A fundamental specification of an 802.11 card is its receive sensitivity. The receive sensitivity is the minimum power level at which a signal can be reliably received. For example, a NIC manufacturer may indicate that their particular card has a receive sensitivity of –96 dBm at 1Mb/sec. If the actual RF energy present at that card were less than –96 dBm, then the card would no longer be able to differentiate between signal and noise. The NIC would not detect the incoming packet at all, and the packet would be lost. But how do vendors measure receive sensitivity and what are the implications of their methods for assessing an 802.11 card's performance?
We asked a major vendor of 802.11 hardware how they measured receive sensitivity in their cards. They told us that to measure receive sensitivity, the WLAN card is placed into an RF-shielded room. This guarantees that the test signal will be the only RF transmission in the room, and no background noise in the environment will interfere with the test. The test receiver is placed on a rotating turntable so that measurements can be taken (and then averaged) for all possible horizontal orientations of the receiving antenna. The vendor then transmits packets at weaker and weaker power levels. As the power level decreases, the bit error rate as measured by the card increases. The receive sensitivity of the card will be the minimum power level at which the bit error rate remained below a certain threshold. Therefore, a lower receive sensitivity value (-93 dBm) is better than a higher one (-85 dBm), since it means that the card was able to “reliably receive” data at lower power levels.
Of course, different data rates, having more and less complex encoding and modulation methods, and being more and less resistant to corruption, will result in different receive sensitivities. As data rate increases, receive sensitivity decreases. To put it another way, the higher the data rate, the stronger the signal strength must be for the packet to be reliably received. This is why 802.11 cards drop to lower data rates when interference is present or when they are at the edges of their coverage range. For example, an 802.11b card might have specifications like this:
Receive sensitivity -95 dBm at 1 Mbps
Receive sensitivity -91 dBm at 2 Mbps
Receive sensitivity -89 dBm at 5.5 Mbps
Receive sensitivity -85 dBm at 11 Mbps
While receive sensitivity might seem like a reliable way of comparing two vendors’ cards, we know of no organization that certifies the veracity of the vendor’s results. Therefore, there is the potential for vendors to manipulate the thresholds of their tests to influence their chipset’s receive sensitivity numbers. For example, a vendor that uses a BER threshold of one error in every 1,000,000,000 bits) will end up with lower receive sensitivities than a vendor that uses a BER threshold of one error in every 100,000,000 even though the second vendor’s card may actually be better at receiving bits. Fortunately, some vendors make their BER threshold available in their card's documentation.
2006年11月2日星期四
[www.connect802.com] RSSI Measurement and dB-Milliwatts (dBm)
Most 802.11 analysis tools and vendors' client management utilities provide a representation of signal strength. Four units of measurement are used to represent RF signal strength in 802.11. These are: mW (milliwatts), dBm ("db"-milliwatts), RSSI (Received Signal Strength Indicator), and a percentage measurement. All of these measurements are related to each other, some more closely than others, and it's possible to convert from one unit to another.
The first two units to consider are the mW and the dBm (pronounced "dee-bee-em" or spoken as "dee-bee milliwatts"). Although these are not the most common units in 802.11, we discuss them first because they are the most basic. Just like a pound is a basic unit for measuring weight, a watt is a basic unit for measuring energy (and, in keeping with metric conventions, a mW is one one-thousandth of a watt). It turns out that measuring RF energy in mW units is not always convenient. This is due, in part, to the fact that signal strength does not fade in a linear manner, but inversely as the square of the distance. This means that if you are a particular distance from an access point and you measure the signal level, then you move twice as far away, the signal will have decreased by a factor of four. This relationship can be characterized as logarithmic, and one can say that "RF power drops off logarithmically."
The "dBm" is a logarithmic measurement of signal strength. Since it is logarithmic, just like the power of the RF signal, as the RF signal's strength changes (logarithmically), the dBm value changes linearly. To put it more generally, if you measure a quantity that changes logarithmically (RF power) with a linear unit (mW), the unit will change logarithmically, which is inconvenient. If you measure a quantity that changes logarithmically with a logarithmic unit (dBm), the unit will change linearly, which is more convenient.
dBm values can be exactly and directly converted to and from mW values. Just like miles and kilometers can be converted directly, so can mW and dBm. The formulas to convert are:
dBm = log(mW) * 10
mW = 10^(dBm/10)
We use dBm because it's much easier to say, and write, "-96dBm" than have to say "0.000 000 000 25 mW". That's a lot of zeroes! You should realize that convenience and ease-of-understanding are two fundamental reasons why the dBm metric is used for RF signal strength, rather than mW.
