Design and Application of MIMO Technology in 3G

The demand for mobile communication air interface bandwidth is increasing. To this end, LTE has selected technologies such as MIMO to achieve the goal of high bandwidth.

Since LTE still needs a long period of time to be commercialized, and the already deployed WCDMA network has consumed a lot of investment by operators, HSPA+ was born as a transitional technology. HSPA+ absorbs many advanced technologies in LTE, and MIMO is an important part of it.

2 Definition and Development History

MIMO, also known as Multiple-Input Multiple-Output (MIMO) system, refers to a communication system that uses multiple antennas at the transmitting and receiving ends at the same time, which exponentially increases the capacity and spectrum utilization of the communication system without increasing the bandwidth.

MIMO technology was first proposed by Marconi in 1908, using multiple antennas to suppress channel fading. In the 1970s, some people proposed to use MIMO technology in communication systems, but the foundation work that greatly promoted MIMO technology in wireless mobile communication systems was completed by scholars from Bell Labs in the 1990s: Telatar gave the MIMO capacity under fading conditions in 1995; Foshinia gave the D-BLAST (Diagonal Bell Labs Layered Space-Time) algorithm in 1996; Tarokh et al. discussed space-time codes for MIMO in 1998; Wolniansky et al. used the V-BLAST (Vertical Bell Labs Layered Space-Time) algorithm to establish a MIMO experimental system in 1998, achieving a spectrum utilization rate of more than 20 bit/s/Hz in indoor experiments, which is extremely difficult to achieve in ordinary systems. These works have attracted great attention from scholars from various countries and have led to the rapid development of MIMO research.

3 Main Technologies of MIMO

Currently, MIMO technology mainly improves wireless transmission rate and quality through three methods:

● Spatial Multiplexing: The system divides the data into multiple parts and transmits them on multiple antennas at the transmitting end. After the receiving end receives the mixed signal of multiple data, it uses the independent fading characteristics between different spatial channels to distinguish these parallel data streams. This achieves the purpose of obtaining a higher data rate within the same frequency resource.

● Transmission diversity technology, represented by Space Time Coding: Jointly encode the data stream at the transmitting end to reduce the symbol error rate caused by channel fading and noise. Space Time Coding increases the redundancy of the signal at the transmitting end, so that the signal obtains diversity gain at the receiving end.

● Beam Forming: The system generates a directional beam through multiple antennas, concentrating the signal energy in the direction of the desired transmission, thereby improving the signal quality and reducing interference to other users.

(1) Spatial multiplexing

Spatial multiplexing technology is to transmit independent signals at the transmitting end, and the receiving end uses interference suppression method for decoding. At this time, the air interface channel capacity increases linearly with the increase of the number of antennas, which can significantly improve the transmission rate of the system, see Figure 1.



Figure 1 Schematic diagram of spatial multiplexing system

When using spatial multiplexing technology, the receiving end must perform complex decoding processing. The main decoding algorithms in the industry are: zero forcing algorithm (ZF), MMSE algorithm, maximum likelihood decoding algorithm (MLD), layered space-time processing algorithm (BLAST, Bell Labs Layered Space-Time).

Among them, the zero forcing algorithm and the MMSE algorithm are linear algorithms, which are relatively easy to implement, but have high requirements on the signal-to-noise ratio of the channel and poor performance; the MLD algorithm has good decoding performance, but its decoding complexity increases exponentially with the increase of the number of transmitting antennas. Therefore, when the number of transmitting antennas is large, this algorithm is not practical; the BLAST algorithm, which combines the advantages of the above algorithms, has the best performance and complexity.

The BLAST algorithm is an effective space-time processing algorithm proposed by Bell Laboratories and has been widely used in MIMO systems. BLAST algorithm is divided into D-BLAST algorithm and V-BLAST algorithm.

D-BLAST algorithm was proposed by GJ Foschini of Bell Labs in 1996. For D-BLAST algorithm, the original data is divided into several sub-data streams, each sub-stream is encoded independently, and is cyclically distributed to different transmitting antennas. The advantage of D-BLAST is that the data of each sub-stream can reach the receiving end through different spatial paths, thereby improving the reliability of the link, but its complexity is too large to be used in practice.

In 1998, GD Golden and GJ Foschini proposed an improved V-BLAST algorithm, which no longer decodes all received signals at the same time, but first decodes the strongest signal, then subtracts the strongest signal from the received signal, then decodes the strongest signal in the remaining signal, subtracts it again, and repeats this cycle until all signals are decoded.

In October 2002, the world's first BLAST chip was launched at Bell Labs, marking the beginning of the commercialization of MIMO technology.

(2) Space-time coding

Space-time coding increases the redundancy of the signal by joint coding at the transmitting end, so that the signal obtains diversity gain at the receiving end, but the space-time coding scheme cannot increase the data rate. See Figure 2 for the system block diagram of space-time coding.



