As a result, many large companies are now grappling with the issue of improving their indoor cellular coverage. This two-part article discusses the technical aspects of cellular coverage, the challenges of providing in-building coverage, and the various approaches to meeting those challenges. In this part, we'll focus on the technology and challenges.
Cellular phones and e-mail terminals have become essential equipment in enterprises, bringing into focus the challenge of providing adequate cellular coverage inside buildings. According to IDC (June, 2006), enterprise telecommunications managers now believe that more than 28 percent of their employees use their cell phones as their primary connection.
3G technologies such as UMTS/HSDPA and CDMA EV-DO are also growing rapidly: the GSM Association announced in June 2006 that the technology now connects 2 billion users worldwide, and projects that there will be some 700 million HSDPA users by 2010.
As device usage soars and applications expand, so does indoor use. In a 2005 study, Japan's NTT DoCoMo found that 70 percent of 3G wireless calls originate from within buildings. Now, many companies are finding that their indoor cellular coverage isn't adequate.
Transmission systems
Providing effective cellular coverage requires distributing the correct frequencies with sufficient signal strength and quality, which means overcoming signal loss (attenuation) and interference. This challenge is greater inside buildings because materials such as steel and concrete (as well as office furniture, equipment, and other objects) attenuate radio signals.
The technical attributes of the signal vary according to the transmission system being used. North American carriers currently employ several systems (GSM, UMTS, CDMA-1xRTT, CDMA-EV-DO, and iDEN) to provide coverage, using four frequency bands:
- Two iDEN bands, 806 to 824 MHz uplink and 851 MHz to 869 MHz downlink, as well as 896 to 901MHz uplink and 935 to 940MHz downlink.
- 850 MHz band used for GSM and CDMA, which ranges from 824 MHz to 849 MHz in uplink and 869 MHz to 894 MHz in downlink
- 1900 MHz frequency band, which ranges from 1850 MHz to 1910 MHz in uplink and 1930 MHz to 1990 MHz in downlink
Due to length considerations, we will confine our discussion to GSM, UMTS/HSDPA, and CDMA-1xRTT/CDMA EV-DO.
The primary goal of any radio system is to provide clear connections to all users who want them. The variables affecting coverage include signal strength, interference (which dictates the spacing and placement of antennas), and capacity (the number of simultaneous calls per cell).
The system will also need to support various data rates based on whether the application is voice or data. It is therefore useful to consider the strengths and weaknesses of cellular systems in terms of how they meet these requirements.
The Carrier to Interference-plus-Noise Ratio (CINR) is a primary measurement of signal effectiveness. The carrier is the desired signal, and the interference can either be noise or co-channel interference. (Co-channel interference is a particular problem when frequencies are reused at short distances).
In order for the receiver to be able to decode the signal, the signal must fall into an acceptable CINR range, which differs with the technology used (i.e., CDMA, GSM, etc.). CINR is expressed in decibels (dBs).
The minimum CINR depends on the kind of modulation (QPSK, QAM) and coding gain. In networks with a frequency reuse of one (same frequency in each cell), the signal characteristics must support so-called soft handoffs, which occur when a receiver is at the edge of a cell in an area where the signal strength from its current cell is almost equivalent to the signal strength from the adjacent cell using the same frequency. Soft handoff is possible if the CINR requirements are -5 dB or less, allowing the two different signals to vary a couple of dBs.
With these basics in mind, let's look at the differences between the predominant cellular technologies.
GSM
GSM uses a 200 KHz channel with eight timeslots per carrier. The timeslots can be used by a control channel (at least one is needed per cell), or by voice channels and the remaining data channels. To increase the capacity of a coverage cell, each cell can include multiple GSM carriers (= TRX).
However, GSM requires a CINR of +9 dB for the modulation and does not have any coding gain, so a soft-handoff will not work in cells, because the two signals from the base stations would violate the CINR requirement (causing too much interference), and the mobile device would not be able to detect these signals. This is why GSM uses different frequencies in order to get the necessary separation.
CDMA 1xRTT
CDMA 1xRTT uses a 1.25 MHz channel. Rather than using separate frequency carriers or different timeslots, it uses code division multiplexing to transmit multiple voice calls on a single carrier.
A CDMA carrier can typically handle 32-48 simultaneous voice calls. Since all mobile phones transmit simultaneously on the uplink (mobile to base station), each mobile device in a CDMA system has to transmit with the right output power in order to preserve the capacity of the cell. To facilitate signal-sharing, CDMA1xRTT uses a very sophisticated power control on the uplink.
One of the advantages in CDMA is the so-called coding gain, which means the information is transmitted over a longer period. The link budget is increased by the coding gain and the CINR requirement is reduced. For voice and low data rates, CINR is negative, so a soft handoff is possible.
For a spreading factor of 128, the coding gain is 21 dB, so if the CINR was +9 dB before spreading, the CINR is -12 dB after spreading. (The spreading factor is the ratio of the chips (CDMA = 1.2288 MChips/s) to baseband information rate.)
CDMA EV-DO
CDMA EV-DO also uses a 1.25 MHz channel and is often used as the second CDMA carrier in CDMA systems. Since EV-DO does not provide circuit voice capability, it is optimized for data and works nicely with CDMA 1xRTT. Downlink speeds can reach up to an impressive 2.4 Mbps, but as a consequence, CDMA EV-DO requires a better signal than CDMA 1xRTT: the link budget is reduced and the CINR requirement is increased.
In fact, data rates of more than 153 Kbps require a CINR greater than "5 dB, and therefore require frequency planning similar to that in GSM (in which adjacent cells use different frequencies). Using the highest data rate of 2.4 Mbps requires a higher CINR than for GSM voice. As a result, soft-handoff for high data rates is not possible with CDMA EV-DO. However, since EV-DO is a data-optimized channel, it can easily carry VoIP traffic instead.