The IEEE 802.11 standard defines a mechanism by which RF energy is to be measured by the circuitry on a wireless NIC. In 802.11b, g, and a, this numeric value is an integer with an allowable range of 0-255 (a 1-byte value) called the Received Signal Strength Indicator (RSSI). Notice that nothing has been said here about measurement of RF energy in dBm or mW. RSSI is an arbitrary integer value, defined in the 802.11 standard and intended for use internally by the physical and data link layers (the hardware in the card and its drivers). For example, when an adapter wants to transmit a packet it must be able to detect whether or not the channel is clear (i.e.: nobody else is transmitting). If the RSSI is below some very low threshold then the chipset decides that the channel is clear. 802.11 does not require that a particular RSSI value correspond to any particular mW value, so each vendor makes this decision on its own. This means that you probably can't compare "signal strength" values between two vendors' chipsets, because those values are based on the RSSI, and different vendors' chipsets associate different power levels with different RSSI values.
To circumvent the complexities (and potential inaccuracies) of using RSSI as a basis for reporting dBm signal strength, it is common to see signal strength represented as a percentage. The percentage represents the RSSI for a particular packet divided by the maximum RSSI value (multiplied by 100 to derive a percentage). If all vendors used that formula for converting RSSI to signal strength percentage, then percentage for signal strength would provide a reasonable cross-vendor metric for use in network analysis and site survey work. However, if vendors do not consistently use the formula above, then we once again end up in a scenario where it's impossible to compare numbers from different vendors. For example, a vendor might hypothetically use a logarithmic function to map RSSI to signal strength, which would cause the signal strength to stay at high values longer as RSSI decreased, and then to drop off very rapidly as RSSI approached zero. Frankly, we don't know the exact details, on a model-number-by-model-number basis, of how each of the many NIC manufacturers map RSSI to signal strength percentage, so its difficult to draw concrete conclusions on this matter.
The first two units to consider are the mW and the dBm (pronounced "dee-bee-em" or spoken as "dee-bee milliwatts"). Although these are not the most common units in 802.11, we discuss them first because they are the most basic. Just like a pound is a basic unit for measuring weight, a watt is a basic unit for measuring energy (and, in keeping with metric conventions, a mW is one one-thousandth of a watt). It turns out that measuring RF energy in mW units is not always convenient. This is due, in part, to the fact that signal strength does not fade in a linear manner, but inversely as the square of the distance. This means that if you are a particular distance from an access point and you measure the signal level, then you move twice as far away, the signal will have decreased by a factor of four. This relationship can be characterized as logarithmic, and one can say that "RF power drops off logarithmically."
The "dBm" is a logarithmic measurement of signal strength. Since it is logarithmic, just like the power of the RF signal, as the RF signal's strength changes (logarithmically), the dBm value changes linearly. To put it more generally, if you measure a quantity that changes logarithmically (RF power) with a linear unit (mW), the unit will change logarithmically, which is inconvenient. If you measure a quantity that changes logarithmically with a logarithmic unit (dBm), the unit will change linearly, which is more convenient.
dBm values can be exactly and directly converted to and from mW values. Just like miles and kilometers can be converted directly, so can mW and dBm. The formulas to convert are:
dBm = log(mW) * 10
mW = 10^(dBm/10)
We use dBm because it's much easier to say, and write, "-96dBm" than have to say "0.000 000 000 25 mW". That's a lot of zeroes! You should realize that convenience and ease-of-understanding are two fundamental reasons why the dBm metric is used for RF signal strength, rather than mW.
