Erfan Majeed

Feedback Generation for CoMP Transmission in Unsynchronized Networks with Timing Offset

Coordinated multipoint (CoMP) transmission is a promising technique in long term evolution (LTE) systems to provide coverage and broadband communication for cell edge user equipments (UEs). However, as signals from multiple transmitters may reach the UE at different times, CoMP networks might experience high time offsets, leading to significant performance loss in closed-loop transmission. In this paper we show that spacing between reference signals (RSs) imposes a phase offset on the transmit covariance matrix in presence of timing offset (TO), affecting the feedback generation. The promised advantages of closed-loop CoMP transmission vanish in presence of TO due to improper precoding matrix index (PMI) selection. Keeping the phase shift close to zero, reliable PMI selection can be guaranteed and as a result near optimum performance is achieved for closed-loop CoMP transmission in unsynchronized networks with TO present.

Cooperative interference mitigation in heterogeneous LTE networks

Small cell transmission is a promised technique to enable the tremendous data growth in next generation long term evolution (LTE) networks. However, emerging inter-cell interference is the dominant limiting factor achieving higher network capacities. In this paper we focus on cooperative interference mitigation techniques in interfering heterogeneous network (HetNet) deployments to assist the cell edge user equipment (UE) suffering from dominant co-channel interference (CCI). As a handover might be refused and conventional precoding matrix index (PMI) selection schemes limit reliable downlink (DL) transmission, the co-scheduled UEs being in good channel condition delegate the precoding vector selection to that victim UE. We provide a practical implementation for the cooperative interference mitigation algorithm keeping computational complexity and feedback overhead low. With numerical results at the link level we show that CCI is almost suppressed and as a result near optimum DL performance is achieved at the cell edge UE in strong interfering HetNet deployments.

Robust FFT Window Replacement in Non-Ideal CoMP Networks with Timing Offset

Coordinated multipoint (CoMP) transmission is one of key technologies to provide broadband communication and coverage for cell edge user equipments (UEs) in long term evolution (LTE) systems. The basic concept behind CoMP transmission are geographically distributed transmitters working jointly in coordinated manner. As the transmit signals reach the UE at different times, CoMP networks might experience high timing offsets (TOs) leading to a fast Fourier transform (FFT) window mismatch. Positive TOs with synchronization signals reaching the UE prior to data signals impose an acceptable performance loss, whereas the presence of negative TOs disables CoMP transmission saturating in error floor at unacceptable bit error rate. In this paper a novel approach to overcome negative TO is proposed that guaranties robust performance while keeping system complexity low. Shifting the FFT window towards the data signal by the estimated TO, the negative TO turns positive and as a result near optimum performance is achieved.

Advanced receiver design for interfering small cell deployments in LTE networks

Small cells are low power nodes, which have lower transmit power than the macro eNodeB. They are regarded as one of the important solutions in long term evolution advanced (LTE-A) to deal with the dramatically increased mobile data traffic. Small cells deployments in LTE-A release 12 with 4 small cells in a cluster suffer from stronger inter-cell interference than heterogeneous network (HetNet) in the previous release of LTE-A, where one small cell exists in a cluster. We investigate the joint least squares channel estimation (JLSCE) over demodulation reference signals (DMRS) spreading in time and frequency resources in LTE-A systems. Furthermore, different receiver structures using least square channel estimation (LSCE) and JLSCE are investigated in small cell deployments with the presence of strong inter-cell interference. Our results show the linear minimum mean square error (LMMSE)-interference rejection combining (IRC) with LSCE is the most efficient receiver structure compared to the enhanced (e)LMMSE-IRC with JLSCE, interference cancellation (IC) with eLMMSE-IRC and maximum ratio combining (MRC) receiver structures for signal detection in small cell deployments.

On the performance of CoMP transmission in unsynchronized networks with timing offset

Coordinated Multipoint (CoMP) transmission is a promising technology in 4G Long Term Evolution (LTE) systems to enable coverage and high data rates for cell edge User Equipments (UEs). However, the presence of Timing Offset (TO) between the received signals from multiple transmitters at the UE could prove prohibitive for CoMP transmission. Such TO affects the estimation of the transmit covariance matrix for Pre-coding Matrix Index (PMI) selection significantly. Especially, the promised advantages of closed-loop CoMP transmission vanishes and it performs even worse than the open-loop in presence of a large TO due to the improper PMI selection. With analytical and numerical results we show that keeping the TO small, the promised advantages of closed-loop CoMP transmission can be guaranteed and as a result near optimum performance is achieved for CoMP transmission in unsynchronized networks with TO present.

