Multi Antenna System
– Antenna Diversity
– Space Division Multiple Access
– First MIMO-Developments and Systems
Power amplifier in basis stations
– Intermodulation
– Digital Predistortion (DPD)
– Crest Factor Reduction (CFR)
Remote Radio Head
Near Field Communication

Advanced Technology for mobile broadband

Multi Antenna Systems

Normal classic radio systems use one antenna for sending and receiving. However, it can be advantageous to use several antennas on the receiving side, for example. Technically speaking, this is called Multiple In, Multiple Out (MIMO) systems. MIMO systems are particularly useful when there is multipath propagation.

Antenna Diversity

Let’s assume a signal reaches an antenna via two paths. Both signals are superimposed at the receiving antenna. If both paths are the same length, the signals will be amplified and reception will be good while if the paths are of different lengths, the signals may cancel each other out. With same strength for both paths, the signal can even be completely erased. This can be a problem, especially with WLAN, where the transmitter and receiver do not move and the transmission situation does not change (the “channels” are constant). For this reason, a technique called antenna diversity was proposed. Two antennas are used in parallel. During reception it is tried out which antenna has a better signal. This can happen during reception of the preamble. During this time you switch between two antennas and choose the one with better reception.

SPACE DIVISION MULTIPLE ACCESS

The case in which two or more complete receivers and transmitters are operated at the same time is more interesting. For example, a system in which signals can be sent via two separate transmitters and receivers with antennas that are sufficiently far apart. This is called a Multiple Input, Multiple Output System (MIMO).

If a signal is broadcast through two antennas at the same time, the signal will amplify in certain directions and cancel out in other directions. This is because the angle of the radiation causes the signals from the two antennas to shift so that they ultimately have a 180° phase shift and thus cancel each other out. So there is a spatial radiation pattern due to two antennas. This pattern can be changed by changing the amplitudes and phases of the two paths. For example, if you shift the signal of one path by 180° compared to the other path, the signal will be canceled perpendicular to the antenna plane. By manipulating the phases, the signal can even be canceled in any direction.

Now imagine that you send out not one signal, but two signals X1 and X2 at the same time. In systems with only one antenna, the signals would completely overlap and it would be impossible to tell them apart. Now suppose you manipulate the signal X1 so that it completely cancels out towards the direction with the angel alpha. Then you can still receive the signal X2 at this angle. So you can manipulate signals X1 and X2 so that they propagate differently in space and can be separated.

In the same way, you can also manipulate the reception of the signals on the receiver side by shifting the phases before they are added together. So you can favor or eliminate different directions. If you now have an environment with multipath propagation, it is conceivable that signals X1 and X2 can be manipulated in such a way that they can be separated from each other. They differ in that they have spread differently in space.

You can distinguish between signals by sending them in different directions or receiving them from different directions. This spatial separation is possible using multiple antennas. This is called Space Division Multiple Access (SDMA). Initial considerations SDMA systems were researched early on at Bell Laboratories.

First MIMO developments and systems

The Indian scientist Arogyaswami Paulraj researched MIMO technology in 1992 while working at Stanford University. He patented a first method for MIMO transmission.

Four years later, engineer Gregory Raleigh recognized the benefits of OFDM in conjunction with MIMO. Above all, he succeeded in mathematically proving that MIMO systems are possible. His principles and algorithms were instrumental in the later introduction of MIMO in OFDM based systems.

Both Paulraj and Raleigh founded companies that developed products using MIMO technology. They were called Iospan and Clearwireless. They were later acquired without having created any commercial products.

WLAN 802.11 n was the first standard to use MIMO

In 2001, Raleigh founded another company called Airgo. He finally managed to build the first chipset in 2003 that successfully used MIMO for WLAN and thereby significantly increased the transmission rate. Airgo has now become the driver of a new WLAN standard called 802.11 n, the first standard in the world with MIMO technology. This was published in 2008. At the same time, MIMO standards for WiMAX were already being developed.

Power Amplifier in Base Stations

As discussed with LTE, there are challenges for the transmitter with downlink LTE signals. This is due to the high Peak to Average Power Ratio (PAPR). As a reminder: Up to 1200 subcarriers can be added at 20 MHz. For example, if all amplitudes have the value 1, very high peak values can occur. (If all carriers have approximately the same phase, the amplitude can be over 1000). Such a PAPR leads to distortion if the power amplifier is not linear. With eNodeB (base stations) you have the challenge of getting a PA (power amplifier) that is as linear as possible while at the same time having a high output power. Unfortunately, PAs reach a saturation range where they are no longer linear.

The average power of the output signal must be adjusted so that possible peaks above the average signal still fall into the linear range. Therefore, for signals with high PAPR there is a high so-called back off, i.e. it must be operated well below the maximum power in order to avoid non-linearities. This has the disadvantage that you need very powerful amplifiers and have to operate them in a very inefficient mode. This is very negative, especially in terms of power consumption.