The IEEE 802.11 standard defines a mechanism by which RF energy is to be measured by the circuitry on a wireless NIC. In 802.11b, g, and a, this numeric value is an integer with an allowable range of 0-255 (a 1-byte value) called the Received Signal Strength Indicator (RSSI). Notice that nothing has been said here about measurement of RF energy in dBm or mW. RSSI is an arbitrary integer value, defined in the 802.11 standard and intended for use internally by the physical and data link layers (the hardware in the card and its drivers). For example, when an adapter wants to transmit a packet it must be able to detect whether or not the channel is clear (i.e.: nobody else is transmitting). If the RSSI is below some very low threshold then the chipset decides that the channel is clear. 802.11 does not require that a particular RSSI value correspond to any particular mW value, so each vendor makes this decision on its own. This means that you probably can't compare "signal strength" values between two vendors' chipsets, because those values are based on the RSSI, and different vendors' chipsets associate different power levels with different RSSI values.
To circumvent the complexities (and potential inaccuracies) of using RSSI as a basis for reporting dBm signal strength, it is common to see signal strength represented as a percentage. The percentage represents the RSSI for a particular packet divided by the maximum RSSI value (multiplied by 100 to derive a percentage). If all vendors used that formula for converting RSSI to signal strength percentage, then percentage for signal strength would provide a reasonable cross-vendor metric for use in network analysis and site survey work. However, if vendors do not consistently use the formula above, then we once again end up in a scenario where it's impossible to compare numbers from different vendors. For example, a vendor might hypothetically use a logarithmic function to map RSSI to signal strength, which would cause the signal strength to stay at high values longer as RSSI decreased, and then to drop off very rapidly as RSSI approached zero. Frankly, we don't know the exact details, on a model-number-by-model-number basis, of how each of the many NIC manufacturers map RSSI to signal strength percentage, so its difficult to draw concrete conclusions on this matter.
[www.connect802.com] Reciprocity Theorem
There is a basic principle of antennae that is so unexpected (to the uninitiated student) that some people refuse to believe it's true the first time they hear it. The principle is called the Reciprocity Theorem. Its consequences are that, if we are using the same input and output gain, then regardless of differences in our antenna gain, if one I can hear you, you can hear me. This month, we'll explore this concept and its implications.
Consider the case where an AP has a 12 dBi omni antenna attached and a client has a 2 dBi omni antenna on a PCMCIA card. Both the AP and the client are using 15 dBm of transmit power. It might not surprise you that the AP's high-gain antenna can push a signal a long way out to the client, but you might guess that the client's low-gain antenna couldn't get a signal back to the AP. You'd be wrong. Antenna reciprocity basically means that the exact same qualities that make an antenna good at transmitting a signal also make it good at receiving a signal. To put it another way, the Rayleigh-Helmholtz reciprocity theorem states:
If an electromagnetic force of some particular magnitude is applied to the terminals of antenna "A" and the received current is measured at some other antenna "B" then an equal current (in both amplitude and phase) will be obtained at the terminals of antenna "A" if the same electromagnetic force is applied to the terminals of antenna "B".
As an analogy for an RF antenna, imagine a paddle sticking up out of the smooth surface of a lake. Another paddle is sticking up at the opposite end of the lake. One paddle begins to oscillate back and forth, creating waves that push on the other paddle, causing it to move. In our analogy, the paddles are antennas and the waves are RF waves. To carry the analogy further, imagine that one paddle is much bigger than the other--it represents our high-gain antenna. When the big paddle oscillates, it makes much bigger waves, causing the smaller antenna to move more even though it's got a smaller surface area. When the little paddle oscillates, on the other hand, the big paddle's increased surface area causes it to move more as well! The analogy fails somewhat because, in reality the increased mass of the big paddle would give it enough inertia that it wouldn't really move more, but for the sake of the analogy, the antennas are massless.
Antenna reciprocity arises from a property of physics equations called "time-symmetry". Time symmetry means that it doesn't matter whether time runs forwards or backwards, the physics equations should work out the same. Time symmetry is one of the touchstones of new physics theories. Any theory that violates time symmetry is called into serious question. To understand the significance of time symmetry, consider a pool table with a white ball near one end and a black ball in the center. The white pool ball is accelerated by the force of impact with the cue stick and travels towards the center of the pool table. In the center, the white ball strikes the black ball in a straight, center-to-center impact. The inertia of the white ball is transferred to the black ball and it is now accelerated away from the white ball in a straight line, leaving the white ball stationary at the point of impact. If you were to make a movie of the two balls striking and then played the movie backwards, it would show exactly the same thing except now it would be the black ball that starred in the opening scene of the movie. If the mass, velocity, and other characteristics of the Amazing Pool Ball Adventure movie were represented through mathematical equations, the equations would not be time dependent. Time could run forward or backward and the results would be identical.
