This paper presents a novel framework for object-oriented integrated sensing and communications (ISAC) channel modeling in urban environments over low-altitude 3D space, which jointly considers unmanned aerial vehicle (UAV)-based and ground-level measurements in conjunction with ray-tracing (RT) technology. Two types of object-oriented channels are considered, namely the static background channel, e.g., buildings, and the dynamic channel, e.g., cars and UAVs. Firstly, a systematic measurement campaign was conducted in the 4 GHz band at 80 m height using a UAV and at ground level using a vehicle. Additionally, photographs of the measurement area are taken by the UAV for an object-oriented 3D model. Based upon channel measurement data and employing RT techniques, the EM parameters of the considered coverage area are accurately calibrated and validated at both heights. This has allowed us to obtain accurate simulations across the vertical airspace of both sensing- and communication-background channels, followed by a comprehensive channel characteristics analysis. By subtracting the background channel from the dynamic channel, experimental results have demonstrated that the angle and distance information of objects can be accurately estimated. Through a case study, it has been shown that the acquired prior scene information can be utilized as a reference to improve object-oriented coverage.

In this work, we present a recent investigation on leveraging large reconfigurable intelligent surfaces (RISs) as anchors for positioning in wireless communication systems. Unlike existing approaches, we explicitly address the uncertainty arising from the substantial physical size of the RIS—particularly relevant when a user equipment (UE) resides in the near field—and propose a method that ensures accurate positioning under these conditions. We derive the corresponding Cramér-Rao bound (CRB) for our scheme and validate the effectiveness of our scheme through numerical experiments, highlighting both the feasibility and potential of our approach.

Next-generation wireless communication systems will incorporate millimeter wave (mmWave) as one of the key technologies to increase data rates. These forthcoming mmWave systems will integrate multiple-input multiple-output (MIMO) technology to ensure sufficient link margin and are expected to be deployed in conjunction with sub-6 GHz systems. Configuring a MIMO communication link usually relies on estimating channel state information (CSI), which is difficult to acquire at mmWave frequencies due to the low pre-beamforming signal-to-noise ratio (SNR). In this paper, we propose a novel approach to estimate mmWave MIMO channels by leveraging out-of-band information from a sub-6GHz band. Utilizing this approach, we develop three channel estimation methods and compare these methods with one using only in-band information. We investigate the influence of theK-factor, the SNR and the number of antennas on the performance of the proposed channel estimation methods through simulations. Additionally, we compare these methods in terms of their computational complexity. Finally, we validate the performance of the proposed methods through channel measurements conducted in an outdoor environment. The results demonstrate that the proposed methods outperform in-band mmWave channel estimation in terms of spectral efficiency, particularly in scenarios of low SNR and high K-factor.

Ray tracing methods are gaining popularity for modeling the propagation conditions of wireless links. This adds them to a series of propagation models for wireless channels, making the choice of the appropriate model difficult. This paper evaluates the reliability of several large-scale fading models by comparison with reference signal received power (RSRP) measurements obtained through a drive test in a low-rise urban environment in the 1.855 GHz band. The models integrate different levels of information regarding the link environment. The results show that all models underestimate the propagation loss and that an increase in available information about the environment does not always enhance the accuracy of a model. For system-level simulation, a statistical approach shows the best results, especially in absence of reliable environmental data.

In this work, we present a wireless localization method that operates on self-supervised and unlabeled channel estimates. Our self-supervising method learns general-purpose channel features robust to fading and system impairments. Learned representations are easily transferable to new environments and ready to use for other wireless downstream tasks. To the best of our knowledge, the proposed method is the first joint-embedding self-supervised approach to forsake the dependency on contrastive channel estimates. Our approach outperforms fully-supervised techniques in small data regimes under fine-tuning and, in some cases, linear evaluation. We assess the performance in centralized and distributed massive multiple-input multiple-output (MIMO) systems for multiple datasets. Moreover, our method works indoors and outdoors without additional assumptions or design changes.

The canonical setup of cell-free massive multiple-input multiple-output (MIMO), where all the access points (APs) serve all the users, does not scale well. In this work, we propose a deep reinforcement learning (DRL) approach to user-centric clustering in which each user is served by only a subset of APs. The clusters are formed such that either a given user demand is satisfied or the network sum rate is maximized. Unlike previous studies, we allow the clusters to vary in size depending on the propagation conditions. We design our DRL framework to be flexible enough to accommodate different performance targets in terms of the sum spectral efficiency, fronthaul capacity and power consumption. By optimizing the AP selection for each user, our proposed scheme is able to achieve the same performance as the canonical setup (upper bound) with significantly lower fronthaul requirements.

In this paper, we consider an access point clustering-based pilot assignment scheme for user-centric cell-free massive MIMO systems. In such a system, the complexity is reduced by performing the pilot assignment for each cluster independently. However, users at the edge of the cluster will experience high interference from users in neighboring clusters. In this paper, we propose two schemes to reduce pilot contamination for cluster-edge users. The proposed schemes can perform per-group pilot assignment with zero or limited inter-cluster cooperation. Numerical results verify that, compared with conventional pilot assignment schemes, the proposed pilot assignment schemes achieve similar performance while having lower complexity and allowing for decentralized processing.

