Michael A. Ruder

Efficient receivers for GSM MUROS downlink transmission

Currently, Multiple Users Reusing One Slot (MUROS) is discussed in 3GPP GERAN as an extension of the GSM standard. In MUROS, two overlaid GMSK signals are transmitted in the same time slot and at the same frequency resource. By this, capacity of existing GSM networks in principle can be doubled and up to four half rate voice users can share one time slot. In this paper, channel estimation and detection is investigated for the MUROS downlink. Two novel channel estimation algorithms are presented, taking into account the specifics of the MUROS downlink. For detection, a joint MLSE of both user signals can be applied in case of noise limited scenarios. For interference limited environments, it turns out that approaches based on the mono interference cancellation (MIC) algorithm for single antenna interference cancellation (SAIC) are more favorable. It is shown that the standard MIC algorithm performs sufficiently well and could be used for a fast introduction of MUROS in existing GSM networks. For enhanced performance, a novel algorithm based on MIC along with successive interference cancellation is proposed. The presented results demonstrate that an even better performance as the GSM reference performance before introduction of SAIC can be obtained for MUROS if well-designed receivers are used.

Combined User Pairing and Spectrum Allocation for Multiuser SC-FDMA Transmission

This paper proposes a strategy that combines user pairing and spectrum allocation for a virtual multiple-input multiple-output (V-MIMO) single-carrier frequency-division multiple access (SC-FDMA) transmission over intersymbol interference (ISI) channels. The users are equipped with a single antenna each, where always two users, composing one pair, transmit their data in the same time slot and frequency band. The receiver at the base station (BS) is assumed to have multiple antennas and separates the users of each pair with a multiuser equalizer. For this, linear minimum mean-squared error (MMSE) equalization and successive interference cancellation (SIC), respectively, are chosen. A solution for the combined optimization of user pairing and spectrum allocation is proposed and evaluated by simulations for a transmission in the uplink of E-UTRA Long Term Evolution (LTE). The simulation results show that the proposed algorithm yields significant gains compared to random user pairing and spectrum allocation. It is also shown that, despite its considerably lower complexity, the performance of the novel algorithm is very close to that of the optimum solution that can only be found by a very complex full search.

User Pairing for Multiuser SC-FDMA Transmission over Virtual MIMO ISI Channels

This paper proposes different criteria for user pairing for a virtual multiple-input multiple-output (V-MIMO) single-carrier frequency-division multiple access (SC-FDMA) transmission over intersymbol interference (ISI) channels. Two users, equipped with a single antenna each, transmit their data in the same time slot and frequency band. The receiver at the base station (BS) is assumed to have two antennas and separates both users with a multiuser equalizer. For this, linear minimum mean-squared error (MMSE) equalization and successive interference cancellation (SIC), respectively, are chosen. Various pairing criteria are proposed and compared by simulations for a transmission in the uplink of E-UTRA Long Term Evolution (LTE). The simulation results indicate that well-designed user pairing strategies yield significant gains compared to random user pairing. It is shown that the proposed novel bit error rate (BER) pairing achieves a significant performance gain compared to capacity pairing for a SIC receiver and higher order modulation.

Joint user grouping and frequency allocation for multiuser SC-FDMA transmission

Abstract This paper proposes strategies for user pairing and frequency allocation for a virtual multiple-input multiple-output (V-MIMO) single-carrier frequency-division multiple access (SC-FDMA) transmission over intersymbol interference (ISI) channels. N u users, equipped with a single antenna each, compose one pair/group and transmit their data in the same frequency band to the base station (BS), equipped with multiple antennas. Multiuser equalization is applied at the BS. The pairs ( N u = 2 ) are either transmitting in subsequent time slots in a time division multiple access (TDMA) scheme using all available bandwidth or in different frequency chunks in a frequency division multiple access (FDMA) scheme. A generalization for N u > 2 users, called user grouping, is given for the FDMA scheme. For the optimization of joint frequency allocation and pairing/grouping, a low complexity solution is proposed and evaluated by simulations for the uplink of Long Term Evolution (LTE). As optimization criteria, the capacity of the resulting V-MIMO channel and the bit error rate (BER) after equalization are evaluated, respectively. The simulation results show that the proposed algorithms yield significant gains compared to random user pairing and frequency allocation and that using BER as an optimization criterion yields a significantly better performance.

LOS and NLOS channel modeling for terahertz wireless communication with scattered rays

In this paper, the wireless communication over indoor Terahertz (THz) channels is studied. The physical mechanisms governing a wireless transmission in the 0.1-10 THz band are a very high molecular absorption and spreading loss which result in a very high and frequency-selective path loss for the line-of-sight (LOS) links. For the non-line-of-sight (NLOS) propagation, a very high reflection loss depending on the shape, material, and roughness of the reflecting surface affects the THz wave propagation. Taking these peculiarities of the THz radiation into account and applying a ray tracing approach for scattered rays, a novel deterministic equivalent channel model is developed that accounts for both the LOS and NLOS propagation cases. Furthermore, the channel capacity of the proposed model is investigated. Simulation results demonstrate that for distances, up to 1 m, data rates in the order of Terabit per second (Tbps) are obtained for a transmit power of 1 Watt. Moreover, the capacity of only the NLOS component is around 100 Gigabit per second (Gbps). These results are highly motivating to develop future wireless THz communication systems.