Intermodulation

The problem of nonlinearity leads to so called intermodulation. This occurs when two frequencies that are close to each other are amplified simultaneously via a nonlinearity. Mixed products of the type k1 x f1 + k2 x f2 are created, where k are integers and f1 and f2 are the corresponding frequencies. Let’s assume f1 is 1000 MHz and f2 is 990 MHz, this results:

2 x 1000 MHz – 990 MHz = 1010 MHz
2 x 990 MHz – 1000 MHz = 980 MHz
3 x 1000 MHz – 2 x 990 MHz = 1020 MHz
3 x 990 MHz – 2 x 1000 MHz = 970 MHz

For multicarrier signals, intermodulation has two negative influences. There is interference between the carriers and resulting errors and there are signals beyond the bandwidth of the OFDM signal.

The figure above shows how shoulder-shaped areas arise on the edges of an OFDM spectrum, which are caused by intermodulation. There are now strict regulations as to how strongly a signal is allowed to “radiate” outside of its own band. There may be another OFDM band right next to the OFDM band, possibly the band of a competing operator. This is the main reason why nonlinearities and the resulting intermodulation must be avoided at all costs.

Digital Predistortion

One way to reduce the problem of non-linearity of the power amplifier is so-called digital predistortion. The signal that is sent is digitally pre-processed so that it compensates for the non-linearity. Values that have a high amplitude are artificially increased even further so that the characteristic curve is linearized in the non-linear range.

There are adaptive algorithms for this that are based on feedback of the output signal. For this purpose, the radio-frequency signal is decoupled behind the power amplifier, modulated down again, digitized and fed to the algorithm. The signal must be sampled with enough bandwidth to also capture the signals caused by the intermodulation. The algorithm now sets a predistortion filter so that intermodulation interference is minimized. DPD involves a lot of effort. You need feedback with a receiver structure and a very fast signal processing unit. It must be ensured that this does not require more electricity than the electricity saved through the DPD.

Crest Factor Reduction (CFR)

Another method to mitigate nonlinearity is the so-called crest factor reduction. Digital preprocessing is also used here, which specifically detects the largest amplitudes in an area and reduces them using a filter process. This reduces the PAPR but also creates errors that affect the modulation accuracy.

Remote Radio Heads

Until now, the base transceiver stations for GSM and the NodeBs for UMTS were in a house directly under the radio tower. The analog radio including the power amplifier was also located here. A thick coaxial cable ran from the radio, known as a transceiver (mixture of transmitter and receiver), to the antennas on the tower. They tried to lose as little power as possible, which is why the cables were almost as thick as an arm.

MIMO has now been introduced with LTE. All antennas now had so-called cross-polarized antennas with two antenna inputs. If even 4×4 MIMO was to be supported, an eNodeB had to serve 4 antennas with appropriate coaxial cables. It was no longer economical to use coaxial cables from the ground. That’s why the entire transceiver, including the power amplifiers, was installed directly on the antennas in its own small housing. These were called remote radio heads. They were digitally connected to the eNodeBs. A new interface was defined specifically for this purpose, which was called CPRI (Common Public Radio Interface). Digital I/Q data was transferred.

CPRI was defined by the manufacturers of the network infrastructure from 2003 onwards. Ericsson, Huawei Technologies, NEC Corporation, Nortel Networks, Alcatel-Lucent and Nokia were active. Multiples of 614.5 Mbit/s are transmitted. Up to 16 streams are defined, resulting in a bit rate of 10 Gbit/s. This transmission usually takes place using fiber optic connections.

RFID und Near Field Communication

One way to transmit information wirelessly is through inductive fields. The transmitter and receiver are in close proximity to each other. The transmitter has a coil that generates an alternating electrical field. This not only creates a signal that contains information, but also energy. This is recorded by a receiver located in the area. Typically, low frequency is used. You would need meter-long antennas for real radiation, but this is not necessary since everything is in the direct induction field. An RFID device consists of a flat printed coil connected to a tiny chip. This typically includes identification and can receive and send data with a relatively small data width. Applications are, as the name suggests, identification.

Originally, RFID was used to protect items of clothing from theft or to identify livestock with an RFID button in the ear, for example. In the 1970s, RFID was used for toll systems to identify vehicles passing through a toll station. All of these systems were initially proprietary.

In the noughties, a community of Philips Semiconductors and Sony began to standardize RFID technology. A uniform frequency (at 13.56 MHz) and a defined protocol for communication with a transmission rate of 424 kbit/s was created. This standard is called Near Field Communication (NFC). The technology is better suited to near field than RF, which suggests real radiation of waves. The NFC specification was completed in 2002. As with Bluetooth and WLAN, an NFC forum was soon created in 2004 in which further development is coordinated. As a mobile phone manufacturer, Nokia was in this forum. Nokia was also the first mobile phone manufacturer to build NFC into a device.

NFC has three different modes:

• Card emulation mode
• Peer-to-peer mode
• Reader or Writer Mode

In card emulation mode, an NFC device, such as a smartphone with NFC, behaves like a chip card containing an RFID chip. This makes it possible to use the smartphone like a bank or credit card.

In peer-to-peer mode, two NFC devices can exchange data with each other when they are in close proximity to each other. This functionality can be used, for example, to pair two devices with each other (e.g. for Bluetooth) or to register with a WLAN access point.

In Reader or Writer Mode, an NFC device can read (or write) information from RFID tags. An RFID tag roughly corresponds to a label that can be attached to a product. Instead of reading a barcode or a QR code, you read the contents of the RFID chip.