Key technologies of WCDMA HSDPA

HSDPA (High speed downlink packet access) is used to realize high speed downlink data service of WCDMA network, which can make the downlink data rate reach 8~10mbps, which is regarded as the main solution of the 3G era. The data rate can reach 20Mbps in the HSDPA system with multiple and multiple out (MIMO) technology. The emergence of HSDPA has aroused great concern of the industry. As the following evolution technology of WCDMA system, many key technologies in HSDPA are similar to CDMA20001XEV/DV and some key technologies in TD-SCDMA, so it is very important to study HSDPA in time for us to fully understand the technology trend in the 3G era.

In the HSDPA technical scheme, the key technologies involved include 4 kinds: Adaptive coding modulation, H-ARQ, Fast cellular selection (FCS), multiple-entry antenna processing (MIMO).

Adaptive coded modulation

Adaptive modulation and Coding (AMC) also belong to the category of link adaptive. The basic principle of AMC is to change the modulation and coding format and make it fit within the system limits and channel conditions, while the channel conditions can be estimated by sending feedback. In AMC system, the higher-order modulation mode and higher coding rate are used in the ideal channel condition, and the lower-order modulation coding method is used in the less ideal channel condition.

The advantages of AMC are as follows: The user in the advantageous position can have higher data rate, thus the average cellular throughput is improved; In the process of link adaptation, the interference level can be reduced by adjusting the modulation coding scheme instead of adjusting the transmitting power.

Currently, AMC faces several challenges. First, AMC is sensitive to measurement error and delay, in order to choose suitable modulation mode, we must first know the quality of channel, and the error of channel estimation may make the system choose the wrong data transmission rate, so that the transmission power is too high, the system capacity is wasted or the BER is increased because the power is too low. Due to the time-varying characteristics of the mobile channel, the delay of the channel measurement report reduces the reliability of the channel quality estimation. In addition, the variation of interference also increases the error of measurement, at this time can seek the combination with other technologies, such as the use of mixed decision feedback retransmission technology (H-ARQ) Can reduce the requirement identification of MCS and sensitivity to measurement error and flow fluctuation. (Computer science)

H-arq

H-arq is also a kind of link adaptive technology. In AMC, an explicit c/i measurement is used to set the modulation encoding format, while in H-arq, the information of the link layer is used for retransmission judgment.

There are many ways to implement h-arq:chase merging, compatible rate drilling turbocodes, and incremental redundancy. The strategy of Chase merging is to send a data group with the same encoding, and then the multiple resend information can be combined with the SNR weighting on the receiving end to obtain the diversity receiving and decoding. Incremental redundancy or h-arq-ii is another way to implement H-arq. This strategy is to send additional redundant information when the first decoding fails, instead of sending the entire data code group back again. H-arq-type-ⅲ is also a kind of incremental redundancy scheme, however, in H-arq-type-ⅲ, each retransmission can be decoded, which is different from harq-ii. In multiple redundant h-arq-type-ⅲ, the different bits are punched each time the redundant information is sent back.

AMC can flexibly select the appropriate MCS based on the UE determination or the information condition provided by the network, but the UE is required to do the exact channel measurement and be affected by the corresponding delay. H-arq can automatically adapt to the change of channel conditions and is insensitive to measurement error and delay. The combination of AMC and H-ARQ can get the best results--AMC provide rough data rate selection and H-ARQ can fine-tune data rates based on data channel conditions.

Fast Cellular selection

FCS is recommended for use in HSDPA. Using Fcs,ue can indicate that one of the best neighborhoods is used for downlink. Determine the "best" honeycomb not only to be based on the conditions of wireless signal transmission, but also to consider the power of the cell in the Activeset and the code word space resources. Generally speaking, at the same time there are many communities in the activeset, but only the most suitable base station to allow the transmission, which can reduce interference to improve system capacity.

At the far edge from the center of the community, each channel has a lower quality. Using the FCS policy you can select a service cell to make the link quality relatively stable. It is through the c/i and the uplink dcch the community instruction information to compare each district. FCS is similar to the physical layer requirements and the selective Diversity Emission (SSDT) in Release99.

