Closed-form approximations of the expected per-terminal signal-to-interference-plus-noise-ratio (SINR) and ergodic sum spectral efficiency of a multiuser multiple-input multiple-output system are presented. Our analysis assumes spatially correlated Ricean fading channels with maximum-ratio combining on the uplink. Unlike previous studies, our model accounts for the presence of unequal correlation matrices, unequal Rice factors, as well as unequal link gains to each terminal. The derived approximations lend themselves to useful insights, special cases and show the combined impact of line-of-sight (LoS) and correlated components. Numerical results show that while unequal correlation matrices enhance the system performance, the presence of strong LoS has an opposite effect. The approximations are general and remain insensitive to changes in the system dimensions, signal-to-noise-ratios, LoS levels, and correlation levels.

In cellular systems, deploying a large number of antennas at the base station (BS), also called massive multiple-input multiple-output (MIMO), can offer a huge improvement in system throughput. This performance gain achieved through coherent beamforming highly depends on having accurate channel state information (CSI) at the BS. Initially, massive MIMO was considered promising only for time division duplexing (TDD) systems because the downlink training overhead in frequency division duplexing (FDD) systems could be large. To reduce the overhead in FDD systems, we propose efficient eigenspace training and precoding. In the downlink training, the prebeamforming matrix is designed based on the spatial fading correlation to probe the channel and minimize the mean squared error (MSE) in channel estimation. In data transmission, the precoding matrix is designed based on instantaneous CSI to manage the inter-user interference. Our algorithms achieve significant savings in the downlink training because the overhead is associated with the dimension of prebeamforming matrix and no longer limited by the number of BS antennas. Compared with joint spatial division and multiplexing, our algorithms offer much lower MSE and, under some conditions, higher sum rates.

Analytical expressions to approximate the expected per-user signal-to-interference-plus-noise-ratio (SINR) and ergodic sum-rate of a multiuser multiple-input-multiple-output system are presented. Our analysis assumes uncorrelated Ricean fading channels with regularized zero-forcing precoding on the downlink. The derived expressions are averaged with respect to the previously unknown arbitrary eigenvalue densities of the complex non-central Wishart distributed channel correlation matrix. To aid the derivation of the expected SINR, we derive analytical expressions for the joint density of two arbitrary eigenvalues of the complex non-central Wishart matrix. Unlike previous studies, our model caters to the presence of a unique Rice factor for each user terminal, making it suitable for analysis of modern systems, such as small cells and millimeter-wave. Our findings suggest that while the presence of strong line-of-sight has an adverse effect on the expected SINR and ergodic sum-rates, increasing the variability of Rice factors enhances the peak rate performance of the system. Our analysis can be applied to arbitrary system dimensions and is seen to remain tight across the signal-to-noise-ratio range considered.

A fundamental building block for supporting better utilization of radio spectrum involves predicting the impact that an emitter will have at different geographic locations. To this end, fixed sensors can be deployed to spatially sample the RF environment over an area of interest, with interpolation methods used to infer received power at locations between sensors. This paper describes a radio map interpolation method that exploits the known properties of most path loss models, with the aim of minimizing the RMS errors in predicted dB-power. We show that the results come very close to those for ideal Simple Kriging. Moreover, the method is simpler in terms of real-time computation by the network and it requires no knowledge of the spatial correlation of shadow fading. Our analysis of the method is general, but we exemplify it for a specific network geometry, comprising a grid-like pattern of sensors. We also provide comparisons to other widely used interpolation methods.

An accurate approximation is developed for the distribution of the instantaneous per-terminal signal-to-noise-ratio (SNR) of a downlink multiuser multiple-input multiple-output system with zero-forcing (ZF) precoding. Our analysis assumes a Ricean fading environment, where we show that the SNR at a given terminal is well approximated by the gamma distribution and we derive its parameters. The analysis relies on densities of an arbitrary eigenvalue and a pair of arbitrary eigenvalues of the uncorrelated complex non-standard, noncentral Wishart matrices. Unlike previous studies, we consider microwave and millimeter-wave channel parameters with a unique Rice factor for each terminal. We demonstrate that stronger line-of-sight adversely impacts the ZF SNR, while increasing the Rice factor variability results in higher peak ZF SNR. Our approximations are insensitive to changes in the system dimension and operating SNRs.