To some readers, the reciprocity theorem may be new. The implications of antenna reciprocity are far reaching and, if this is the first time you've encountered the concept, the implications may be too hard to accept without proof. In fact, not only is reciprocity demonstrable in the lab and in real-world installations, but the physicists of the world can provide mathematical proof that the theorem holds true. If the antennae and the space between them are replaced with a network of linear, passive, bilateral impedances, then the current through the network can be calculated in accordance with standard practices in electronics theory. Whether on paper or in practice, given the same input power on both ends, "If you can hear me, then I can hear you!"
Next month, we'll discuss some of the real-world implications of antenna reciprocity
Consider the case where an AP has a 12 dBi omni antenna attached and a client has a 2 dBi omni antenna on a PCMCIA card. Both the AP and the client are using 15 dBm of transmit power. It might not surprise you that the AP's high-gain antenna can push a signal a long way out to the client, but you might guess that the client's low-gain antenna couldn't get a signal back to the AP. You'd be wrong. Antenna reciprocity basically means that the exact same qualities that make an antenna good at transmitting a signal also make it good at receiving a signal. To put it another way, the Rayleigh-Helmholtz reciprocity theorem states:
If an electromagnetic force of some particular magnitude is applied to the terminals of antenna "A" and the received current is measured at some other antenna "B" then an equal current (in both amplitude and phase) will be obtained at the terminals of antenna "A" if the same electromagnetic force is applied to the terminals of antenna "B".
As an analogy for an RF antenna, imagine a paddle sticking up out of the smooth surface of a lake. Another paddle is sticking up at the opposite end of the lake. One paddle begins to oscillate back and forth, creating waves that push on the other paddle, causing it to move. In our analogy, the paddles are antennas and the waves are RF waves. To carry the analogy further, imagine that one paddle is much bigger than the other--it represents our high-gain antenna. When the big paddle oscillates, it makes much bigger waves, causing the smaller antenna to move more even though it's got a smaller surface area. When the little paddle oscillates, on the other hand, the big paddle's increased surface area causes it to move more as well! The analogy fails somewhat because, in reality the increased mass of the big paddle would give it enough inertia that it wouldn't really move more, but for the sake of the analogy, the antennas are massless.
Antenna reciprocity arises from a property of physics equations called "time-symmetry". Time symmetry means that it doesn't matter whether time runs forwards or backwards, the physics equations should work out the same. Time symmetry is one of the touchstones of new physics theories. Any theory that violates time symmetry is called into serious question. To understand the significance of time symmetry, consider a pool table with a white ball near one end and a black ball in the center. The white pool ball is accelerated by the force of impact with the cue stick and travels towards the center of the pool table. In the center, the white ball strikes the black ball in a straight, center-to-center impact. The inertia of the white ball is transferred to the black ball and it is now accelerated away from the white ball in a straight line, leaving the white ball stationary at the point of impact. If you were to make a movie of the two balls striking and then played the movie backwards, it would show exactly the same thing except now it would be the black ball that starred in the opening scene of the movie. If the mass, velocity, and other characteristics of the Amazing Pool Ball Adventure movie were represented through mathematical equations, the equations would not be time dependent. Time could run forward or backward and the results would be identical.
To some readers, the reciprocity theorem may be new. The implications of antenna reciprocity are far reaching and, if this is the first time you've encountered the concept, the implications may be too hard to accept without proof. In fact, not only is reciprocity demonstrable in the lab and in real-world installations, but the physicists of the world can provide mathematical proof that the theorem holds true. If the antennae and the space between them are replaced with a network of linear, passive, bilateral impedances, then the current through the network can be calculated in accordance with standard practices in electronics theory. Whether on paper or in practice, given the same input power on both ends, "If you can hear me, then I can hear you!"
Next month, we'll discuss some of the real-world implications of antenna reciprocity
[www.connect802.com] How To Get More Range: Output Power, Antenna Gain, or Receive Sensitivity?
