Kiran Kuchi

Performance of PZF and MMSE Receivers in Cellular Networks With Multi-User Spatial Multiplexing

This paper characterizes the performance of cellular networks employing multiple antenna open-loop spatial multiplexing (SM) techniques. We use a stochastic geometric framework to model distance depended inter cell interference. Using this framework, we analyze the coverage and rate using two linear receivers, namely, partial zero-forcing (PZF) and minimum-mean-square-estimation (MMSE) receivers. Analytical expressions are obtained for coverage and rate distribution that are suitable for fast numerical computation. In the case of the PZF receiver, we show that it is not optimal to utilize all the receive antenna for canceling interference. With α as the path loss exponent, Nt transmit antenna, Nr receive antenna, we show that Nt ⌈(1-2/α) (Nr/Nt-1/2)⌉ receive antennas for interference cancellation and the remaining antennas for signal enhancement (array gain). For both PZF and MMSE receivers, we observe that increasing the number of data streams provides an improvement in the mean data rate with diminishing returns. Also transmitting Nr streams is not optimal in terms of the mean sum rate. We observe that increasing the SM rate with a PZF receiver always degrades the cell edge data rate while the performance with MMSE receiver is nearly independent of the SM rate.

Performance Analysis of ML Detection in MIMO Systems with Co-Channel Interference

In this letter, we analyse a receiver technique for MIMO systems with flat fading channel, in which the Co-Channel Interference (CCI) along with white Gaussian noise is modelled as correlated Gaussian noise. This correlated Gaussian noise is whitened and then Maximum likelihood detection (MLD) is performed over the desired users streams. An upper bound on the bit error rate (BER) is derived to understand the performance of this approach. Using the upper bound, we show that the covariance matrix of the interference has to be rank deficient for the BER to reduce with SNR. Monte Carlo simulations are performed to verify the diversity order predicted by the upper bound.

Coverage and rate in cellular networks with multi-user spatial multiplexing

In this paper we consider a multi-user spatial multiplexing (SM) cellular network, where Nt streams are transmitted to Nt users in the cell. Specifically, we obtain the coverage and rate expressions for a system employing zero-forcing (ZF) receiver. Compared to single stream transmission (SST), it is interesting to see that SM degrades the rate for a notable percentage of users. For the case of two and four receiver antennas, the increase in mean rate of SM is modest compared to single stream transmission (SST) while SST provides a gain over SM for cell edge users.

Limiting Behavior of ZF/MMSE Linear Equalizers in Wideband Channels with Frequency Selective Fading

This letter establishes the near optimality of the linear equalizers (LEs) in multiple antenna systems with frequency selective fading. First, we show that a zero-forcing (ZF) LE employing N_r receiver antennas (where N_r>;1), achieves a fixed signal-to-noise-power-ratio (SNR) of Nr-1/N0 in an i.i.d. Rayleigh fading channel with infinite number of taps. The LE sacrifices one degree-of-freedom (DOF) for mitigating the multi-path fading channel and leaves remaining Nr-1 DOF for array gain. The maximum performance loss with respect to the matched filter bound (MFB) is shown to be 10 log (Nr-1/Nr) which becomes small in systems with large antenna arrays. Next, simulation is used to compare the performance of minimum-mean-square-error estimation (MMSE) and ZF receivers. It is shown that the MMSE based LE provides a significant advantage over ZF LE only for single antenna systems and the performance difference between the two methods becomes arbitrarily small with an increase in the number of receiver antennas.

An iterative MIMO-DFE receiver with MLD for uplink SC-FDMA

Single carrier frequency division multiple access (SC-FDMA) based systems such as long term evolution (LTE) uplink experience severe inter symbol interference (ISI) in highly frequency selective channels. The conventional linear receivers such as the linear minimum mean square error(LMMSE) based receiver fail to address this issue, thereby degrading the performance of the system. This paper proposes a new iterative multiple input multiple output decision feedback equalizer (MIMO-DFE) which does not require the estimation of any parameters for canceling ISI. It also derives a closed form expression for the residual noise covariance after equalization which is used for maximum likelihood detection (MLD) across streams. Simulations show that the proposed receiver yields significant performance gains especially in the high signal-to-noise-ratio (SNR) regime with only a modest increase in computational complexity.

