Orthogonal Resource Partitioning FDMA,TDMA,D-CDMA

(a)FDMA: The allocated bandwidth W is divided into K frequency slots to form K user channels, with each channel occupying a bandwidth of W/K and transmitting their each own signal over entire time domain. No interferencee exist among any two users, therefore these channels are of orthogonality.

Disadvantage: require more than one transceivers at the base station, which is bulky

(b)TDMA: All users in the system share the entire bandwidth W, but transmit signal in a chronological sequence, with no interference among any two users, namely orthogonality.

However,TDMA requires all the K users be synchronized to maintain a common timing reference, which is impractical for exacting synchronization.

To allow a tolerable discrepancy in synchronization, guard time is applied between two time slots,during which no signal is transmitted, but leading to a limit on the maximum cell range.

Two challenges: (1)difficult to transmit the wideband signal due to multipath fading seen by the transmitted signal,which requires equalization to reduce the inter-symbol interference.It is complicated. (2) *Electromagnetic interference.

Advantage: a single RF(costly) at the base station.

(c) D-CDMA: all user share the entire bandwidth and the entire time duration but with different code channels.

Functions of Spread Spectrum:
(1)Interference Suppression;(2)Low probability of Interception;(3)However, No Impact on Channel Noise.

Requirements on Technique:(1)Synchronization;(2)Feasible only in the forward link,namely downlink channel, from the base station.

At the receiver side, the system must know:(1)sequence information;(2)Synchronization, Timing information.

Tips: In D-CDMA, the number of available channels is equal to the available orthogonal code, which is determined by the Spreading Factor,SF. Thus the system is code-limited. Furthermore, the complete orthogonality of code sequence requires perfect synchronization between Ck(t) and Cj(t), which is more feasible in the downlink than in the uplink channel, since the signals originate from the same position.

Random -CDMA_R-CDMA

If we relax the orthogonality between Ck(t) and Cj(t), namely Ck(t)*Cj(t) with small cross-correlation, the system is no longer limited by the number of codes but the interference. That’s interference-limited system.

Assume in the system, the users are far away from the base station, the received power at the BS is S. Using R-CDMA, the BS received signal of user 1, who is the reference user. Due to the small cross-correlation among codes ,the power of other users in the system will not disappear totally, but contributed to interference power.With users in the system increasing, the interference power level is also climbing. Once the interference floor exceeds the maximum tolerable interference level of the system, the performance of the reference user will be poor. Thus, the R-CDMA system is interference-limited, rather than code-limited. In the noise-limited system, the SNR penalty will translate into higher transmitted power or reduced cell range;however, in the interference-limited system, the penalty will translate into capacity loss.

Moreover, in the flat fading situation, the SNR penalty is larger than in AWGN situation in terms of the same BER. Thus, for the same maximum tolerable interference level, in the flat fading situation,the number of supportable users in a single system is lowered. How many users in the system you can support depends on much interference you can tolerate,which is determined by the physical design. By diversity combining method, the flat fading curve will be pushed steeper,thus reducing the SNR penalty. For the same tolerable interference level, the number of supportable users will increase.

In order for R-CDMA to work, it’s important to maintain the transmitted power, which is realized by the power control mechanism at the receiver side, telling the transmitter whether the transmitted power is high or low. The power control, aiming at suppressing the interference, in R-CDMA system is fast control mechanism; whereas the power control mechanism in GSM is slow, with the purpose of saving battery.

For releasing confusion, note that PG is the operational parameter of R-CDMA, different from spreading factor as the parameter for D-CDMA. SF governs the maximum number of orthogonal spreading code sequence, while PG determines the ability to suppress multi-user interference.

Then the two possible ways to improve the capacity is(1) to increase the spreading factor but at the expense of the bandwidth; and (2)to reduce the require SINR by optimizing the physical design,e.g.coding and diversity.

Since R-CDMA imposes loose constraint on orthogonality among codes, it is usually applied in the uplink channel.