Purdue's router IP addresses (think of private/public IP addresses discussed in class)? Would these router IP addresses be reachable from the outside world? Try it yourself by going to traceroute.org and performing traceroute from Stanford University to the routers inside Purdue. What do you observe? What happens if you traceroute from Stanford to the lab machine you have been using? Make sense of your findings.