Receiver Concepts and Resource Allocation for OSC Downlink Transmission

Voice services over Adaptive Multi-user channels on One Slot (VAMOS) has been standardized as an extension to the Global System for Mobile Communications (GSM). The aim of VAMOS is to increase the capacity of GSM, while maintaining backward compatibility with the legacy system. To this end, the Orthogonal Sub-channels (OSC) concept is employed, where two Gaussian minimum-shift keying (GMSK) signals are transmitted in the same time slot and with the same carrier frequency. To fully exploit the possible capacity gain of OSC, new receiver concepts are necessary. In contrast to the base station, where multiple antennas can be employed, the mobile station is typically equipped with only one receive antenna. Therefore, the downlink receiver design is a very challenging task. Different concepts for channel estimation, user separation, and equalization at the receiver of an OSC downlink transmission are introduced in this paper. Furthermore, the system capacity must be improved by suitable downlink power and resource allocation algorithms. Making realistic assumptions on the information available at the base station, an algorithm for joint power and radio resource allocation is proposed. Simulation results show the excellent performance of the proposed channel estimation algorithms, equalization schemes, and joint radio resource and power allocation algorithms in realistic VAMOS environments.

MMSE Beamforming for SC-FDMA Transmission over MIMO ISI Channels with Linear Equalization

We consider transmit beamforming for single-carrier frequency-division multiple access (SC-FDMA) transmission over frequency-selective multiple-input multiple-output (MIMO) channels. The beamforming filter is optimized for minimization of the sum of the mean-squared errors (MMSEs) of the transmitted data streams after equalization. Here, MIMO minimum mean-squared error linear equalization (MMSE-LE) is considered. We prove that for SC-FDMA transmission eigenbeamforming diagonalizing the overall channel together with a nonuniform power distribution is the optimum beamforming strategy in the MMSE sense. The derived optimum power allocation is similar in spirit to classical results for the optimum continuous-time transmit filter for linear modulation formats obtained by Berger/Tufts and Yang/Roy. Moreover, we present a modification of the beamformer design to mitigate an increase of the peak-to-average power ratio (PAPR) which is in general associated with beamforming. Simulation results demonstrate the excellent performance of the proposed beamforming algorithm.

Cramer-rao lower bound for channel estimation in a MUROS/VAMOS downlink transmission

Voice over Adaptive Multi-user Channels on One Slot (VAMOS) is an extension of the Global System for Mobile Communications (GSM) standard, where two overlaid Gaussian minimum-shift keying (GMSK) signals are transmitted in the same time slot and on the same frequency. The overlaid signals are usually assigned different powers to combat slow fading and propagation loss. For channel estimation, specific training sequences are used for both signals. In the downlink, at each mobile station the channel coefficients and the sub-channel power imbalance ratio (SCPIR) have to be estimated. In this paper, the Cramer-Rao lower bound (CRB) for the training sequence based joint estimation of the SCPIR and the channel coefficients at the mobile station is derived and compared with the CRB for a conventional GSM transmission. The results are compared with the performance of channel estimation algorithms designed for VAMOS. It turns out that these algorithms perform very close to the CRB.

Codebook aided user pairing and resource allocation for SC-FDMA

This paper proposes the use of codebook aided algorithms for joint user pairing and resource allocation for a virtual multiple-input multiple-output (V-MIMO) single-carrier frequency-division multiple access transmission. Two users, equipped with a single antenna each, transmit their data in the same time slot and frequency band to a base station with two receive antennas. Finding the optimum solution to the joint resource allocation and user pairing problem is NP-hard. Therefore, two codebook aided algorithms with low complexity are proposed, where one of them is based on the Global Vector Quantization (GVQ) algorithm. The algorithms are compared with several suboptimal algorithms by simulations for a transmission in the uplink of Long Term Evolution (LTE). It is shown that the proposed algorithms exhibit a low complexity and a performance close to that of the optimum scheme.

Single Antenna Interference Cancellation for GSM/VAMOS/EDGE Using

Different schemes for single antenna interference cancellation (SAIC) in the Global System for Mobile Communications (GSM) have been proposed so far. Most of these schemes work best in synchronous interference environments and are limited to Gaussian minimum-shift keying (GMSK) modulation. In this paper, we propose several modifications of the conventional GSM receiver, which target the harmful effects of asynchronous co-channel interference (ACCI). Based on the observation that the impairment caused by ACCI cannot be modeled as additive Gaussian noise, which was assumed for the derivation of the metric used in state-of-the-art receivers, we propose modeling the impairment as Generalized Gaussian noise (GGN). This leads to modifications of the conventional branch metrics of the reduced-state equalizer and the decoder employed at the receiver. Both proposed modifications are applicable to any linear modulation alphabet and do not require any modification of the GSM air interface. Hence, the proposed novel branch metrics are also applicable in Enhanced Data Rates for GSM Evolution (EDGE) systems which employ 8-ary phase-shift keying (8-PSK) modulation and where the SAIC algorithms tailored for GMSK cannot be used. Furthermore, we show that the proposed branch metric modifications also improve the performance of the new Voice Services over Adaptive Multiuser channels on One Slot (VAMOS) extension of GSM in ACCI environments. Based on simulation results, the performance of the proposed schemes is compared with that of an unmodified receiver employing the conventional Gaussian metric. We show that the proposed receivers outperform the unmodified receiver in ACCI environments and discuss the complexity entailed by the proposed branch metric modifications.