This month's "Essential Wi-Fi" addresses the issue of increasing range by increasing output power. The specific example given is of switching to higher gain antennas. In that article, we discuss how simply increasing output power may not result in the increase in usable range that you expect. It turns out that the usable range of an AP is dependent on several factors. In this column, we examine those factors and discuss the best way to increase the range of an AP in an indoor environment.
We all know of vendors who sell access points with extremely high power output. For example, one vendor sells an AP that outputs a full watt of power (the FCC maximum) with no external amplifiers. Another vendor sells a PCMCIA card that outputs 250 mW with no external amplification (most PCMCIA cards are around 30 mW). Clients using these devices and expecting dramatic range increases may be disappointed. An 802.11 link is two-way; it doesn't do any good for the one device to be able to blast a signal a long way if the other device can't get a signal back. For example, a 100 mW AP might be able to blast its Beacon packets through several walls. A regular PCMCIA card with 30 mW of output power could receive those Beacons and the user would see the AP in his or her list of available wireless networks. But when the user tried to connect to the AP, the user's card wouldn't have enough power to get back to the AP, and the connection would always fail.
This example points out a fundamental limitation of high-output APs. It might seem at first like a high-powered AP is a great, cost-effective way of increasing the range of your network. An casual site survey, which only measured receive signal strength, would even confirm that the AP's coverage had increased (if all you intend to do is receive packets, then the site survey is right). But as soon as clients actually tried to connect to the network, the flaw would become obvious. In general, there's not much point in having an AP that is much more powerful than the clients that it will serve. All of the AP's extra transmit power will go into pushing signal into areas where the clients can't get a signal back to the AP, and that essentially wastes that coverage, and all the money you spent on a high-powered AP. Typical wireless client cards have an output power between 15 and 30 mW, so we at Connect802 usually design networks with APs that have approximately this output power.
Increased output power is useful if you know that all of the devices in the network will be using the same power output. For example, if you've got two wireless bridges at the end of a point-to-point link, you could increase the usable range of the link by putting an equal amplifier on both of the bridges. This circumstance is difficult to guarantee in the more typical one-AP, many clients scenario, so it's usually better to design that type of network with APs that only put out as much power as the lowest-powered client that you expect to use the network. Another option is to design the network so that clients using maximum power (say, 30 mW) will get maximum data rates (say, 54 Mbps) and clients using less power will still be able to connect, but only at lower data rates.
Increased power output in the AP doesn't result in increased range if clients are not able to get a signal back to the AP, but there's a second parameter in this equation that we have so far neglected: receive sensitivity. Receive sensitivity determines the weakest signal that a device can reliably receive. If an AP combines increased power output with a proportionally better receive sensitivity, then it can not only transmit a more powerful signal to the client, but it can also receive the client's weaker signal. For example, consider a client that transmits at 15 mW and has a receive sensitivity of -72 dBm. Now consider an AP that has a transmit power of 30 mW. The AP is transmitting 3 dB "louder" than the client, so its receive sensitivity must be 3 dB better than the client's (-72 dBm minus 3 dB = -75 dBm) to offset the client's weaker transmit power. If this relationship holds true, then the combination of increased transmit power and increased receive sensitivity will result in better range. To put it simply, just making the AP talk louder doesn't work, but by increasing the AP's receive sensitivity, we've made it both talk louder and listen harder.
Unlike transmit power, the receive sensitivity of an radio is not directly adjustable; it's a fundamental property of the engineering of the radio. But the receive sensitivity of an 802.11 radio differs depending on the data rate that the radio is using, with higher data rates requiring more signal power. Typically, an 802.11 network is designed towards a certain minimum desired data rate, and then radios are purchased with an output power and a receive sensitivity that allows them to achieve that data rate in the required coverage area.
At this point, we have discussed the relationship between usable range, transmit power, and receive sensitivity. Essentially, we are considering two link budgets: one from the client to the AP and one from the AP to the client. These link budgets give a maximum range from the AP to the client and from the client to the AP. The usable range of the AP is limited by the smaller of these two ranges. If the AP-to-client range is already greater than or equal to the client-to-AP range, then increasing the AP's transmit power won't result in more usable range because increasing transmit power only increases the AP-to-client side of the link. Increasing transmit power with a corresponding increase in receive sensitivity will result in more usable range because increasing receive sensitivity also increases the client-to-AP side of the link.