If you use the unit selection between Node-b, the Harq state and scheduling table synchronization need to be implemented after HSDPA scheduling and terminal readiness. A method of transmitting State synchronization is achieved through the physical layer of airborne propagation. If FCS can choose to change the Node-b, then all node-b will be able to detect the physical layer of the uplink to send signals, which contradicts the conventional uplink power control strategy. This strategy does not determine that uplink send signals can be detected by all node-b. There are two ways to solve this problem: using an improved uplink power strategy, the UE's transfer power can be increased when any node-b is needed; the second way is to use a conventional power control strategy, but add a power offset to ensure that the transmission state can be detected by the new node-b. In both of these ways, the second approach is preferred. However, it is important to assess how much power offsets are required and their performance impact on the system as a whole.

MIMO Technology

Multi-input (MIMO) systems use multiple antennas simultaneously on the sending and receiving side, which is more beneficial than using more than one antenna on the transmitter side. In MIMO system, the peak throughput can be improved by code multiplexing technology.

Using code multiplexing technology, the channel/scrambling pairs allocated for Hs-dsch are used to modulate m independent data streams (m for the number of antennas sent). Data reused with the same channel code and scrambling code must be distinguished by spatial parameters, which requires at least m antennas at the receiving end. In theory, the peak transmission rate used for code multiplexing is m-fold of single antenna transmission. Code multiplexing can be combined with code multiplexing technology and a lower-order constellation modulation such as 16QAM to achieve a moderate rate of data transmission, without the use of code multiplexing technology, to achieve the same data rate may need to use 64QAM modulation. Compared with the same rate achieved by using the single antenna transmission plus the higher order constellation modulation, the code multiplexing technique can reduce the eb/n0 requirements and improve the performance of the whole system.

When focusing on the MIMO technology used in HSDPA, it focuses on the representative open-loop mode of MIMO. A set of downlink channels (n) are shared among multiple users in a HSDPA sent by a conventional single antenna. The same number of downlink channels are used, but each code word is reused for m times, and each code word is used to modulate different data substrings. Special data is encoded at a higher coding rate, rate matching and interleaving.

For the joint detection of UE, a reply path estimation is performed between each transceiver antenna pair. In the flat fading channel, the characteristic of the channel can be determined by the MP's reply-path factor. In the frequency selective fading channel, the channel characteristic can be characterized by LPM, in which L is the finger number of the rake receiver. The channel estimation can be obtained by the correlation operation of the receiving signal and m orthogonal pilot sequence. Compared with the conventional single antenna receiver, the complexity of channel estimation increases MP times. For data detection, each antenna must be followed by a matched filter for n spread-spectrum codes. Generally speaking, each antenna needs to be LP a diffuser. For each sub stream of the MN data stream, corresponding to the LP amplifier output, each output is weighted by the complex conjugate of the corresponding channel estimation, and then combined together to form a sufficient statistic. This process is called space-time rake reception, which is the extension of single antenna rake receiving in the case of multi antenna processing.

Each quantity (scalar) in the full statistics (vector) of M data substrings that share the same code word contains spatial multiple access disturbances, however, in the flat fading channel, because the orthogonality of the code word is maintained in the channel transmission process, as a group of these substrings are not subject to the interference of the substring generated by other code words. For each group in the M-coded substring, multiuser detection is used to eliminate the effect of Mai. The multiuser detection methods that can be used include maximum likelihood detection and verticalblast detection. The maximum likelihood detection method can be deduced directly by the noise variance of the full statistic vector, but the complexity of maximum likelihood detection is exponentially increasing with m, so the suboptimal but less complex v-blast is a feasible method. The V-blast detector consists of two parts: a linear transformation and a serial interference canceller, linear transformation eliminates mai by forcing 0 algorithm or least mean square error criterion, after the linear transformation, the highest signal-to-noise ratio coded symbol in the sub stream is detected and the corresponding signal in the sufficient statistic is extracted. Using corrected full statistics, linear transformations and interference cancellation are repeated until all substrings are detected. After the MIMO detector, the MN substring is restored to a high-speed data stream, the solution is mapped to a bit, and then the solution is interleaved and decoded.