Gregory Morozov

Evaluation of Joint Transmission CoMP in C-RAN based LTE-A HetNets with large coordination areas

Coordinated Multi-Point (CoMP) transmission is considered as one of the key technology enhancements for Rel-11 LTE-A systems. CoMP transmission is realized by exchanging coordination information between a set of transmission nodes, forming a so-called CoMP cluster. The size of the CoMP cluster is typically limited, which leads to cell-edge performance issues appearing at the CoMP cluster boundaries. However, in new radio access network (RAN) architectures such as Cloud RAN (C-RAN) the effect of cell-edge user performance degradation at the CoMP cluster boundaries can be minimized by considering larger CoMP coordination areas. In this paper, we investigate the Joint Transmission (JT) CoMP performance in a C-RAN implementation of LTE-A HetNet with large CoMP cluster sizes. Some approaches to minimize the CoMP processing requirements for this case are also discussed. As a baseline for comparison, the LTE-A time-domain enhanced Inter-Cell Interference Coordination (eICIC) scheme which has no coordination boundary issues is considered.

Quasi-deterministic approach to mmWave channel modeling in a non-stationary environment

There is increasing faith that mmWave technology will be part of 5G wireless networks in the wide frequency range 30-90 GHz. Experimental measurements are used to model mmWave channels addressing issues like human body shadowing or reflections due to moving vehicles. In this paper a new quasi-deterministic (Q-D) approach is introduced for modeling mmWave channels. The proposed channel model allows natural description of scenario-specific geometric properties, reflection attenuation and scattering, ray blockage and mobility effects. This new channel modeling approach is of utmost importance for further measurement campaigns planning, channel model characterization, system level simulations and network access capacity estimations.

Performance evaluation of dynamic point selection CoMP scheme in heterogeneous networks with FTP traffic model

High traffic demand is considered as one of the key challenge in modern wireless communication systems which are typically limited by amount of available spectrum. Recently heterogeneous networks (HetNet) have been proposed as an attractive approach to cope with this problem, by enabling cell splitting gain via deployment of an additional layer of pico-cells. While low power and small antenna height of pico-cell base stations simplify search of the site location, the efficiency of cell splitting gain in HetNet is typically limited by small coverage area of pico-cells. To address the issue of HetNet performance in LTE-A Rel-11, coordinated multipoint (CoMP) traffic management schemes were considered. In particular a combination of dynamic point selection (DPS) with dynamic point blanking (DPB) was identified as a promising mechanism to provide cell load balancing and interference mitigation in HetNet. Dynamicity of CoMP schemes is expected to provide performance benefits in various scenarios including non-full buffer traffic models, where scheduling decisions of DPS/DPB can be based on the instantaneous cells loading conditions. In this paper we provide detailed system level performance analysis for DPS/DPB CoMP scheme in HetNet scenario with non-full buffer FTP traffic model. The performance results are conducted for different traffic loadings of the cells and DPS/DPB scheduling decision granularities.

Advanced interference suppression receiver for LTE-advanced systems

Co-channel interference is considered as one of the most important issues of modern wireless communication systems. In order to address the interference issue in LTE-A systems, the linear interference rejection combining (IRC) receiver was introduced. Inter-cell interference suppression in the IRC receiver is provided via simple beam forming on the antenna elements of the receiver, which are calculated using the estimated interference covariance matrix. During the related LTE Rel-11 study item in the 3GPP RAN1 Working Group, the IRC receiver was identified as a promising mechanism for improving the cell-edge user throughput. However, for interfering links employing MIMO schemes with multiple spatial layers or in the case of more than one dominant interfering node, the interference suppression capability of the linear IRC receivers is substantially reduced. In this paper we investigate the performance of advanced non-linear IRC receiver structures for interference suppression. More specifically, we propose a maximum-likelihood detection (MLD) scheme to suppress more than one dominant interfering links. The simulation results show that the proposed MLD receiver structure, employing additional knowledge about the modulation scheme of the interfering layers, improves the cell-edge user throughput (defined as the 5th percentile value of the cumulative distribution function) by approximately 23% compared to the baseline linear IRC receiver.