Train commuters use smart devices that are continuously connected to the Internet via cellular networks. As such, end users’ service perception depends on the interruption-free delivery that access networks can provide. Setting up a wireless reception inside the metal cabin of a train is a challenging task; more so when travelling in rural areas, where mobile coverage is limited, at high speeds. In the past decade, several technical advances, e.g., active repeaters in trains and along tracks, have been deployed to improve cellular coverage. The resulting overall performance of these active elements can be evaluated only through measurement. Thus, we present a methodology that can directly evaluate under different scenarios the performance gain that an in-train repeater solution delivers. The method delivers synchronous results for both cases of with and without in-train repeater. This then provides the means to drill down to specific cases at high spatial granularity, e.g., the impact of different cellular deployments. In this paper, we validate the methodology using extensive laboratory measurements, and then present the results of our real-world drive tests. Based on the results obtained from the real-world measurement, we can highlight the strengths and shortcomings of an in-train repeater system. Our results show that the static evaluation of the carriage can only partially reflect the scenarios observed in the dynamic measurement on the railroad track. The new method enables swift and resourceful measurement campaigns for the design, optimization, and maintenance of an in-train repeater system.

Overcoming the Vehicle Penetration Loss (VPL), introduced by the thermal insulation of windows, is a critical factor for providing seamless connectivity in high-speed mobility scenarios. Several frequency-selective surface solutions for VPL reduction exist on the market. In this paper, we introduce a framework for analyzing the VPL and its dependency on the Angle of Arrival (AoA) utilizing cellular network measurements within real-world vehicular scenarios. Through the fusion of interpolated outdoor and indoor Received Signal Reference Power (RSRP) measurements collected along the track, the high-resolution VPL profile is derived. Exploiting the knowledge of the network and environment layout, the AoA at each measurement location is determined, enabling us to analyze the VPL’s AoA dependency.Built solely upon mobile network measurements and geospatial data, the suggested framework is well-suited for extensive crowd-sourced data utilization. It also seamlessly applies to various vehicular contexts, encompassing trains and cars alike. Operating across all frequencies, the framework’s scope expands beyond our LTE focus to encompass future generations such as 5G, 6G, and beyond.The presented approach is validated through a comprehensive real-world measurement campaign conducted on a moving train. The evaluation on a per-band basis offers additional insights into the frequency dependency. These findings provide valuable guidance for optimizing the positioning of BSs for track-side deployment, as well as comparing different VPL reduction solutions.

We consider a user-centric cell-free massive MIMO system in which many distributed access points are connected to a central processing unit to serve multiple users. In practice, linear precoding schemes can achieve high spectral efficiency in massive MIMO systems. However, for cell-free massive MIMO, these methods suffer from a large training overhead and high computational complexity. We reduce the complexity of the zero-forcing precoder and improve the performance of the maximum ratio transmission precoder by dividing the users into groups and assigning different frequency resources to each group. Further, we reduce the training overhead by assigning highly interfering users to different groups. We propose a graph coloringbased user grouping scheme with an adjustable group size. Numerical results show that the proposed scheme can achieve about 96% spectral efficiency of zero-forcing while requiring only 32% processing complexity of zero-forcing. Further, the proposed scheme improves the performance of maximum ratio transmission by 24%. Finally, we investigate the impact of the coherence block length.

Distinguishing between reconfigurable intelligent surface (RIS) assisted paths and non-line-of-sight (NLOS) paths is a fundamental problem for RIS-assisted integrated sensing and communication. In this work, we propose a pattern alternation scheme for the RIS response that uses part of the RIS as a dynamic part to modulate the estimated channel power, which can considerably help the user equipments (UEs) to identify the RIS-assisted paths. Under such a dynamic setup, we formulate the detection framework for a single UE, where we develop a statistical model of the estimated channel power, allowing us to analytically evaluate the performance of the system. We investigate our method under two critical factors: the number of RIS elements allocated for the dynamic part and the allocation of RIS elements among different users. Simulation results verify the accuracy of our analysis.

Recently, there has been increased interest in using network traces from crowdsourcing for anomaly detection or operator benchmarking. Likewise, the inference of the network topology, particularly the localization of basestations, has been an active field of research. In this work, we conduct a large-scale analysis of a probabilistic basestation localization technique operating on crowdsourced Timing Advance measurements collected through an Android application. We evaluate the performance using ground-truth locations for 4,946 LTE cells, covering a representative subset of the infrastructure of an Austrian operator. Careful analysis confirms the well-known quality issues of crowdsourced data, with 5.8% of samples effectively reporting negative multipath components. Incorporating the notion of erroneous measurements into our system model ensures that the presence of such outliers is reflected in the posterior confidence of the estimator. Overall, this adapted approach significantly outperforms the baseline estimators from the literature, achieving an accuracy of above 97.5% in a cell-identification framework and mean distance errors of around 100 m. Deviations from multipath and erroneous samples remain the primary error source, given that for a low-multipath subset, the performance can be further improved to 99% and 50 m, respectively.

