The automotive industry perceives high-resolution radar sensors as one of the backbones of autonomous driving technology. With human safety being at stake, the topic of calibration is of the utmost importance. Yet, realizing possibly large volumes of accurate measurements of devices at 77 / 79 GHz, with 15 cm radiation aperture or more, is a challenge. Applying the Fraunhofer formula in such case results in a prohibitive direct far-field range length of 11.5 m. To address that issue, this paper introduces a new measurement system, consisting of a short-size focal length offset-fed compact antenna test range (CATR), interfaced with an analog echo generator. With a chamber size of only 0.9 m × 2 m × 1.6 m, the setup is designed to test apertures up to 30 cm size. Measurements are presented which involve a 4-D imaging radar on chip (RoC). Results obtained in the CATR and in a reference 7 m far-field range are compared and show excellent agreement. Reducing the cost and size of the test environment offers the possibility to measure more, and in particular to perform full angular calibrations, with the target being swept through the complete azimuth and elevation region. Such procedure is compared to a typically used diagonal calibration approach and is demonstrated to bring significant improvement in the compensation of sensor biases, with a 10 dB increase in peak-to-side-lobe ratio.

A pulse-width and pulse-position modulator (PWPM) IC for RF carriers from 170 MHz to 2.8 GHz is presented. The IC features a digital 5 bit pulse-width (PW) and 6 bit pulse-center (PC) input interface, updated at the RF carrier frequency, a small arithmetic unit, two delay locked loops with a new phase detector, two phase selectors, CMOS pulse logic and two differential binary RF outputs. At 900 MHz, a 14 MBd 256-QAM (112 Mb/s) signal with an EVM of 1.83 % and an ACLR of -45 dB is shown. At 2016 MHz, an ultra-broadband 504 MBd 16-QAM (2.016 Gb/s) signal with an EVM of 13.5 % and a BER of 1.5·10 -4 is demonstrated. The IC is implemented in a 28 nm fully depleted silicon-on-insulator (FDSOI) CMOS technology and runs from a 1.0 V supply. It consumes 38 mW at 900 MHz and 58 mW at 2016 MHz carrier frequency.

This paper presents an all-digital transmitter solution particularly suited for Massive multiple-input and multiple-output (MIMO) systems for mobile communications. Massive MIMO is a key candidate to address the challenges of future mobile communication standards, especially to provide higher capacity in dense urban scenarios. While the required communication theory is elaborated to a great extend, the transceiver hardware complexity remains a potential economical show-stopper. This paper demonstrates that all-digital transmitters can be employed to reduce size, cost, and engineering effort of heavily parallelized transmit architectures. Therefore, today's all digital transmitter concepts are analyzed and improvements are suggested to increase performance and feasibility. The realized setups prove that the key specifications of mobile communication standards can be met utilizing dedicated integrated circuits or even by using off-the-shelf FPGAs and their high-speed interfaces. We show that we can generate 8 parallel 5 MHz LTE signals at 2.6 GHz out of a single FPGA with an ACPR of 48 dB with a coding efficiency of 50% using only binary waveforms.

All-digital transmitter is a hot topic since a few years. One of the key elements of all-digital transmitters is the modulator, which modulates the baseband IQ signal onto a carrier frequency in the form of digital binary signals. The combination of delta-sigma modulation and pulse width modulation is a very strong candidate. However, the regular sampled pulse width modulator introduces non-linear distortions and images also known as aliasing effects. Due to these non-linear distortions and images, the signal integrity is violated especially when the power spectral density of the baseband input signal is not symmetrical around DC. In this paper, we propose a baseband injection method with low complexity to pre-compensate these unwanted emissions. This novel method does not have to run at the operating frequency of pulse width modulator but can run at a much lower frequency. Moreover, each term of the non-linear distortions and images can be individually compensated without mutual influence. In Matlab simulations we show that all the compensated non-linear distortions and images are suppressed by over 20 dB and disappear below the noise floor of delta-sigma modulator.

This paper describes a novel concept to encode amplitude and phase information in binary waveforms for switch-mode power amplifiers (SMPAs). A combination of a phase-modulated delta-sigma modulation (DSM) and pulse-width modulation (PWM) makes it possible to adapt signal generation to the power amplifier limits. The proposed system overcomes the inherent signal quality limits of digital systems with fixed clock frequencies by applying a phase modulated clock. We demonstrate that it is possible to achieve the performance of conventional concepts at less than half of the bit rate. Simulation and measurement results are given for important figures of merit of mobile communication signals with high peak-to-average power ratio (PAPR). A proof of concept is implemented in a laboratory setting using a conventional FPGA.

This paper presents new architectures for RF pulse width modulation (RF-PWM). They allow for very efficient and simple implementation of this class of modulators. Even for the 2.6-Ghz band, the modulator can now be built with standard components. The new concepts support binary and M-ary output alphabets. One variant is capable of generating M-ary RF-PWM signals by only deploying one switching device. The new architectures are derived analytically and are illustrated by simulation results. A recent hardware implementation of this concept proofs the efficiency of this approach.

We show a universal analog signal conditioning approach for switch-mode power amplifiers. Three novel RF-PWM concepts are implemented on PCB, which work with a single RF input. The modulation-inherent distortion is compensated by a dedicated predistortion. We evaluated the performance for base stations of mobile communications in terms of ACLR, EVM, coding efficiency and channel power from 400MHz to 3 GHz. For a standard UMTS signal a coding efficiency of 82%, added EVM of 3% and ACLR of 49 dB are achieved.

This paper discusses RF-PWM for mobile communication applications focusing on maximal achievable coding efficiency. Coding efficiency is one of the most crucial metrics in evaluating modulation concepts for switch mode amplification. It is shown that the achievable efficiency is mostly dependent on the chosen modulator waveform and the magnitude probability density function of the wanted signal. We determine theoretical limits and unveil unexpected benefits of unipolar waveforms. The results are applied to realistic communication test signals with various system parameter sets.

In this paper, we present a novel so-called class-O distributed base station system where the remote unit and the central unit are connected via an optical link. In contrast to existing systems, we transmit a binary RF signal over the optical link which significantly reduces the complexity of the remote unit. The conversion from an analog to a binary signal is implemented using a fast pulse-width-modulator in the RF domain. For downlink this pulse-width-modulator is combined with a class-S amplifier, meeting the 3GPP ACLR requirements within an excellent 10dB electrical output power range. The uplink shows a dynamic range of 65dB for a 5MHz LTE signal. Lab measurements show robust optical transmission of the RF signals over up to 40km fiber. It is shown that both downlink and uplink are capable of supporting simultaneous transmission and reception, respectively, of non-contiguous frequency bands.

In this paper the effect of on-air-combining on the signal quality in distributed transmitter systems like active antennas is described. Active antennas containing a multitude of parallel, independent transmitter chains are emerging in the mobile communication base station market. Due to the independent transmitter chains, each signal is subject to different distortions by the various transmitter components like D/A-converters, filters, amplifiers, noise sources etc. These multiple signals are then radiated by individual antenna elements, which, according to their phase and amplitude offsets, then combine on air; forming the required beam pattern. In this paper the effect of the combining process is described; a system simulation tool in Matlab and Simulink is described and the measured verification-results of an implemented test system are presented.

In this paper a statistical analysis of the impact of phase- and amplitude-distortions on the performance of active antenna arrays for mobile communication base stations is presented. The cause and influence of different contributors to these distortions is described and the allowable phase- and amplitude-margins that need to be reached for mobile communication base station applications are derived. Finally a test setup of a distributed active antenna is implemented and it is shown, that the required accuracy can be achieved with standard RF-circuit design and components.