Figure 2 Schematic block diagram of space-time coding system

Space-time coding is mainly divided into space-time trellis code and space-time block code.

Space-time trellis code can enable the system to obtain diversity gain and coding gain at the same time without sacrificing the system bandwidth. However, when the number of antennas is constant, the decoding complexity of space-time trellis code increases exponentially with the increase of diversity degree and transmission rate.

In order to reduce the decoding complexity of the receiver, Alamouti proposed the concept of space-time block code (STBC). STBC allows the receiving end to use only simple linear processing for decoding, thereby reducing the complexity of the receiver.

(3) Beamforming

Beamforming technology is also called smart antenna. It performs phase weighting on the correlation of the output signals of multiple antennas, so that the signals form in-phase superposition (Constructive Interference) in a certain direction and phase cancellation (Destructive Interference) in other directions, thereby achieving signal gain, see Figure 3.



Figure 3 Signal simulation effect of directional smart antenna

When the system transmitter can obtain channel state information (such as TDD system), the system will adjust the phase of the signal transmitted by each antenna according to the channel state (same data) to ensure the maximum gain in the target direction; when the system transmitter does not know the channel state, a random beamforming method can be used to achieve multi-user diversity.

4 Advantages and disadvantages of the three technologies and application scenarios

Spatial multiplexing can maximize the average transmission rate of the MIMO system, but can only obtain limited diversity gain. When used when the signal-to-noise ratio is small, high-order modulation methods such as 16QAM may not be used.

Wireless signals will be frequently reflected in dense urban areas and indoor coverage environments, making the fading characteristics between multiple spatial channels more independent, thereby making the effect of spatial multiplexing more obvious.

In suburban and rural areas, wireless signals have fewer multipath components and the correlation between spatial channels is greater, so the effect of spatial multiplexing is much worse.

Space-time coding of the transmitted signal can obtain additional diversity gain and coding gain, so that high-order modulation methods can be used in wireless environments with relatively low signal-to-noise ratios, but the rate dividend brought by spatial parallel channels cannot be obtained. Space-time coding technology can also perform well in situations where wireless correlation is large.

Therefore, in the actual use of MIMO, spatial multiplexing technology is often used in combination with space-time coding. When the channel is in an ideal state or the correlation between channels is small, the transmitter adopts a spatial multiplexing transmission scheme, such as dense urban areas and indoor coverage; when the correlation between channels is large, a space-time coding transmission scheme is adopted, such as suburban and rural areas. This is also the method recommended by 3GPP in FDD systems.

Beamforming technology can achieve better signal gain and interference suppression when it can obtain channel state information, so it is more suitable for TDD systems.

Beamforming technology is not suitable for dense urban areas, indoor coverage and other environments. Due to reflection, on the one hand, the receiving end will receive too many path signals, resulting in poor phase superposition effect; on the other hand, a large number of multipath signals will make it difficult to estimate DOA information.

5 Application of MIMO Technology in 3G

The combination of spatial multiplexing technology and space-time coding technology enables MIMO to play a good role in different usage scenarios. It is precisely because of this that the 3GPP organization has incorporated MIMO technology into the HSPA+ standard (R7 version).

Considering the comprehensive consideration of cost and performance, MIMO in HSPA+ adopts a 2×2 antenna mode: dual antennas are used for downlink transmission and dual antennas for reception; in order to reduce the cost and size of the terminal, a single antenna is used for uplink transmission. In other words, the utility of MIMO is mainly used in the downlink, and the uplink is only used for transmission antenna selection.

In HSPA+, MIMO specifies the downlink precoding matrix, including 4 forms:

● Spatial Multiplexing.
● Space Time Block Coding.
● Beam Forming.
● Transmit Diversity.

In actual use, the base station automatically selects the use according to different wireless environments.

In the uplink of HSPA+, MIMO technology has two antenna selection schemes, namely open loop and closed loop.

● The open-loop solution is TSTD (Time Switched Transmit Diversity), where uplink data is sent alternately between antennas to avoid fast fading of a single channel, see Figure 4.



Figure 4 Open-loop antenna selection solution

● In the closed-loop solution, the terminal must send reference symbols from different antennas, and the base station will measure the channel quality and then select the antenna with good channel quality to send data, see Figure 5.



Figure 5 Closed-loop antenna selection solution

MIMO technology can greatly improve spectrum utilization, allowing the system to transmit higher-speed data services in a limited wireless frequency band. As the inventor of MIMO technology, Alcatel-Lucent first proposed to add MIMO technology to the 3GPP standard and actively promoted the application of MIMO technology in HSPA+. We believe that MIMO technology will occupy an important position in future mobile networks.