The goal of every mobile telephone service provider is to conduct as many simultaneous calls as possible (think greed). In most wireless technologies, only one party is permitted to transmit a signal at a given frequency in a defined geographical location, which works fine for applications like broadcasting. (Having two different stations simultaneously transmitting Channel 6 would really cause a headache.) But cellular technology is different.
In the United States, each cellular provider is allocated 25 MHz of spectrum, 12.5 MHz for transmitting (called the downstream) and 12.5 MHz for receiving (called the upstream). Cellular telephony is a full duplex systemboth parties can talk at the same time (think husband and wife) because transmitting and receiving are allocated their own frequencies.
In first-generation cellular systems, each phone conversation is allocated 30 kHz of spectrum. Therefore, each 12.5 MHz of bandwidth can handle 416 simultaneous phone calls as shown in Figure 75. If the cellular service providers were to follow the broadcast model, only 416 total calls could be conducted simultaneously in a given geographical area (an MSA or RSA). Letting only 416 people talk at once in, say, Southern California, would not even satisfy the demands of Beverly Hills.
The good news is that there is no need for cellular service to follow the broadcast model. Since a person on a mobile call only needs their allocated frequency within the cell they are currently in, there is no reason somebody else on the other end of town cannot be using that same exact frequency in an entirely different cell. The concept of multiple users operating at the same frequency, at the same time, and in the same geographic area, is called frequency reuse, and it is what separates mobile telephony from non-cellular wireless communications.
Figure 75 Frequency division multiple access.
For frequency reuse to work properly, it is imperative that each mobile phone only put out enough power to reach the cell site of the cell it is in. If it puts out too much power, it will not only reach the intended cell site, it will reach unintended cell sites, which others may be using at the same frequency for a totally different conversation. This strict limitation on transmitted power, called power management, however, is an advantage in that low power transmission means that the cellular phone's battery will last longer and therefore people can talk longer (always a good thing) between charges.
Referring back to Figure 71, users located in the cells marked with the letter A can both be using the same exact frequency to conduct their own separate conversations. Here is a challenging question: how come adjacent cells cannot conduct different conversations at the same frequency (and the same time)? Imagine that you are a cellular caller on the border between cells and you are communicating with one cell site, but the power level received at the other cell site is almost as great, causing interference to anyone using that frequency in that cell. Because of this potential interference, identical frequencies in adjacent cells cannot be used simultaneously.
Once again there is a tradeoff to be made. To avoid the possibility of interference, cells using the same frequency at the same time must be as far away as possible. Conversely, if the cellular provider wants to make as much money as possible (and they do), the cells must be as close together as possible, so more people can talk simultaneously. In practice, the number of cells of separation, which depends on many things, ranges anywhere from 4 to 21.
As mentioned above, each 12.5 MHz of bandwidth is broken down by frequency into 416 different channels, with one conversation per channel. This dividing up of the frequency band is known as frequency division multiple access or FDMA. FDMA is a type of air interface. Think of air interface as a way of manipulating signals to maximize the capacity of the allocated bandwidth. In the case of FDMA, the manipulation breaks up the allotted frequency band into smaller frequency sub-bands. Is FDMA the only air interface? Just wait.
Having 416 different possible conversations at one time (in a given cell) is fine, but what if there were a way to get more than 416 possible simultaneous conversations at one time out of the same 12.5 MHz frequency allocation? With the new digital technologies available, there are.
The first of these new digital technologies, or air interfaces, is known as time division multiple access or TDMA. TDMA takes the same 30 kHz bandwidth discussed above and breaks it down into time slots, as shown in Figure 76. Notice that the horizontal axis is labeled with "Time." Several conversations can take place simultaneously in the same frequency band because each conversation is periodically allocated a short time slot in which to transmit its message. As you can imagine this requires some sophisticated signal processing, but it does result in higher cell site capacity. Some systems break up the channel into as many as eight different time slots, which theoretically increases the call-carrying capacity of the system eightfold.
Figure 76 Time division multiple access.
Another way to increase call capacity is with an air interface known as code division multiple access or CDMA. CDMA uses a technology (explained in detail shortly) called spread spectrum. In essence, spread spectrum stamps each RF signal with a unique destination address. As a result, many signals can coexist in the same frequency band at the same time since each receiver can only decipher the intended signal (by its address). And because of the miracle of digital technology, more conversations can be crammed into a given bandwidth with CDMA than any other currently employed air interface. Figure 77 is a graphical depiction of CDMA.
Figure 77 Code division multiple access.
Referring to Figure 77, when the RF signal in a CDMA system gets the "address" imprinted on it, the spectrum it occupies gets bigger. For instance, a signal that occupies 30 kHz before the address is applied might occupy 1 MHz after the address is applied. This "spreading" of the occupied frequency is why it is called spread spectrum. At first thought, it might seem that having a signal occupy more frequency than it does in its original form is a mistake. However, even though it does occupy a greater frequency band than in its original form, the system can now pile many signals on top of each other because they can all be distinguished by their "addresses." In this manner, more total signals can fit into a given frequency band, and that is, after all, the goal of every service provider.
Another air interface, and one that has been around for awhile, is called cellular digital packet data or CDPD. Unlike the other air interfaces that try and increase the amount of voice traffic, CDPD is only concerned with data. In fact, CDPD is a packetized data service, which works well for short, bursty data like e-mail. It is meant to enable computers (and not people) to talk to each other.
In a typical configuration, a laptop computer is outfited with a special CDPD modem (usually installed in the PCMCIA expansion slot). The CDPD air interface uses the same infrastructure (see Figure 72) and frequency bands as regular cellular phones, with the PSTN interface essentially replaced by an Internet router. The unique feature of CDPD is that it only occupies unused channels to send the data.