In this paper, we consider the problem of activity detection in grant-free code-domain non-orthogonal multiple access (NOMA). We focus on performing activity detection via subspace methods under a setup where the data and pilot spreading signatures are of different lengths, and consider a realistic frame-structure similar to existing mobile networks. We investigate the impact of channel correlation on the activity detection performance; first, we consider the case where the channel exhibits high correlation in time and frequency and show how it can heavily deteriorate the performance. To tackle that, we propose to apply user-specific masking sequences overlaid on top of the pilot signatures. Second, we consider the other extreme with the channel being highly selective, and show that it can also negatively impact the performance. We investigate possible pilots’ reallocation strategies that can help reduce its impact.

Cellular networks are the information backbone of nomadic societies commuting between different places. Providing reliable connections for commuters is one of the main challenges for today’s railroad operators, who deploy different solutions to their railroad cars. In this paper, we present a benchmarking methodology for comparing different solutions in an operational network under the regular operation of the cabin. We avoid confounding in the result by conducting spatial interpolation and mapping onto a reference path combined with path segmentation. We further cluster the data according to landscape class to make the results more comparable. To validate our approach, we conducted a test run along the 490km Vienna-Innsbruck rail while onboard a prototypical empty train. The train under the test provided three differently configured cabins regarding wireless communications, namely, standard windows, the first generation of modified windows, and a repeater deployment. We placed two phones in identical positions inside each cabin and compared the measurement results with the data collected using an outside antenna. The statistical analysis of the LTE signal data collected while the train was running along the track shows that the data collected from different test runs can be aggregated even when the train runs in opposite directions. Further, the statistical analysis of the per-seat measurements reveals that the measurements at different positions can be accumulated. The vehicle penetration loss can passively be tuned by modifying the window coating. The passive solutions do not change the overall distribution and differ only by the delta of the window attenuation in the observed scenario. This simplifies future refurbishment planning, as improvements can be deduced from current measurements. The performance of the active solution, which on average outperforms the passive one, is limited by the maximum output power of the chosen repeater system. In areas with excellent coverage, the passive solution outperforms the active one. Future developments in coating modifications might enable similar performance in areas of moderate coverage.

The accurate and reliable localization of transmitter locations from crowdsourced measurements has enabled the large-scale analysis of the previously hidden network layout. Recent work has shown that signal-strength measurements from drive-test campaigns also unveil the antenna orientations - pro-viding an open-source network twin that can act as the backbone of coarse user-equipment positioning, operator benchmarking, or the generation of coverage and performance maps. In this work, we extent this drive-test based scheme to the regime of large-scale crowdsourced datasets and conduct an assessment of orientation inference on 4,950 sectors from a live LTE network. Fusing signal-strength and geometry-based features in a probabilistic Deep Learning image processing framework tackles the challenging characteristics of such noisy crowdsourced data collected in uncontrolled conditions. We further use transfer learning and weight sharing to extend our approach to allow for joint inference of sectors mounted onto the same base station. On the test dataset - representative of a complete network deployment - our selective predictors achieve median errors as low as 7.3° with 95 percentiles below 21°.

Non-orthogonal multiple access (NOMA) transmission techniques promise reduced latency, massive user deployments, and higher data rates. To investigate how these per-formance enhancements translate to practical deployments in a network, a reliable system-level abstraction for downlink two-user power domain NOMA with successive interference cancellation at the receiver is needed. A simple and efficient abstraction is described and then validated through comparison with link-level simulation results. With the developed abstraction, the effects of the deployment of NOMA techniques in a wireless network with a large number of machine type communication users are evaluated in terms of cell throughput, number of users served simultaneously, and fairness.

We propose novel tree-structured quantization approaches for points on Grassmann and Stiefel manifolds. Such manifold quantizers find application, e.g., for channel state information quantization in limited feedback multiple-input multiple-output wireless communications. The proposed quantizers allow to trade-off quantization distortion for complexity by pruning the width of the quantization tree. We compare the rate-distortion performance and quantization complexity of the proposed quantizers to existing single stage and recursive multi stage quantizers, showing that the proposed methods lie in-between these two extremes.

Vehicular communications has steadily grown in the last years playing a key role in the development of road safety applications, assisting users to improve traffic efficiency, or providing infotainment systems. There are two main Radio access technologies (RATs) that support vehicular communications: the Dedicated Short Range Communication (DSRC) and the Cellular vehicle-to-everything (C-V2X) standard. Among its use cases, the C-V2X technology allows communication between users, that is, vehicle-to-vehicle (V2V) communication, but also the vehicle-to-infrastructure (V2I) communication. In our work, we focus on enhancing the quality of the V2I link through V2V communications. We use idle users to boost the signal coming from the Base Station (BS) and therefore to improve the coverage of the overall network. Furthermore, we analyze a scenario where both standards coexist: the cellular users enhance the signal from the BS while DSRC users act as interferes in the relay-assisted link. To model the interference of the DSRC users, we consider both perfect and imperfect Carrier Sense Multiple Access (CSMA). Our scenario is a 2-dimensional (2D) Manhattan grid where BSs, vehicular users, and streets are placed randomly. We provide analytical expressions of the coverage probability for the direct link and the individual parts of the relay-assisted link by leveraging tools from stochastic geometry (SG). We compare the analytical results to Monte Carlo system-level simulations in order to validate our model.

