(Originally posted at http://www.von.com/blogs/guest/2008/12/a-system-view-of-wireless.aspx )
As with most things wireless, end-user performance is impacted by a wide range of factors: power, proximity to base station/access point, number of concurrent users, presence of interferers, backhaul capacity, and so on.
Wireless systems are comprised of a transmitter and a receiver. Other than in broadcast media like FM radio, each device normally performs both the transmit and receive functions. Each transmitter requires energy to send its signal. The strength of a transmission is measured in Watts. The amount of energy available to transmit differs greatly, depending on the device type.
A handheld device is constrained by its size, the capabilities of its battery, and then further by regulatory bodies, such as the FCC in the United States. It is useful to think of these devices as “whisperers” in a wireless network.
Conversely, a cellular base station, access point or FM radio transmitter has very different parameters. They draw power from the grid, so battery life is not a concern. They’re not held to the ear or in close proximity to the body, so concerns about electromagnetic radiation are alleviated. There are regulatory constraints regarding their maximum allowed transmit power, as transmitting energy indiscriminately would result in interference to other transmitters. These would be the “shouters” in a wireless network.
To effectively close a link, receivers on both sides are required to be able to receive this transmitted energy coherently. Many things come into play here, including the effect of RMS delay spread (multipath), Doppler shift (mobility), but for the purposes of this blog, let’s assume these to be neutral. The ability to listen for and receive energy is known as receive sensitivity and is measured in dBm. In this case, lower (a bigger negative number!) is better, as it indicates an ability to receive a weaker signal. For example, an Rx sensitivity of -98dBm is better than an Rx sensitivity of -95dBm by 3dB, or a factor of two. In other words, at a specified data rate, a receiver with a -98dBm sensitivity is able to receive signals that are half as weak as receivers with a -95dBm receive sensitivity.
Most contemporary wireless communication systems have adaptive modulation schemes that are designed to ensure reliable transmission by altering the encoding of the data and backing off on the modulation rate depending on the link quality — which includes, among many other things, the received signal strength. These modulation schemes have some pretty funky names (64QAM, 16QAM, QPSK, BPSK), but what this means is that in order to get the highest possible rates, users have to be pretty much beside the access point.
Unfortunately, this is an area where there is a disconnect between heightened expectations and real-world performance as quoted performance specs are based on theoretical maximums. Despite the fact that an 802.11n access point may have a theoretical maximum rate of 300mbps, the user at the edge of coverage may be operating only at a 1mbps rate. As a result, throughout the coverage area of an access point, or a base station, the overall performance of any wireless system is diminished by the distribution of users operating at their respective rates.
A simple way of looking at this is to take a population of users and assume that they’re evenly distributed across the coverage area of a particular cell, which looks something like a bull’s eye with four concentric circles, each representing a different modulation rate. With thanks to Archimedes, we know that the coverage area of the smallest circle is [π(r)^2]. The area of Circle 2 is then [π(r)^2 - Circle 1 area], and so on. Even if we assume that the increase in radius is equal for each modulation rate, the coverage area within which users would be able theoretically to achieve maximum rates is only 6.25 percent of the total coverage area. In this example, 43.75 percent of the users would be at the lowest rate. What this means is that you would have seven times the number of users operating at the lowest rates as opposed to the highest rates. In reality, this ratio of 1:7 is extremely conservative. Therefore, it is more useful to build models and set end-user expectations that account for average modulation rates throughout the entire cell.
While the real world does not look like the scorched earth model depicted, hopefully this helps to illustrate why average data rates to all the end users throughout the cell of any wireless system typically are a fraction of the peak rates which can be attained only at the center of the cell! Stay tuned, there’s more to come on this topic.
That’s my .02!
Martin Suter is vice president of business development at BelAir Networks, a provider of broadband mesh solutions for Wi-Fi, WiMAX, 4.9 GHz Public Safety and 5.9 GHz ITS networks. Previously, Martin was the CEO at Cohda Wireless, where he raised the company’s profile and negotiated a licensing deal with a Fortune 100 vendor in its core franchise. Prior to Cohda, he was vice president of business development at MeshNetworks Inc., a classic tech transfer/disruptive technology success story that achieved a major liquidity event for its investors in Q4/2004 with its acquisition by Motorola. Martin also was responsible for building several high profile alliances with and for leading technology companies, including Fujitsu, Microsoft, Netscape, Sun Microsystems, and Teradata. Additionally, Martin has successfully negotiated technology transfer, distribution and/or licensing deals with companies like 3Com, BioChem Pharma, Dow Chemical, Exodus, Fujitsu, IBM, Microsoft, Motorola, Netscape and Sun.