Reliable implicit feedback generation in unsynchronized CoMP transmission

Coordinated multipoint (CoMP) transmission is a key technique in 4G long term evolution (LTE) systems to enable broadband communication and coverage for user equipments (UEs) at the cell edge. However, as the transmit signals will reach the UE at several and varying delays, unsynchronized CoMP transmission might experience large timing offsets (TOs), imposing a significant performance loss. In this paper we focus on the impact of implicit feedback generation in closed-loop CoMP transmission operating in joint processing mode. To enable the expected advantages, reliable channel state information is crucial for practical realization of closed-loop CoMP transmission. We show that a spacing in terms of subcarriers between the reference signals in the orthogonal frequency division multiplexing symbols imposes a phase offset on the transmit covariance matrix in presence of TO, affecting the feedback generation and limiting the expected performance. We provide a solution to overcome this limitation by estimating and compensating this phase offset. We show analytically and numerically that keeping the phase offset close to zero, reliable feedback generation can be guaranteed and, as a result, near optimum performance is achieved in closed-loop CoMP transmission in unsynchronized networks with TO present.

On the performance of advanced receivers in unsynchronized small cell LTE networks

Next generation long term evolution (LTE) systems evolve towards small cell networks to boost spectrum efficiency through extreme frequency reuse. To mitigate the impact of the resulting tremendous co-channel interference, advanced receivers supporting network assisted interference cancellation and suppression (NAICS) represent a promised technique to meet the ubiquitous cell edge free experience. However, as the interfering signals will reach the cell edge user equipment (UE) at varying delays, unsynchronized small cell networks might experience large timing offsets (TOs) and considerably limit the expected downlink (DL) performance. In this paper we focus on reliable channel estimation for diverse types of NAICS receivers operating in rich interfering small cell LTE networks. With analytical results we derive the impact of TOs on channel estimation, show that the promised performance of each of the NAICS receivers vanishes when neglecting the varying delays and against expectations perform worse than conventional linear interference suppression receivers. However, estimating and compensating the impact of the TOs, reliable channel estimation can be achieved and as a result the expected DL performance of each of the investigated NAICS receivers can be guaranteed in unsynchronized small cell LTE networks.

Reliable feedback generation in unsynchronized joint-processing CoMP transmission networks

Coordinated multipoint (CoMP) transmission is a key technique in 4G long term evolution (LTE) systems to provide the ubiquitous broadband communication with cell edge free experience at the user equipment (UE). However, as the transmit signals will reach the UE at several and varying delays, practical CoMP networks might experience large timing offsets (TOs), imposing a significant performance loss. In this paper we focus on implicit feedback generation for practical CoMP joint processing (JP)-joint transmission (JT) networks suffering from TOs. We show that conventional implicit feedback schemes prohibit the expected gain in the downlink, getting outperformed by open-loop transmission schemes and call the benefits of channel state information (CSI) feedback for CoMP JP-JT networks into question. Analytically we prove that the mapping of CSI-reference signal (RS) on different subcarriers imposes multiple phase offsets on the covariance matrix, leading to improper precoding vector selection, while imbalanced TOs among the transmitters limit coherent signal demodulation at the UE. With numerical results we show that estimating and compensating the phase offsets reliable precoding vector selection and phase feedback for coherent signal demodulation can be obtained and as a result considerable improvements for practical CoMP JP-JT networks can be achieved in presence of varying TOs.

Impact of Antenna Configuration on Feedback Generation for Non-Ideal CoMP Transmission

Coordinated multipoint (CoMP) transmission is a promised candidate in long term evolution advanced (LTEA) systems to assure coverage and broadband communication for cell edge user equipments (UEs). However, the presence of timing offsets (TOs) between the received signals from multiple transmitters might limit the performance of CoMP transmission. The promised advantages of closed-loop CoMP transmission vanish for the system configuration with four transmit antennas, showing significantly poorer performance than the transmitter equipped with two antennas for large TOs. The reason is that the transmit covariance matrix used for implicit feedback generation is affected by a phase shift caused by TO after orthogonal frequency division multiplexing demodulation, leading to improper precoding matrix index (PMI) selection and therefore limiting the beamforming gain at the cell edge UE. On the other hand the system configuration with two transmit antennas is robust against that phase shift, despite the TO. Keeping the phase shift in systems with four transmit antennas low, reliable PMI selection can be guaranteed and as a result the promised CoMP transmission benefits can be achieved in a practical realization with TO present.

Cooperative implicit feedback generation in interfering heterogeneous LTE networks

Next generation long term evolution (LTE) systems evolve towards small cell transmission to enable the tremendous data growth. However, emerging interference is the dominant limiting factor achieving higher network capacities. In this paper we focus on cooperative implicit feedback generation in interfering heterogeneous networks (HetNets) to assist the cell edge user equipment (UE) suffering from dominant co-channel inter-cell interference. As a hand-over might be refused and conventional precoding matrix index (PMI) selection schemes limit reliable downlink (DL) transmission, the co-scheduled UEs delegate the PMI selection to that victim UE, trimming the codebook. We provide a practical implementation for the cooperative PMI selection, keeping computational complexity and feedback overhead low. With numerical results at the link level we show that the suggested cooperative PMI selection algorithm mitigates co-channel inter-cell interference and as a result enables reliable DL transmission at the victim UE.