Fixed point-to-point microwave links (P-P links) are widely used for vital backhaul connections in cellular networks. One frequency range allocated to P-P links by the International Telecommunications Union ranges from 24.25GHz to 29.5GHz. The same frequency range was also allocated to mobile services, with bands dedicated to 5G new radio (NR). Potential outages or performance drops in P-P links due to interference from 5G NR are therefore a concern to mobile network operators. We characterize the link-level performance of a commercial off-the-shelf (COTS) P-P link experimentally through measurements. We install a COTS P-P link on our campus. Through an additional experimental transmitter, we generate a 5G NR interference signal, allowing us to measure the link performance depending on signal-to-interference ratio (SIR) in the field. We measure the data rate that the P-P link achieves with respect to the interferer’s location and transmit power. We establish a relation between the P-P link’s SIR and data rate based on our measurements. We further measure the P-P link’s antenna pattern in an anechoic chamber. To investigate the impact of interfering 5G NR user equipments (UEs) in typical cellular network deployment, we perform system-level simulations based on our link-level measurements. We simulate the impact of 5G NR users on a P-P links backhaul connection in typical cell geometries. We identify the cell’s key parameter that determines the P-P link’s achievable data rate.

Providing accurate estimates for Key Performance Indicators (KPIs) of mobile cellular networks is crucial to fulfill agreed-upon service requirements. Typically, such performance maps are based on measurements, which are expensive and only offer sparse samples of the underlying ground-truth. Therefore, we require interpolation schemes to further increase the spatial resolution in the area of interest. In dense urban areas, the propagation environment is often governed by tunneling effects along streets and blockages introduced by buildings. However, these aspects are typically not accounted for in state-of-the-art interpolation approaches. As such, we propose a propagation-aware interpolation scheme based on Gaussian Process Regression (GPR), which utilizes available geospatial information via the diffusion kernel. Based on simulations in a stochastic Manhattan Grid (MG) environment, we show that our approach is well equipped to handle environments dominated by tunneling effects and significantly outperforms the omnidirectional Radial Basis Function (RBF) kernel. Further, the proposed graph-based scheme offers a high degree of flexibility in modeling the environment. Therefore, it is not specific to the MG scenario but can easily be extended to generic blockages and real-world street layouts. We show this by applying our approach to a set of Reference Signal Received Power (RSRP) measurements from an operational LTE network collected in a drive-test campaign in Vienna, Austria — again resulting in an error reduction.

Access to reliable public data about the networking topology in a given area is an essential requirement for applications such as end-user localization; likewise, the subject is a significant area of research. For one, results from eNodeB localization can act as ground truth data for system-level simulations; for another, they can benefit the evaluation of advanced pathloss models. Although the orientation of a given sector antenna plays a crucial role for such use-cases, this aspect of network topology inference has not been prominently addressed in the literature. As such, we propose a Bayesian scheme operating on end-user Reference Signal Received Power measurements from live LTE networks, to infer the orientation of the serving sector antenna. Hereby, we rely on a state-of-the-art 3GPP system model, with the sector orientation as the primary free parameter. The stochastic treatment allows us to directly incorporate the shadow-fading statistics into the estimation scheme, thus actively addressing the main limitations of Received Signal Strength-based approaches. We examine our approach on a set of simulations before validating it on an extensive drive-test data set collected in Vienna, Austria, with ground truth orientations confirmed by a major Austrian Mobile Network Operator. On this data set, our shadow-fading aware approach achieves a median absolute error of$18^{\circ}$, which can be further reduced to$11^{\circ}$by incorporating the generic assumption of non-overlapping sectors. This shows, that end-user measurements are a suitable basis to expose this aspect of the network topology to researchers and industry observers.

The deployment of large antenna arrays at the base station is a prerequisite for operating at the millimeter wave bands. The large antenna arrays are required to achieve high beamforming gains and thus to guarantee sufficient received signal power. In massive multiple input multiple output systems hybrid precoding is utilized to reduce power consumption and cost. Conventional precoding schemes require full channel knowledge at the transmitter, which is hard to obtain, especially in frequency division duplex systems. In this paper, we propose a novel hybrid precoding scheme, which requires only the information acquired during the initial access phase. Furthermore, we derive an analytical expression for characterizing the beam detection capabilities of the system.

