2 useful starting points. One from quicken over at the IDC club on........
ragingbull.com
WCDMA and cdma2000
Two Nokia engineers have written a detailed chapter in a new book "The Mobile Communications Handbook", Jerry Gibson, IEEE Press 1999, 2nd Edition. The chapter is "An Overview of cdma2000, WCDMA, and EDGE", Tero Ojanpera and Steven Gray, page 36-1. I attempt below to describe the differences and similarities between the two CDMA technologies for those interested. Comments and corrections to my understanding are appreciated. To the telecom engineers, this is probably too simplified.
For another good review of wideband features vs. narrow band CDMA I recommend Bill's site at telecomtechstocks.com
Thanks also to Corpgold for the recent patent information.
Before I start, some background (as much for me as anybody).
Code Division multiple access is a transmission technique that handles multiple carriers [lines of communication] by breaking up or spreading each transmission into a train of pulses called chips. How each transmission is broken up is determined by a unique code so that a receiver (e.g. phone) on the other end that knows this code can rebuilt the original transmission into a recognizable one. Without this knowledge, the frequency band [the range of radio frequencies to work with] would sound like empty noise to a receiver. The channels are decoded with correlators: intelligent trial and error searches that look for the signal that matches a code. [note: I believe Corpgold has posted in the past that cdma2000 is restricted to a correlational filter while WCDMA can use a faster matched filter; this chapter does not use this terminology or make this distinction]. As the number of carriers increases in the available bandwidth, performance is said to degrade "gracefully" as opposed to TDMA which has a dead-end limit on the # of carriers. Also, because there are all these noisy transmissions coded in the available bandwidth, interference becomes very sensitive to the power of each transmission and CDMA hence requires a lot of efficient power management. As transmissions bounce of structures they can arrive at the receiver delayed and weakened. Recognizing this multipath fading as it is called is important since these signals can possibly be reconstructed. This is what a RAKE receiver does and it basically amounts to having an extra antenna in the cell phone.
Some definitions:
forward link: a transmission from base station to mobile (phone)
reverse link: mobile to base station.
And now the chapter:
Among the first points the authors make is that there is no longer a distinction between the WCDMA proposals from Japan and Europe. They are now one.
WCDMA and cdma2000 similarities: Common to both standards is the basic concept of spreading signal based on codes discussed above. This idea actually was first used for antijam communications in hostile environments. Variable spreading techniques, error coding, and modulation of signal with phase shifts are also common to both standards. I think it is important to identify this commonality since the point has been made on this board that Nokia chose IDC in part because they believed they would minimize Qualcomm intellectual property. If this is true, then Qualcomm has little if any property in this similarities group. One would also assume that some technologies unique to WCDMA (discussed below) are also not Qualcomm intellectual property. Far be it for me to conclude anything myself though.
Bandwidth: 3G proposals specify that emissions must fall within a 5 MHz bandwidth so both standards must comply to this. The two standards do not fill this 5 MHz restriction equally however. It turns out that the calculated minimum channel separation for non-overlapping carriers in WCDMA (given desired data rates, chip rates and other restrictions) happens to be 4.999712 MHz; for cdma2000 it is below 4 MHz (~3 IS-95 channels). The WCDMA chip rate has also been selected for backward compatibility with GSM (all of Europe) and PDC (Japan's Personal Digital Cellular). cdma2000 chip rate is a derivation from IS-95 (cdmaOne, USA and Korea).
Frame structure: More basic differences between WCDMA and cdma2000 are in structure, i.e. the specifics of the communication framework.
In WCDMA, time multiplexing is used. Time multiplexing is similar to TDMA where communication channels are broken up and cascaded in sequence except instead of just voice channels, large CDMA-coded blocks are inserted in the time slots. For the forward link, each block consists of control information (pilot, power, rate information) and data. For the reverse link, separate blocks for control and data are used. Blocks are broken up based on a unique code and spread out in the available spectrum as discussed above. There is no multicarrier capability as in cdma2000 (see below); instead different time slots are provided. All transmissions are either spread out in the available 5 MHz or time slotted.
One of the advantages of CDMA is the ability to provide different services with different requirements in the pipe on demand. Data rates will therefore vary depending on the demand (e.g. video vs. voice). As the data rate requirements increase the spread coding requirements increase. WCDMA will appropriately build the coded blocks based on these requirements, assigning each to different time slots as the demand dictates. WCDMA also specifies room for short (10 ms) bursts of packet data (for internet applications). Dedicated channels for longer bursts of data can also be specified.
In cdma2000 no time multiplexing is used. Instead there are separate channels, each coded for spreading within 1.25MHZ of spectrum (same as IS-95): fundamental, supplemental, pilot, and dedicated control channels. Two additional carriers are added each with 1.25 MHz bandwidth, to use up about 3.75 MHz of the available 5 MHz. Alternatively, spreading over about 3.6 MHz can be used with just one carrier. IS-95 limited data rates are used on the fundamental channel, but higher rates are possible in the supplemental channel, again with appropriately more complex coding, as with WCDMA [there are coding methods used in WCDMA and cdma2000 that are common between the two and some that are unique; these are presently beyond my understanding]. Burst modes for packet data are also used in cdma2000. Service requirements are accommodated by coding or channel allocations (unlike WCDMA which accommodates increase service demands with coding or using new time slots).
That's it. I am a little closer to understanding Goldberg's 11/18/99 press release.
quicken
...and a 1997 article from the CDMA trade association:
cdg.org |
