Troubleshooting Multipath

An in-phase or out-of-phase RF wave cannot be seen, so we must look for the effects of multipath in order to detect its occurrence. When doing a link budget calculation, in order to find out just how much power output you will need to have a successful link between sites, you might calculate an output power level that should work, but doesn't. Such an occurrence is one way to determine that multipath is occurring.

Another common method of finding multipath is to look for RF coverage holes in a site survey (discussed in Chapter 11). These holes are created both by lack of coverage and by multipath reflections that cancel the main signal. Understanding the sources of multipath is crucial to eliminating its effects.

Multipath is caused by reflected RF waves, so obstacles that more easily reflect RF waves, such as metal blinds, bodies of water, and metal roofs, should be removed from or avoided in the signal path if possible. This procedure may include moving the transmitting and receiving antennas. Multipath is likely the most common "textbook" wireless LAN problem. Administrators and installers deal with multipath daily. Even wireless LAN users - because they are mobile - experience problems with multipath. Users may roam into an area with high multipath, not knowing why their RF signal has been so significantly degraded.


Solutions for Multipath

Antenna diversity was devised for the purpose of compensating for multipath. Antenna diversity means using multiple antennas, inputs, and receivers in order to compensate for the conditions that cause multipath. There are four types of receiving antenna diversity, one of which is predominantly used in wireless LANs. The type of transmission diversity used by wireless LANs is also described below.

• Antenna Diversity - not active

• Switching Diversity

• Antenna Switching Diversity – active

• Phase Diversity

• Diversity Transmission

Figure 9.6 illustrates an access point with multiple antennas to compensate for multipath.


Antenna diversity is made up of the following characteristics that work together to compensate for the effects of multipath:

1. Antenna diversity uses multiple antennas on multiple inputs to bring a signal to a single receiver.

2. The incoming RF signal is received through one antenna at a time. The receiving radio is constantly sampling the incoming signals from both antennas to determine which signal is of a higher quality. The receiving radio then chooses to accept the higher quality signal.

3. The radio transmits its next signal out of the antenna that was last used to receive an incoming signal because the received signal was a higher quality signal than from the other antenna. If the radio must retransmit a signal, it will alternate antennas until a successful transmission is made.

4. Finally, each antenna can be used to transmit or receive, but not both at the same time. Only one antenna may be used at a time, and that antenna may only transmit or receive, but not both, at any given instant.


Most access points in today’s wireless LANs are built with dual antennas for exactly this purpose: to compensate for the degrading effects of multipath on signal quality and throughput.


Hidden Node

Multiple access protocols that enable networked computing devices to share a medium, such as Ethernet, are well developed and understood. However the nature of the wireless medium makes traditional methods of sharing a common connection more difficult.

Collision detection has caused many problems in wired networking, and even more so for wireless networks. Collisions occur when two or more nodes sharing a communication medium transmit data simultaneously. The two signals corrupt each other and the result is a group of unreadable packet fragments. Collisions have always been a problem for computer networks, and the simplest protocols often do not overcome this problem. More complex protocols such as CSMA/CD and CSMA/CA check the channel before transmitting data. CSMA/CD is the protocol used with Ethernet and involves checking the voltage on the wire before transmitting. However, the process is considerably more difficult for wireless systems since collisions are undetectable. A condition known as the hidden node problem has been identified in wireless systems and is caused by problems in transmission detection.

Hidden node is a situation encountered with wireless LANs in which at least one node is unable to hear (detect) one or more of the other nodes connected to the wireless LAN. In this situation, a node can see the access point, but cannot see that there are other clients also connected to the same access point due to some obstacle or a large amount of distance between the nodes. This situation causes a problem in medium access sharing, causing collisions between node transmissions. These collisions can result in significantly degraded throughput in the wireless LAN, as illustrated in Figure 9.7.


Figure 9.7 illustrates a brick wall with an access point sitting on top. On each side of the wall is a wireless station. These wireless stations cannot hear each other's transmissions, but both can hear the transmissions of the access point. If station A is transmitting a frame to the access point, and station B cannot hear this transmission, station B assumes that the medium is clear and can begin a transmission of its own to the access point. The access point will, at this point, be receiving transmissions that have originated at two points and there will be a collision. The collision will cause retransmissions by both stations A & B, and again, since they cannot hear each other, they will transmit at will thinking the medium is clear. There will likely be another collision. This problem is exacerbated with many active nodes on the wireless LAN that cannot hear one another.

Troubleshooting Wireless LAN Installations

Nulling

The condition known as nulling occurs when one or more reflected waves arrive at the receiver out-of-phase with the main wave with such amplitude that the main wave's amplitude is cancelled. As illustrated in Figure 9.4, when reflected waves arrive out-ofphase with the main wave at the receiver, the condition can cancel or “null” the entire set of RF waves, including the main wave.


When nulling occurs, retransmission of the data will not solve the problem. The transmitter, receiver, or reflective objects must be moved. Sometimes more than one of these must be relocated to compensate for the nulling effects on the RF wave.


Increased Signal Amplitude

Multipath conditions can also cause a signal’s amplitude to be increased from what it would have been without reflected waves present. Upfade is the term used to describe when multipath causes an RF signal to gain strength. Upfade, as illustrated in Figure 9.5, occurs due to reflected signals arriving at the receiver in-phase with the main signal. Similar to a decreased signal, all of these waves are additive to the main signal. Under no circumstance can multipath cause the signal that reaches the receiver to be stronger than the transmitted signal when the signal left the transmitting device. If multipath occurs in such a way as to be additive to the main signal, the total signal that reaches the receiver will be stronger than the signal would have otherwise been without multipath present.


It is important to understand that a received RF signal can never be as large as the signal that was transmitted due to the significance of free space path loss (usually called path loss). Path loss is the effect of a signal losing amplitude due to expansion as the signal travels through open space.

Think of path loss as someone blowing a bubble with bubble gum. As the gum expands, the gum at any point becomes thinner. If someone were to reach out and grab a 1-inch square piece of this bubble, the amount of gum they would actually get would be less and less as the bubble expanded. If a person grabbed a piece of the bubble while it was still small (close to the person's mouth, which is the transmitter) the person would get a significant amount of gum. If the person waited to get that same size piece until the bubble were large (further from the transmitter), the piece would be only a very small amount of gum. This illustration shows that path loss is affected by two factors: first, the distance between transmitter and receiver, and second, the size of the receiving aperture (the size of the piece of gum that was grabbed).