In this paper, we investigate the outage performance of an intelligent reflecting surface (IRS)-assisted non-orthogonal multiple access (NOMA) uplink, in which a group of the surface reflecting elements are configured to boost the signal of one of the user equipments (UEs), while the remaining elements are used to boost the other UE. By approximating the received powers as Gamma random variables, tractable expressions for the outage probability under NOMA interference cancellation are obtained. We evaluate the outage over different splits of the elements and varying pathloss differences between the two UEs. The analysis shows that for small pathloss differences, the split should be chosen such that most of the IRS elements are configured to boost the stronger UE, while for large pathloss differences, it is more beneficial to boost the weaker UE. Finally, we investigate a robust selection of the elements’ split under the criterion of minimizing the maximum outage between the two UEs.

The evaluation of mobile network performance is based on real-world measurement data. This data originates from different sources, such as drive-test, drone-measurements, and crowdsourced data. As measurements are not available at all locations, spatial interpolation is necessary to estimate spatial service coverage. Gaussian Process Regression (GPR) presents itself as a useful tool for performance metrics map reconstruction and for delivering prediction quality with it, allowing network operators to determine areas in which new measurements, e.g., drive-tests, will be useful. However, it comes at the cost of low scalability with growing data sets, expected with crowdsourced data. We aim to limit the computational effort in the GPR prediction resulting from updates in the data set, by measurement clustering and averaging while reducing the measurement and Global Positioning System (GPS) location noise. We investigate two scenarios with different measurement distributions and the influence of cluster identification, as required for real data measurements. Based on a desired error of the GPR prediction, we can determine the number of clusters required in the area of interest and the number of points needed inside each cluster for sufficient noise reduction.

Abstract Issues of energy efficiency, latency and reliability have engaged researchers' attention in wireless communications for a long time due to their importance in the dependable delivery of data over wireless networks. The emergence of M2M communication and IoT has intensified this attention because energy efficiency and latency are critical factors affecting their performance. In this paper, a scheduler is designed by utilising the probability density function of the signal‐to‐noise ratio of Rayleigh fading channels to define a threshold used for resource allocation. This threshold, combined with the mean SNR of an M2M device, determines whether or not an M2M device is eligible for scheduling, given its instantaneous channel conditions. The trade‐off between energy efficiency, latency and packet drop of this proposed scheduling strategy is investigated. The performance of the proposed scheduler is compared to round robin, maximum throughput, and proportional fair schedulers. Compared to these standard scheduling strategies, the scheduler provides the advantage of trading off latency for energy efficiency by tuning the threshold parameter.

In this paper, we discuss the temporal behavior of several key parameter indicators (KPIs) in LTE downlink (DL) based on multiple long-term static measurement campaigns conducted under various conditions, including rural and urban indoor and outdoor scenarios. We publish part of the measurements as an open dataset. We conclude that the reference signal received power (RSRP) does not show any time-of-day patterns and thus can be modeled as time-invariant for limited periods of time (from several hours to multiple days) followed by occasional anomalies in the form of jumps of several dB. The other KPIs, for example, throughput, exhibit repeated diurnal patterns with significant differences between peak hours (full cell in the afternoon and evening) and off-peak hours (empty cell in the early morning). Even under model idealizations, such KPIs should be thus modeled at least as cyclostationary random processes. We are not aware of any other paper that would provide static LTE DL measurement data covering several weeks and would thus allow the scientific community to study the diurnal patterns without a need to perform own measurements.

Open datasets with measurements in cellular mobile networks are rare. In this paper, we share 90 hours (over 600 000 samples with time-resolution of 500 milliseconds) of smartphone measurements conducted in a dense urban environment in the downlink of an operational LTE network. To capture, on the one hand, the short-term fluctuations of the measured key parameter indicators, and on the other hand, also time-of-day effects, we chose a static indoor setup. Despite the static configuration, the dataset offers a rich variability over time. The dataset can be appealing for machine learning applications (e.g., throughput prediction), simulations verification, or for educational purposes, as it contains detailed statistics throughout different LTE layers, which may help students to better comprehend the concepts of communication in LTE downlink.

Uplink non-orthogonal multiple access (NOMA) has been proposed as an efficient technique to support massive connectivity and reduce access-latency. However, due to the inherent multiuser interference within such a system, iterative joint detection is required, which is of high-complexity. In this paper, we exploit the propagation properties of wireless channels to reduce the detection complexity. In particular, when neighboring spreading-blocks on the time-frequency grid experience similar channel conditions, then it is possible to reuse the calculated filter weights between them. We propose four detection strategies and compare them across a wide range of time- and frequency-selectively. Then, assuming the base station is equipped with a sufficient number of antennas, we replace the MMSE filter with a lower-complexity approximation using Neumann series expansion. The results show that our strategies incur only a small performance loss, while substantially cutting down complexity.