Partially adaptive arrays application for MU-MIMO mode in a MmWave small cells

In this paper we provide detailed system level performance analysis of MU-MIMO mode in the millimeter wave (mmWave) small cells (Wi-Fi hotspots) environment. Traditional way of MIMO implementation assumes a single RF chain per antenna element, with all spatial processing done in the baseband. To overcome high pathloss in the mmWave bands, the large-aperture, very high gain arrays with a large number of elements (several hundreds) are required. Therefore we introduce partially adaptive arrays with reduced number of degrees of freedom which are implemented based on the modular antenna arrays (MAA) architecture. The scalable practical design of MAA allows creation of large-aperture, high-power antenna arrays with reduced number of RF chains at cost of decreasing the adaptability. Employing recently proposed quasi-deterministic (Q-D) model for millimeter-wave channels, it was shown that partially adaptive MAA have very small degradation in comparison with ideal fully adaptive array (FAA) in a realistic scenarios. It was shown that application of MU-MIMO mode in a mmWave small cells allows achieving up to 15-30 Gbps total throughput per cell in a multipath environment (university campus scenario) with practical antenna array design.

Analysis of IEEE 802.16m and 3GPP LTE Release 10 technologies by Russian Evaluation Group for IMT-Advanced

The IMT-Advanced standardization process is executed by International Telecommunications Union Radiocommunication Sector (ITU-R) to define requirements to the next 4-th generation mobile communication systems and certify technologies meeting these requirements. This work presents results of the IEEE 802.16m and 3GPP LTE Release 10 technologies evaluation obtained by the Russian Evaluation Group (REG) as part of the ITU-R IMT-Advanced standardization process. The evaluation was done mainly by system level simulations and considered four mandatory test environments of IMT-Advanced: indoor, microcellular, base coverage urban and high speed. Cell spectral efficiency and cell-edge user spectral efficiency were evaluated for downlink and uplink transmissions. It was found out that performance of both IEEE and 3GPP technologies meets the IMT-Advanced requirements for all the considered evaluation scenarios. This conclusion was conveyed by the REG to ITU in the evaluation report submitted in June 2010.

CS/CB CoMP scheme with semi-static data traffic offloading in HetNets

Co-channel interference is considered as one of the most important issues limiting the performance of cellular systems. In LTE-A systems, the Coordinated Multi-Point (CoMP) transmission technology was introduced to effectively manage data traffic at the transmission nodes for reducing co-channel interference. CoMP transmission is realized by exchanging coordination information via backhaul links, typically available between the transmission nodes. However, in practical networks the backhaul links may have capacity constraints hindering the use of CoMP techniques with high-rate information exchange. In scenarios where ideal backhaul link is not available, coordinated scheduling and coordinated beamforming (CS/CB) CoMP schemes can be used. However, the conventional CS/CB CoMP scheme only assumes interference coordination between the transmission nodes which significantly limits the expected performance benefits that can be additionally achieved by using traffic management techniques. In this paper, we propose a CS/CB CoMP scheme with semi-static traffic offloading. Traffic management in the proposed scheme is achieved by using the cell range expansion concept that takes into consideration a limited set of the transmission points that can be simultaneously coordinated to reduce interference. For comparison, the LTE-A time-domain enhanced inter-cell interference coordination (eICIC) scheme with conventional cell range expansion is considered.