Increasing transmit power can increase usable range in another way that is somewhat independent of receive sensitivity. Each 802.11 data rate requires a certain minimum signal-to-noise ratio. If the signal is too weak to achieve the required signal-to-noise ratio for a given data rate, increased transmit power is the only answer. Increasing receive sensitivity or antenna gain doesn't work because both of those options will amplify the noise as well as the incoming signal, leaving the signal-to-noise ratio the same. In summary, if range is limited by interference, you must at least increase transmit power to the point where the clients receive the minimum signal-to-noise ratio for the desired data rate. At that point, range may still be limited by the AP's receive sensitivity or the client's transmit power or receive sensitivity.
Increasing antenna gain is another way of increasing range. Increasing antenna gain differs from increasing transmit power because of the principle of antenna reciprocity. Put one way, antenna reciprocity states that the same qualities that increase an antenna's gain when it transmits also increase its gain when it receives. This means that an increase in antenna gain on either the client or the AP increases BOTH the client-to-AP side of the link AND the AP-to-client side of the link.
Putting a higher-gain antenna on the access point can be a cheap way of increasing usable range for all clients of that AP, regardless of the clients' transmit powers and receive sensitivities. The tradeoff is that increasing antenna gain changes the antenna's coverage pattern, which may limit the maximum antenna gain that can be achieved. More than about 6-10 dB of antenna gain usually results in coverage area that is small enough to preclude any link except one with fixed endpoints at pre-determined locations.
What conclusions can we draw from this analysis? Putting a higher-gain antenna on the AP increases performance for all clients of the AP, but may not offer enough extra power to significantly increase coverage, especially in indoor environments. Increasing the AP's power output might seem like a cheap way of increasing range, but in reality, a complex link budget relationship exists between transmit power and receive sensitivity on the client and AP that dictates the usable range for each client. Simply dropping a 1000 mW into the center of a building is unlikely to provide the results that the AP's vendor would promise or that naive WLAN administrator might expect. Buying an AP with a very good receive sensitivity (for example, an 802.11g AP with receive sensitivity of -75 dBm or lower at 54 Mbps, -95 dBm or better at 6 Mbps) is a cost-effective way of maximizing the effectiveness of the clients' power output.
We all know of vendors who sell access points with extremely high power output. For example, one vendor sells an AP that outputs a full watt of power (the FCC maximum) with no external amplifiers. Another vendor sells a PCMCIA card that outputs 250 mW with no external amplification (most PCMCIA cards are around 30 mW). Clients using these devices and expecting dramatic range increases may be disappointed. An 802.11 link is two-way; it doesn't do any good for the one device to be able to blast a signal a long way if the other device can't get a signal back. For example, a 100 mW AP might be able to blast its Beacon packets through several walls. A regular PCMCIA card with 30 mW of output power could receive those Beacons and the user would see the AP in his or her list of available wireless networks. But when the user tried to connect to the AP, the user's card wouldn't have enough power to get back to the AP, and the connection would always fail.
This example points out a fundamental limitation of high-output APs. It might seem at first like a high-powered AP is a great, cost-effective way of increasing the range of your network. An casual site survey, which only measured receive signal strength, would even confirm that the AP's coverage had increased (if all you intend to do is receive packets, then the site survey is right). But as soon as clients actually tried to connect to the network, the flaw would become obvious. In general, there's not much point in having an AP that is much more powerful than the clients that it will serve. All of the AP's extra transmit power will go into pushing signal into areas where the clients can't get a signal back to the AP, and that essentially wastes that coverage, and all the money you spent on a high-powered AP. Typical wireless client cards have an output power between 15 and 30 mW, so we at Connect802 usually design networks with APs that have approximately this output power.
Increased output power is useful if you know that all of the devices in the network will be using the same power output. For example, if you've got two wireless bridges at the end of a point-to-point link, you could increase the usable range of the link by putting an equal amplifier on both of the bridges. This circumstance is difficult to guarantee in the more typical one-AP, many clients scenario, so it's usually better to design that type of network with APs that only put out as much power as the lowest-powered client that you expect to use the network. Another option is to design the network so that clients using maximum power (say, 30 mW) will get maximum data rates (say, 54 Mbps) and clients using less power will still be able to connect, but only at lower data rates.