MMSE-Prewhitened-MLD Equalizer for MIMO DFT-Precoded-OFDMA

A low-complexity equalizer which uses a combination of space-frequency minimum-mean-square-error-estimation (MMSE) filter and a pre-whitened maximum likelihood detector (MLD) is proposed for discrete Fourier transform precoded orthogonal frequency division multiple accesses (DFT-precoded-OFDMA) systems employing multi-stream spatial multiplexing (SM). We show that this receiver behaves like an optimum MLD in channels with low frequency selectivity (flat fading) and the performance converges to that of MMSE in channels with high frequency selectivity. Further, we analytically characterize the performance of the zero-forcing (ZF) linear equalizer (LE) in an i.i.d. channel with infinite amount of frequency selectivity for the case when the number of receiver antennas Nr is greater than the number of transmitter antennas/streams Nt. The ZF-LE is shown to provide a per-stream post-processing signal-to-noiseratio (SNR) of Nr-Nt/NtN0 for Nr >; Nt. Additionally, simulation is used to compare the bit error rate (BER) performance of ZF and MMSE based receivers.

Exploiting Spatial Interference Alignment and Opportunistic Scheduling in the Downlink of Interference-Limited Systems

In this paper, we analyze the performance of single-stream and multistream spatial multiplexing (SM) systems employing opportunistic scheduling in the presence of interference. In the proposed downlink framework, every active user reports the postprocessing signal-to-interference-plus-noise power ratio (post-SINR) or the receiver-specific mutual information (MI) to its own transmitter using a feedback channel. The combination of scheduling and multiantenna receiver processing leads to substantial interference suppression gain. Specifically, we show that opportunistic scheduling exploits the spatial interference alignment (SIA) property inherent to a multiuser system for effective interference mitigation. We obtain bounds for the outage probability and the sum outage capacity for single-stream and multistream SM employing real or complex encoding for a symmetric interference channel (SIC) model. The techniques considered in this paper are optimal in different operating regimes. We show that the sum outage capacity can be maximized by reducing the SM rate to a value less than the maximum allowed value. The optimal SM rate depends on the number of interferers and the number of available active users. In particular, we show that the generalized multiuser SM (MU SM) method employing real-valued encoding provides a performance that is either comparable or significantly higher than that of MU SM employing complex encoding. A combination of analysis and simulation is used to describe the tradeoff between the multiplexing rate and the sum outage capacity for different antenna configurations.

An Iterative DFE Receiver for MIMO SC-FDMA Uplink

Ever since the adoption of the single carrier frequency division multiple access (SC-FDMA) scheme for the uplink in the Long Term Evolution (LTE) standard, there has been much interest in improved receiver algorithms for the same. Successive interference cancelation (SIC) based turbo receivers (Turbo SIC) have been proposed to tackle the multi-stream interference (MSI) for transmission modes such as spatial multiplexing (SM) and multiuser (MU) multiple input multiple output (MIMO). In addition, decision feedback equalizers (DFE) have been discussed as an alternative to the simple linear minimum mean square error (LMMSE) based receivers to suppress intersymbol interference (ISI) caused by the frequency selective multi-path channel. This paper proposes a DFE receiver variant based on parallel interference cancelation (PIC) with maximum likelihood detection (MLD) and ordered SIC (OSIC) principles to suppress ISI and MSI. Simulations show that the proposed receiver performs better than the existing SIC and DFE based receivers and achieves the matched filter bound (MFB) for both coded and uncoded systems.

A novel RACH mechanism for dense cellular-IoT deployments

Cellular Internet of Things (C-IoT) is one of the emerging 5G technologies. 4G LTE being the most promising cellular technology so far and is a suitor for serving C-IoT devices. In future millions of C-IoT devices will be deployed in the coverage of a single 4G LTE base station. However, the existing random access (RACH) mechanism in 4G LTE is not designed to connect millions of devices. Hence, in this paper, we propose a novel RACH mechanism that allows millions of C-IoT devices to associate with the base station within the existing 4G LTE framework. 3GPP has proposed an extended access barring mechanism to solve this problem for machine type communication (MTC). We compare the performance of the proposed RACH mechanism with the existing 3GPP extended access barring mechanism through analysis. Further, through simulation results, we show that the proposed mechanism is faster and saves power when compared to existing 3GPP extended access barring mechanism under perfect synchronization.

Performance of cloud radio networks with clustering

Cloud radio networks employ coordinated transmission among base stations (BSs) to reduce the interference effects. The practical limitations in implementing coordination results in suboptimal systems with limited performance. In this paper, we analyze the performance of a cloud network with clustering, where geographically close BSs form a clustered cloud. Coverage probability and rate distributions are analyzed using stochastic geometric models. Specifically, we develop approximations for the coverage probability of a typical user in a clustered cloud network with zero-forcing dirty paper coding. The adverse effect of finite clusters on the achievable rate is quantified.