A novel combined CSI feedback mechanism to support multi-user MIMO beamforming schemes in TDD-OFDMA systems

In modern Orthogonal Frequency Division Multiple Access (OFDMA) systems Multi-User Multiple-Input Multiple-Output (MU-MIMO) techniques are employed to increase the cell and user throughput without additional bandwidth or transmit power requirements. The increased throughput in the downlink is achieved by simultaneous transmission to multiple mobile stations (MSs) over the same time-frequency resources. MU-MIMO beamforming techniques at the base station (BS) transmitter are used to mitigate interference from one data stream to another. To enable the MU-MIMO beamforming operation in the downlink, channel state information (CSI) feedback from the MS is required. In this paper we propose a novel CSI feedback mechanism which combines two existing CSI feedback schemes to achieve significant system performance improvements in MU-MIMO beamforming for OFDMA systems operating in Time Division Duplex (TDD). The employed CSI reporting schemes are the quantized, codebook-based feedback scheme and the sounding-based feedback scheme. In the first step, the performance of the two feedback schemes is evaluated separately through system-level simulations. It is shown that for accurate calculation of the MU-MIMO beamforming weights, the sounding-based CSI feedback scheme is preferable in deployment scenarios with low noise and interference levels. For the remaining deployment scenarios, the quantized codebook-based reporting mechanism outperforms the sounding-based feedback scheme. In the second step, the proposed combination of the considered feedback mechanisms using a simple feedback selection scheme is evaluated through extensive simulations. The simulation results presented in the paper show that the proposed combined CSI feedback mechanism not only outperforms the existing CSI feedback schemes if used separately but also leads to marginal performance degradation compared to the system performance achieved with prefect CSI knowledge at the BS transmitter.

Prediction Model for Turbo-Coded OFDMA-Systems Employing Rate Matching and HARQ

The availability of an accurate and simple performance prediction model for OFDMA-based mobile broadband communications systems is critical for many practical applications such as link adaptation and link-to-system mapping for system-level performance evaluations. Previous work on performance prediction models has dealt with predefined and fixed modulation and coding rate schemes. However, recent advances in forward error correction coding techniques in present and next generation mobile broadband communications standards render the existing prediction schemes inappropriate due to the variable coding rate property inherent to the design and operation of such systems; the variable coding rate is due to the rate matching and HARQ processes employed in modern wireless communications standards. The conventional prediction models are inappropriate due to the complexity required for calculating the prediction model parameters in the case of variable coding rates. In this paper, we present a novel model for the performance prediction of turbo-coded OFDMA-based systems employing variable coding rate. The proposed model can be used as a link-to-system mapping, a.k.a. physical layer (PHY) abstraction, in system level simulations of next generation mobile broadband standards such as IEEE 802.16m and LTE-Advanced.

Multi-Point Single-User MIMO Transmission Scheme for Communication Systems beyond LTE-Advanced

Downlink Coordinated Multi-Point (CoMP) transmission schemes are often considered as an effective means to minimize the impact of co-channel interference from neighboring transmission point(s) on the performance of cell-edge users. For Long-Term Evolution Advanced (LTE-A), CoMP was defined in Rel 11 and supports three types of transmission schemes: Coordinated Scheduling and Coordinated Beamforming (CS/CB), Joint Transmission (JT) and Dynamic Point Selection (DPS). For JT CoMP with single-user MIMO (SU-MIMO), it is generally assumed that the user equipment (UE) receives each MIMO layer from multiple transmission points which apply joint precoding. In practice, however, such joint precoding is not always desirable from the performance perspective and system constraints. Therefore, in this paper we propose an alternative approach of supporting JT CoMP for SU-MIMO. More specifically, it is shown that for spatial multiplexing SU-MIMO transmission schemes it is beneficial to transmit MIMO layers with per-point precoding independently from the neighboring (i.e., non-serving) point(s) in addition to the MIMO layer(s) transmitted by the serving point. For the proposed multi-point SU-MIMO transmission scheme, an original proportional fair scheduling algorithm is developed to achieve load balancing across the cooperating transmission points. Using system-level simulations it is demonstrated that the considered multi-point SU-MIMO transmission scheme can improve the system performance and, therefore, may be considered as a candidate enhancement for next generation (5G) cellular communication systems beyond LTE-Advanced.