We describe three in-field data collection efforts yielding a large database of RSSI values vs. time or distance from vehicles communicating with each other via DSRC. We show several data processing schemes we have devised to develop opportunistic Vehicle-to-Vehicle (V2V) propagation models from such data. The database is limited in several important ways, not least, the presence of a high noise floor that limits the distance over which good modeling is feasible. Another is the presence of interference from multiple active transmitters. Our methodology makes it possible to obtain, despite these limitations, accurate models of median path loss vs. distance, shadow fading, and fast fading caused by multipath. We aim not to develop a new V2V model, but to show the methods enabling such a model to be obtained from in-field RSSI data, without elaborate measurement design and the associated deployment cost. Finally, models based on field data allow for capturing the multiple effects of an increasing number of simultaneous V2V transceivers under typical extreme traffic scenarios.

We propose a method for estimating channel parameters from RSSI measurements and the lost packet count, which can work in the presence of losses due to both interference and signal attenuation below the noise floor. This is especially important in the wireless networks, such as vehicular, where propagation model changes with the density of nodes. The method is based on Stochastic Expectation Maximization, where the received data is modeled as a mixture of distributions (no/low interference and strong interference), incomplete (censored) due to packet losses. The PDFs in the mixture are log-Gamma, according to the commonly accepted model for wireless signal and interference power expressed in dBm. This approach leverages the loss count as additional information, hence outperforming maximum likelihood estimation, which does not use this information (ML-), for a small number of received RSSI samples. Hence, it allows inexpensive on-line channel estimation from ad-hoc collected data. The method also outperforms ML- on uncensored data mixtures, as ML- assumes that samples are from a single-mode PDF.

This paper evaluates and compares the performance of single- user (SU) and multi-user (MU) transmission for downlink multiple-input multiple-output (MIMO) channels in terms of system energy efficiency (EE). We introduce power control algorithms to maximize EE. Specifically, to optimize the power allocation, we consider the problem of EE maximization with the satisfaction of the minimum spectral efficiency (SE) gain. Antenna selection is taken into account to further enhance the EE performance for both SU and MU systems. Our results reveal that we should turn off extra antennas at the transmitter, which are originally used for diversity gain but incur large circuit power consumption. Our comparisons between SU and MU differ from conventional comparisons, which focus on SE. Jointly considering EE and SE, we show that SU is more desirable when the transmit power is low, while MU is favored in the case of high transmit power.

We describe three in-field data collection efforts yielding a large database of RSSI values vs. time or distance from vehicles communicating with each other via DSRC. We show several data processing schemes we have devised to develop Vehicle-to-Vehicle (V2V) propagation models from such data. The database is limited in several important ways, not least, the presence of a high noise floor that limits the distance over which good modeling is feasible. Another is the presence of interference from multiple active transmitters. Our methodology makes it possible to obtain, despite these limitations, accurate models of median path loss vs. distance, shadow fading, and fast fading caused by multipath. We aim not to develop a new V2V model, but to show the methods enabling such a model to be obtained from in-field RSSI data.

The development of spectrum measurement infrastructure that can produce real-time geographic maps of spectrum usage is important to the deployment of future wireless systems. Such an infrastructure also provides the basis for creating a spectrum database in support of dynamic spectrum access. We are developing algorithms that can aggregate, classify and geographically map collected RF power measurements so as to support the formation of the spectrum database. A fundamental building block to creating this database is using a collection of spectrum measurements to infer the expected power levels at locations where there was no measurement infrastructure. This project is currently developing a pathloss-based interpolation scheme that can accurately estimate the power levels at locations bounded by four spectrum sensors deployed in a rectangular pattern.

In emulating a multi-node wireless network, received interference can be represented by combining the multipath responses of the interfering links. Each multipath response can be described by a set of mean-squared amplitude of the multipath components and relative delays. The number of filter taps required per link to emulate the actual (`true') channel is a function of the channel bandwidth W and RMS delay spread τrms. Assuming each per-link channel to have an exponentially decaying power delay profile, this value is about 4Wτrms. We propose to emulate each link using n uniformly-spaced taps of equal mean-square gain. For this case, the required number of taps is only 2Wτrms, while maintaining the important characteristic (i.e., the CDF of total power, taken over the fading) of the true channel. We derive this result analytically and confirm it by simulation. Improving on this 50% reduction in required taps, we further show that the loss in accuracy is significantly low so long as the total number of taps is the order of 16 or more. For large values of Wτrms, this can lead to even more reduction in n and, thus, further limit the cost and complexity of emulators.