Increased power output in the AP doesn't result in increased range if clients are not able to get a signal back to the AP, but there's a second parameter in this equation that we have so far neglected: receive sensitivity. Receive sensitivity determines the weakest signal that a device can reliably receive. If an AP combines increased power output with a proportionally better receive sensitivity, then it can not only transmit a more powerful signal to the client, but it can also receive the client's weaker signal. For example, consider a client that transmits at 15 mW and has a receive sensitivity of -72 dBm. Now consider an AP that has a transmit power of 30 mW. The AP is transmitting 3 dB "louder" than the client, so its receive sensitivity must be 3 dB better than the client's (-72 dBm minus 3 dB = -75 dBm) to offset the client's weaker transmit power. If this relationship holds true, then the combination of increased transmit power and increased receive sensitivity will result in better range. To put it simply, just making the AP talk louder doesn't work, but by increasing the AP's receive sensitivity, we've made it both talk louder and listen harder.
Unlike transmit power, the receive sensitivity of an radio is not directly adjustable; it's a fundamental property of the engineering of the radio. But the receive sensitivity of an 802.11 radio differs depending on the data rate that the radio is using, with higher data rates requiring more signal power. Typically, an 802.11 network is designed towards a certain minimum desired data rate, and then radios are purchased with an output power and a receive sensitivity that allows them to achieve that data rate in the required coverage area.
At this point, we have discussed the relationship between usable range, transmit power, and receive sensitivity. Essentially, we are considering two link budgets: one from the client to the AP and one from the AP to the client. These link budgets give a maximum range from the AP to the client and from the client to the AP. The usable range of the AP is limited by the smaller of these two ranges. If the AP-to-client range is already greater than or equal to the client-to-AP range, then increasing the AP's transmit power won't result in more usable range because increasing transmit power only increases the AP-to-client side of the link. Increasing transmit power with a corresponding increase in receive sensitivity will result in more usable range because increasing receive sensitivity also increases the client-to-AP side of the link.
Increasing transmit power can increase usable range in another way that is somewhat independent of receive sensitivity. Each 802.11 data rate requires a certain minimum signal-to-noise ratio. If the signal is too weak to achieve the required signal-to-noise ratio for a given data rate, increased transmit power is the only answer. Increasing receive sensitivity or antenna gain doesn't work because both of those options will amplify the noise as well as the incoming signal, leaving the signal-to-noise ratio the same. In summary, if range is limited by interference, you must at least increase transmit power to the point where the clients receive the minimum signal-to-noise ratio for the desired data rate. At that point, range may still be limited by the AP's receive sensitivity or the client's transmit power or receive sensitivity.
Increasing antenna gain is another way of increasing range. Increasing antenna gain differs from increasing transmit power because of the principle of antenna reciprocity. Put one way, antenna reciprocity states that the same qualities that increase an antenna's gain when it transmits also increase its gain when it receives. This means that an increase in antenna gain on either the client or the AP increases BOTH the client-to-AP side of the link AND the AP-to-client side of the link.
Putting a higher-gain antenna on the access point can be a cheap way of increasing usable range for all clients of that AP, regardless of the clients' transmit powers and receive sensitivities. The tradeoff is that increasing antenna gain changes the antenna's coverage pattern, which may limit the maximum antenna gain that can be achieved. More than about 6-10 dB of antenna gain usually results in coverage area that is small enough to preclude any link except one with fixed endpoints at pre-determined locations.
What conclusions can we draw from this analysis? Putting a higher-gain antenna on the AP increases performance for all clients of the AP, but may not offer enough extra power to significantly increase coverage, especially in indoor environments. Increasing the AP's power output might seem like a cheap way of increasing range, but in reality, a complex link budget relationship exists between transmit power and receive sensitivity on the client and AP that dictates the usable range for each client. Simply dropping a 1000 mW into the center of a building is unlikely to provide the results that the AP's vendor would promise or that naive WLAN administrator might expect. Buying an AP with a very good receive sensitivity (for example, an 802.11g AP with receive sensitivity of -75 dBm or lower at 54 Mbps, -95 dBm or better at 6 Mbps) is a cost-effective way of maximizing the effectiveness of the clients' power output.
訂閱：
文章 (Atom)