Benchmarking mobile network operators based on crowdsourced data instead of conducting costly drive tests, can drastically increase the spatial and temporal coverage of the area of interest. If a user is subject to a tariff-limit, then the results from crowdsourcing might be misleading-effectively providing tariff measurements and not network benchmarks. Hence, detecting whether a tariff is the limiting factor in one specific broadband measurement is crucial to ensure that such crowdsourced approaches are credible. Although machine-learning techniques are promising for this use case, collecting a large number of labeled samples is non-trivial, especially if we aim to cover a wide range of test conditions.In this paper, we use a reference LTE cell, which we can fully control, to conduct automated measurements over a wide range of signal strengths. By carefully engineering different features, we show that this training data set can be separated almost perfectly. We then implement a semi-supervised learning approach based on this feature vector, and use label spreading to classify the validation data set that we collected from several walk tests.Providing unlabeled data increases the accuracy of the classifier to 99%, thus helping to bridge the gap between lab-data and real-world tests. This shows that automated measurements collected from a reference cell under lab conditions are - when combined with unlabeled outdoor data - a sufficient basis for a robust and accurate tariff limit detection scheme.

Enhanced Mobile Broadband (eMBB) is just around the corner, and its performance aspect makes it interesting for Vehicle-to-Vehicle (V2V) communication. Fair benchmarking of communication networks has been a research topic for decades. In contrast to wired networks, mobile networks lack a quick, accurate, and data-saving tool to describe the properties of the connection in a vehicle. Therefore, this work aims to create a crowdsourcing-ready and easily expandable software platform capable of analyzing the current network conditions in a quick and data-saving way. Based on literature research, we identified CRUSP as quick, data-saving, and accurate method to characterize mobile communication networks. Afterwards, we realized Constant Rate Ultra Short Probing (CRUSP) in a flexible software platform for performing, gathering, and analyzing measurements. The evaluation of the software was performed in a configurable and shielded reference cell and delivered an average error of 2.44% to iPerf3, a popular reference tool. In conclusion, the results indicated that the developed software operates accurately in Long-Term Evolution (LTE) networks.

A state-of-the-art method of characterizing a mobile network operator's performance or of benchmarking multiple operators is to measure the achievable throughput along a particular route. It does not matter if the measuring devices are mounted on a train, car, bus, or truck. During longer measurement campaigns, there will undoubtedly be some intervals, during which the vehicle remains stationary. In these static phases, the impact of small-scale fading on the throughput is uneven.Although small-scale fading is a well-known phenomenon, the body of literature that examines its impacts on the measured throughput is scarce. Our lab experiments in LTE have shown that the throughput differences between nearby points reach tens of Mbit/s while the small-scale throughput pattern remains stable over several days.Furthermore, we examine the impact of static measurements on the throughput distributions collected from several LTE drive tests. We find that the small-scale patterns remain stable even in an outdoor environment; thus, the static samples introduce a significant bias in the overall distribution. If we exclude the zero-speed intervals, then we obtain identical distributions from two receivers, which are 20 cm apart.

To provide reliable estimates of wireless network performance along a specific route, repeated measurements should always be conducted to determine the consistency and stability of the relevant key parameter indicators. If the same speed profile cannot be guaranteed during each repetition, then the alignment of the measured signals becomes a challenging task. To tackle this problem, we introduce the concept of the reference path and propose a modified dynamic time warping algorithm. By combining both ideas, we obtain a novel RP-DTW algorithm (modified dynamic time warping with a reference path) that aligns repeated measurements optimally. To demonstrate its usefulness, we apply the algorithm to measurements of received signal power in operational LTE networks of three Austrian operators, which we collected while being onboard a train that traverses between Vienna and Gmünd. We publicly offer our measurements as an open dataset.

With the rising interest in vehicular communications many road safety applications have been developed over the last years. Road safety applications demand low end-to-end latency which can be supported by the large bandwidth available in the millimeter-wave (mm-wave) band. However, with growing carrier frequency the wireless network coverage degrades dramatically. In our work, we focus on enhancing the vehicle-to-infrastructure (V2I) link through idle vehicular users. We enable idle users to act as relays and to boost the signal from the Base Station (BS) to the users with poor quality links and therefore enhance the performance of the entire network. We analyze this approach in a 2-dimensional (2D) Manhattan grid where micro-cells and vehicular users are placed randomly. We consider a part of the users to be idle and select the one who maximizes the coverage improvement to boost the signal from the BS depending on the street, BS and user density. Based on techniques of stochastic geometry, we derive an analytical expression for the coverage probability of the direct link as well as the relay-assisted link and compare the analytical results to Monte Carlo system level simulations in order to validate our model.

Inter-cell interference remains to be a bottleneck for conventional cellular networks as cell-edge users continue to suffer from poor performance. Cell-free massive MIMO is a novel network architecture that suppresses inter-cell interference by eliminating cell boundaries. It promises uniform performance throughout the coverage area, enabled by the coherent joint transmission from multiple distributed antennas. To make the network scalable, a user-centric approach is adopted where each user is served by a cluster of nearby access points (APs). In this work, we study the impact of cluster formation on the total fronthaul requirement, which is a limiting factor in practical coordinated distributed systems. We also investigate its effect on the guaranteed quality of service (QoS) of the network. Using our proposed algorithms, we look into the optimal cluster sizes for a given scenario and show that good performance can be achieved even with relatively small user-centric clusters, which then translates to fronthaul savings.