This letter evaluates three quantization techniques for downlink multiple-user multiple-input multiple-output (MU-MIMO) systems with limited feedback. The required feedback bits for a specified rate loss are quantified, as well as the complexity for each technique. Furthermore, the net capacity, which incorporates the effect of the overhead, is studied. The analysis and simulation results reveal the advantages and drawbacks of each quantization method and demonstrate under what conditions to use one of them rather than the other.

Prior research has shown that delayed channel state information (CSI) severely degrades the throughput performance of multi-user (MU) multiple-input multiple-output (MIMO) transmission, resulting in a preference for single-user (SU) transmission. This paper investigates how the relative performance of MU-MIMO and SU-MIMO is affected by 1) different Doppler spectra, and 2) high-order prediction methods for channel estimation. Simulation results reveal that the often-used Jakes-Clarke spectrum is the worst case for MU performance. By evaluating the impact of other Doppler spectra and high-order channel prediction on MU-MIMO performance, we demonstrate that MU can tolerate larger CSI delays and retain its theoretical advantage over SU in practice.

Typically, a multipath channel response can be characterized as a sum of Rayleigh-fading ''rays'', each defined by a time delay and a mean-square amplitude. Therefore, the channel response can be largely described by a power delay profile (PDP), which is the set of mean-square ray amplitudes and relative delays. Here, we address the following question: Given an actual (or ''true'') PDP, PDP(τ), which may have many rays, is there a 3-ray (i.e. 3- tap) equivalent response, derivable from PDP(τ), that can be used to accurately estimate the average bit error rate, <;BER>, vs. receiver input signal- to-noise ratio, SNR? The results reported here give an affirmative answer, e.g., for <;BER> values down to = 10-4, the required SNR using a 3-tap equivalent channel response is less than 1.1 dB larger than that required for the ''true'' channel. This agreement can be improved upon, suggesting further work on deriving and evaluating equivalent 3-tap channels. We discuss the benefits of such simplifications for hardware emulators as well as for simulation and analysis.

A novel quantization method, sparse coding quantization (SCQ), is proposed for downlink multiuser multiple-input multiple-output (MU-MIMO) systems. Compared to conventional vector quantization (VQ), SCQ utilizes a linear combination of several codewords rather than a single one to represent the channel matrix. Both analytical and simulation results reveal that the proposed technique can achieve the same sum rate performance as VQ at a reasonable cost in feedback overhead. Thus, SCQ is more practical because it significantly reduces the time and space complexity for generating, searching and storing the codebook. % The net capacity, which indicates the tradeoff between sum rate and overhead, is also studied.

Widely regarded as one of the most promising emerging technologies for driving the future development of wireless communications, cognitive radio has the potential to mitigate the problem of increasing radio spectrum scarcity through dynamic spectrum allocation. Drawing on fundamental elements of information theory, network theory, propagation, optimisation and signal processing, a team of leading experts present a systematic treatment of the core physical and networking principles of cognitive radio and explore key design considerations for the development of new cognitive radio systems. Containing all the underlying principles you need to develop practical applications in cognitive radio, this book is an essential reference for students, researchers and practitioners alike in the field of wireless communications and signal processing.

This chapter develops the fundamental capacity limits and associated transmission techniques for different cognitive radio network paradigms. These limits are based on the premise that the cognitive radios of secondary users are intelligent wireless communication devices that exploit side information about their environment to improve spectrum utilization. This side information typically consists of knowledge about the activity, channels, encoding strategies, and/or transmitted data sequences of the primary users with which the secondary users share the spectrum. Based on the nature of the available side information as well as regulatory constraints on spectrum usage, cognitive radio systems seek to underlay, overlay, or interweave the secondary users' signals with the transmissions of primary users. This chapter develops the fundamental capacity limits for all three cognitive radio paradigms. These capacity limits provide guidelines for the spectral efficiency possible in cognitive radio networks, as well as practical design ideas to optimize performance of such networks.

Stochastic models for path loss are most often of the form PL = A + B log(d), where A and B are empirical constants derived from data via least-squares fitting, and d is the path distance. For in-building environments, however, many investigators have noted the added effects of transmission through interior walls and floors. Here, we represent this effect by a third term, which is exponentially related to log(d), and we model its impact on path loss. The database we use is obtained using a well-validated ray-tracing tool, which we apply to single floors of four very distinct office buildings. The context is an ad hoc wireless network, wherein both the transmit and receive locations can be anywhere on the floor. The resulting model consists of the median path loss, involving three fitting constants A, B and C; and a log-normal variation about the median, with its standard deviation being a function of distance. The structure and details of the model are shown to be remarkably similar across the four distinct buildings studied.