Code division multiple access

from Wikipedia, the free encyclopedia

The code division multiplex method ( Code Division Multiplex , CDM or Code Division Multiple Access , CDMA ) is a multiplex method that enables the simultaneous transmission of different user data streams on a common frequency range. The essential property of the commonly used frequency range is a larger bandwidth than the useful data stream occupies. For frequency spreading and to distinguish the different data streams are specific spreading codes used.

Applications of CDMA in the range of the digital signal transmission in mobile radio networks of the third generation (3G) such as CDMA2000 and UMTS . Other areas of application for synchronous CDMA are the Global Positioning System (GPS) and Galileo satellite navigation systems .


Generation of a CDMA signal

With CDMA, the focus is on the separation and differentiation of different data streams transmitted in parallel over a shared and dedicated frequency band. To distinguish the data streams are coded with special spreading codes, these code sequences additionally having certain properties such as orthogonality and, in certain applications, based on pseudo randomness , whereby the original useful data streams can be obtained separately from one another on the receiver side through correlation with the spreading code sequence. In contrast to the classic multiplex methods such as frequency division multiplex and time division multiplex , the individual data streams are superimposed both in the frequency domain and in the time domain in the code division multiplex.

The figure on the right shows the generation of a spread-spectrum signal: The source data sequence shown above, consisting of four bits ( 1-0-1-0 ) with a bit duration of T b, is multiplied by a higher-frequency spreading code sequence to produce the spread-spectrum signal shown below and to receive the transmission signal to be transmitted. The individual elements of the spread data sequence are referred to as chips and have a length of T c . In the figure, the chip rate (cps) is 10 times as high as the bit rate (bps) to be transmitted , which results in a band spread of roughly a factor of 10. The resulting bandwidths are, depending on the application, in the range of a few MHz.

For the code division multiplexing method, there must be a separate spread data sequence for each subscriber, which has the lowest possible cross-correlation with all other code sequences used . Such code sequences are also referred to as orthogonal . One example is the strictly orthogonal Walsh codes . In general, the relationship that the longer the code sequence, the lower the user data rate. In return, the number of possible users increases or the transmission power decreases. Depending on the application, the transmitters can be assigned code patterns of different lengths, as is done with UMTS. For a gross data rate of 1.92 Mbps, a code sequence with a length of 4 bits is used in this method. At 30 kbps the code sequence is 256 bits long. A constant chip rate of 3.84 Mcps is used.

Synchronous CDMA

Walsh code with four chips, as used in synchronous CDMA

In the case of synchronous CDMA, all subscribers must align the code used for band spreading with one another in exactly the same time. This means that no so-called code phase shifts are permitted. This is only possible with certain technical applications, for example with the radio link ( downlink ) from the Base Transceiver Station (BTS) to the individual mobile subscribers according to the CDMA2000 mobile radio standard.

In the synchronous case, it is possible to use strictly orthogonal sequences such as the Walsh code in order to minimize mutual crosstalk ( Multi User Interference , MUI). In addition, pseudo-random sequences such as gold sequences are also used with synchronous CDMA , for example in the UMTS mobile radio standard and in the Global Positioning System (GPS).

Asynchronous CDMA

With asynchronous CDMA there is no way to control the code phase shift between the participants. In this case, it is no longer the cross-correlation between the individual code sequences that is decisive, but the odd cross-correlation , which allows any shifts between the code sequences.

In asynchronous CDMA, pseudo-random data sequences with a significantly longer period than a symbol interval are typically used for band spreading. These code sequences are not optimized for particularly good mutual orthogonality, which means that there are strongly different correlation properties from symbol to symbol. This weakness has to be compensated for by an appropriate channel coding . The advantage of asynchronous CDMA, however, is that a large number of code sequences, although only approximately orthogonal, are available for a large number of different subscribers.

One area of ​​application of asynchronous CDMA is the radio link ( uplink ) from the mobile subscribers to the Base Transceiver Station (BTS) according to the CDMA2000 mobile radio standard. However, general spread spectrum methods such as Direct Sequence Spread Spectrum (DSSS) also belong to the asynchronous CDMA method. In the case of DSSS, the focus is on spreading the spectrum compared to code division multiplexing and multiple use.