In this work, we investigate the potential of deep feedforward neural networks for user position estimation for a multi-path directional channel model in presence as well as absence of the line of sight path. Furthermore, we take advantage of a triplet network architecture combined with a triplet loss function, which allows us to exploit intrinsic properties of channel state information. We show that despite the high-dimensional nature of CSI, the proposed network can be trained to learn from the low-dimensional space and with less amount of training samples compared to the long-established approaches in the literature for fingerprinting localization. Finally, in order to investigate the performance and emphasize the benefits of triplet loss, we compare it to another network based on classification.

Railway communications has attracted significant attention from both academia and industries due to the booming development of railways, especially high-speed railways (HSRs). To be in line with the vision of future smart rail communications, the rail transport industry needs to develop innovative communication network architectures and key technologies that ensure high-quality transmissions for both passengers and railway operations and control systems. Under high mobility and with safety, eco-friendliness, comfort, transparency, predictability, and reliability. Fifth-generation (5G) technologies could be a promising solution to dealing with the design challenges on high reliability and high throughput for HSR communications. Based on our in-depth analysis of smart rail traffic services and communication scenarios, we propose a network slicing architecture for a 5G-based HSR system. With a ray tracing-based analysis of radio wave propagation characteristics and channel models for millimeter wave (mmWave) bands in railway scenarios, we draw important conclusions with regard to appropriate operating frequency bands for HSRs. mymargin Specifically, we have identified significant 5G-based key technologies for HSRs, such as spatial modulation, fast channel estimation, cell-free massive multiple-input-multiple-output (MIMO), mmWave, efficient beamforming, wireless backhaul, ultrareliable low latency communications, and enhanced handover strategies. Based on these technologies, we have developed a complete framework of 5G technologies for smart railways and pointed out exciting future research directions.

We propose a novel recursive multi-stage approach to Grassmannian quantization. Compared to the commonly employed single-stage quantization, our method has the advantage of significantly decreasing the number of codebook searches required for quantization and, thus, reducing the complexity. On the downside, the multi-stage approach causes a slight rate-distortion degradation compared to single-stage quantization. We analyze the rate-distortion performance of the proposed recursive quantization approach, considering random vector quantization within the individual stages. We furthermore propose a bit-allocation optimization amongst the stages of the quantizer, given a constraint on the total number of quantization bits.

Delay diversity (DD) is a low-complexity and flexible transmit diversity technology for coded offset quadrature amplitude modulation based filter bank multicarrier (OQAM/FBMC) transmission systems. However, the introduction of delay parameters in DD-OQAM/FBMC systems leads to increased channel frequency selectivity, which makes the channel estimation problem more complicated. In this paper, we analyze and develop preamble-based channel estimation methods in DD-OQAM/FBMC systems from the perspective of frequency-domain, transform-domain, and time-domain aspects. Besides, channel estimators that can obtain the minimum channel estimation mean square error under least squares (LS) and linear minimum mean square error (LMMSE) criteria are derived according to different levels of channel knowledge. Specifically, the proposed frequency-domain method estimates each channel frequency response individually based on a simplified frequency-domain channel model. The transform-domain method first estimates the equivalent single-input-single-output channel observed at the receiver assuming subchannel flatness, and then uses a window function in the time domain to separate the channels of different transmit antennas. The time-domain method is constructed based on an accurate time-domain channel model to directly estimate the channel impulse response without making any assumption about the subchannel flatness. A series of simulation experiments is conducted to verify the effectivity of these channel estimation methods for DD-OQAM/FBMC systems.

Numerical simulation of wireless communications systems is an important means within the process of development and evolution of mobile communications specifications. We offer a software suite to the academic society for free under an academic use license that enables the community to perform simulations of relevant scenarios for 5G and beyond in a reproducible manner. To cover a variety of potential scenarios for future mobile communications systems, we offer the Vienna 5G Link Level (LL) Simulator and the Vienna 5G System Level (SL) Simulator. In this contribution, we perform a verification of our LL and SL simulators which allows us to claim simulation results to be valid. As the Vienna 5G SL Simulator relies on link performance simulation results carried out with the Vienna 5G LL Simulator, we further consider their interaction in the context of verification.

We present a novel localization method based on directional beams, as available in novel massive MIMO transmission techniques instead of radius information, and derive a least squares (LS) estimation method. The new method is a direct LS method that can be solved by a linear set of equations rather than an iterative method required for radius information. In a further step, we also show how to transform radius information into virtual beams to apply the proposed method. Finally, we evaluate the accuracy of the new methods by simulations.