CDPD was originally meant as an overlay to first-generation analog systems in the United States. (When I hear the word "overlay," I think same hardware, new software.) As mentioned above, these first-generation systems have 416 30-kHz channels within a given cell. Rarely are all 416 channels in use at any instant in time (except Friday afternoon rush hour). CDPD uses a special scanning receiver to constantly monitor which, if any, of the 416 channels are available for sending data. In general, voice traffic has priority, so whenever there is a conflict over a channel, the talker wins. This probably stems from the fact that conversations are more time-critical than bursty data. (Who cares if you get an e-mail to your laptop five seconds later?)
With CDPD, because it can only use unallocated channels, the data being sent is constantly hopping from channel to channel. One of the consequences of this is the potential for interference with voice calls. One way around this is to avoid the channel hopping altogether and assign specific channels (among the 416) just for CDPD. The downside with this approach is that it takes away from the voice capacity of the cell. For a service provider to do this, it must make economic sense in terms of increased revenue from the service.
Another interesting thing about CDPD is that it is an always-on connection. There is no need to "dial up" from your laptop. Unfortunately for users of CDPD, the maximum data transfer rate is only 19.2 Kbps. At the time it came out that was pretty fast, but in light of all the recent increases in data throughput, it is hard to know what the future holds for this air interface. The good news is that CDPD can coexist with the newer air interfaces like TDMA and CDMA. And there are companies working with advanced modulation techniques that claim to increase the CDPD data rates up to 400 Kbps, which is screaming fast. Stay tuned.
Cellular Phone Block Diagram
At this point you are probably wondering how a cellular phone works and so I have included a block diagram of one in Figure 78. It is a block diagram of a generic digital cellular phone. Because of all the possible variations, it is not meant to be an exact functional diagram of any particular digital phone, but rather it shows the main functions contained within most digital phones. Keep in mind that some of these "generic" blocks will differ in function and location within the system depending on things like the air interface used.
Figure 78 Block diagram of a digital cellular phone.
Referring to the lower left portion of Figure 78, you will see a microphone. This is where the whole process starts. The microphone just converts sound (air movement) into an analog voltage. Being a digital phone, it cannot remain an analog voltage very long and so one of the first things the analog signal does is get converted to a digital signal by an analog to digital (A/D) converter. A/D converters change the constantly varying analog voltage into a corresponding string of 1s and 0s. Without going into too much detail, let's just say that the higher the (analog) voltage, the more 1s there are and the lower the analog voltage, the more 0s there are.
After the A/D converter, comes the transmitter (Tx) digital signal processor (DSP). The DSP is the real brains of the mobile phone and performs many operations to the digital bit stream. One of the operations it performs is speech encoding. Speech encoding is concerned with speech quality and compression. Recall that compression is used to eliminate redundant information.
Another function of the DSP is channel encoding, which modifies the bit stream to compensate for any errors that might occur during its transmission through the air. The DSP may also be involved in encrypting the bit stream so that no one else can overhear the conversation.
Finally the DSP almost always performs some sort of interleaving. As you will soon learn, when a signal is sent from the mobile unit to the basestation, more than just the conversation information is sent. The interleaving function of the DSP interleaves the voice information with any other information sent.
After the DSP, the bit stream enters the equalizer. The purpose of the equalizer is to compensate for any frequency-dependent impairments that occur during transmission through the air such as phase and amplitude distortion.
After the equalizer, the signal is ready to enter the world of RF. Here the digital bit stream is combined with an RF signal by a modulator. (In this block diagram, you don't see the source of the RF because it is assumed to be contained within the modulator.) In a digital system, the modulation can be frequency modulation or more typically phase modulation. So what comes out of the modulator is just a sine wave with its phase modulated all over the place. However, the frequency of this sine wave is not yet at the frequency of transmission so it gets sent to an upconverter.
The upconverter is really just a mixer that is used to change the frequency of the signal. The "intermediate frequency" RF signal (coming out of the modulator) is combined in the upconverter with another sine wave coming from the synthesizer. The synthesizer generates a frequency such that the output of the upconverter is at the exact frequency of transmission.
Why do cellular phones use a synthesizer? Because they are required to transmit at multiple carrier frequencies. Recall that one way or another, all cellular systems break up their allotted frequency range into multiple sub-bands (using FDMA). Each of these sub-bands requires their own, different RF carrier frequency. The synthesizer must be able to generate all these different frequencies at a moment's notice for the system to work properly.
After the upconverter, the signal is in the exact form it needs to be to be sent wirelessly (it is at the right frequency and the digital information has been modulated onto it). The only problem now is that the signal is too small and so it is sent through a power amplifier which increases the signal to the appropriate power level. Remember that in cellular systems power management is used to ensure the signal is at just the right power level. It is not shown in Figure 78, but the power amplifier is in reality a variable gain amplifier whose gain is controlled by the DSP from information it receives from the basestation. This is how power management is realized in a mobile phone.
Now the signal is ready to be sent wirelessly, but one more thing must be done first. The signal needs to be filtered so that no unwanted frequencies are transmitted. The signal is sent to a duplexer, which is just a double filter (one for the transmit band and one for the receive band). Finally the signal is sent out the antenna to find its way in the wireless world.
On the return path, the signal does many of the things it did on the transmit path only in reverse. The signal first enters the antenna, and because it is quite small by this time, the very next thing it does is get amplified by a low noise amplifier. Then the downconverter lowers the carrier frequency and the demodulator strips away the RF leaving only a digital bit stream.
Once in digital form, the signal goes through the equalizer again and then on to the DSP, where many of the previous DSP functions are done in reverse. Finally the bit stream is sent to a digital to analog (D/A) converter and then to a speaker where you get to hear those magic words: "I'm not here right now, so please leave a message."