For spreading-based multiple access, whether orthogonal (OMA) or non-orthogonal (NOMA), the spreading sequences (signatures) are selected from a predefined codebook. When operating in a cellular system, intercell interference will be inherently present between close base stations that share the same resources. If the codebook is reused across the different cells, then intercell interference can cause a full collision of the interfering users in the code-domain, thus deteriorating their performance, especially those at the cell-edge. In this paper, we propose a method for reducing intercell interference by means of jointly designed codebooks. The criterion for the code-book design is to minimize the maximum cross-correlation between the signatures of the interfering cells. In order to obtain such code-books, we employ the algorithm of alternating projection. We finally apply the method to two interfering cells for both NOMA and OMA systems, with the users being uniformly distributed in the cells, and show that it can provide a considerable gain to the cell-edge users.

In this paper, we propose a novel optimization framework for resource and numerology allocation in multi-user scenarios. The goal of our optimization is to equalize the users' achievable rates. Our optimization includes intersymbol and intercarrier interference, channel estimation error and interband interference as a result of mixed numerology. The optimization can be formulated as an integer linear program. To reduce the computational complexity of the optimization, we also propose a linear relaxation of the problem. We investigate the performance of our methods by numerical simulations, demonstrating significant gains over an LTE-compliant scenario with fixed numerology and a heuristic approach.

Throughput is one of the most important key performance indicators in cellular mobile networks. In wired networks, there are applications that estimate available throughput based on various low-volume traffic patterns. In wireless networks, however, these techniques have not proven itself to work. Common throughput measurement tools estimate the available throughput by downloading as much data as possible for a predefined time. If the signal-to-noise ratio is high and the cell load is low, then such approach becomes expensive in terms of the consumed data volume, which users usually have to pay for. In our previous work, we proposed constant rate ultra short probing (CRUSP); in the controlled LTE test cell, we confirmed that CRUSP can estimate available throughput while consuming less than 2 MB. In this paper, we report on several drive tests (vehicular measurements) and analyze the performance of three Austrian operators in order to verify whether CRUSP can also be successfully applied to real live LTE networks. Due to the highly dynamic nature of the mobile networks, we lack the ground truth when measuring throughput in live cells. To tackle this problem, we scheduled the same number of longer iPerf3 measurements, and then compared the throughput distributions of both tools using the Kolmogorov-Smirnov test. We find that CRUSP yields the same distribution as iPerf3 if there is no UDP throttling in force. With UDP throttling, CRUSP does not match iPerf3 perfectly, but its distribution is still close. Hence, we can use CRUSP for operator benchmarking. In most cases, CRUSP saves more than 90% of measurement time and more than 90% of data volume.

Periodic Channel Quality Indicator (CQI) feedback consumes much of the uplink resources. In scenarios such as urban festivals, sport events, and Olympics football World Cups, which pose additional challenges to the wireless networks due to the heavy traffic load, channel estimation strategies have to be implemented to overcome the problem of signalling overhead while satisfying certain performance bounds. To reduce the CQI feedback overhead, we propose a limited feedback selection scheme. The proposed scheme permits the Base Station (BS) to obtain CQI from a subset of users under substantially reduced feedback overhead and estimate the channel for the remaining users. We cast the problem of CQI estimation by exploiting the theory of Gaussian Process Regression (GPR) which benefits from the correlation property of the macroscopic shadow fading. I.e., users that are close to each other have a correlated channel. The results show that with the proposed approach, a significant reduction in feedback is achieved while keeping the BLock Error Ratio (BLER) below 10% threshold.

In 5G cellular networks, a lightweight, online, real-time, and precise user flow prediction can improve the scheduling of adaptive radio and core network resources. We compare six different methods for short-to-medium-time prediction that fulfill these requirements. In our scenario, we predict the flow of users in a small cell indoor network. We then compare the performance of each method to predict the number of users per access point for several discrete time intervals without any intermediate information. To benchmark the performance of each method in live networks, we collected data from an operational WLAN network at the university over a period of one semester. The results show that common Markov-based solutions perform well only on very short prediction horizons. Beyond that, they are even outperformed by naive predictors. Solely the machine learning approach based on a neural network outperforms all other methods for any prediction horizon. This method enables a real-time and precise user flow prediction at the edge nodes of the network, as the complex task of training can be performed centralized and offline.

In recent years, the demand of train-commuting nomadic users for the Internet has been a huge challenge for train and mobile operators. Consequently, train operators have started to deploy in-train repeater systems to increase mobile users' service quality. These systems generally follow the structure of static deployments. However, in contrast to a static deployment where the system is set up once using constant parameters that are based on an initial measurement, the train will commute in dense urban city centers and sparsely populated rural areas. We propose a holistic model that helps to understand the role of each system component, namely, cabin pathloss, repeater model, window penetration loss, and outdoor pathloss, in which each individual element can be measured independently. We parametrize the model by measuring an existing in-train repeater deployment of a high-speed train. Finally, we have analyzed the operation of the complete system for two extreme cases of deployment strategy: simple rural and optimized track side. We show that the main benefit of the tested system is visible in the nonoptimized rural deployment. However, the optimized deployment also benefits from the repeater setup.

In recent years, many public high-speed trains have been upgraded with active in-train repeaters in order to improve mobile service coverage for train passengers. With a repeater in place, the user equipment perceives a wireless channel with a direct path and an indirect path through the repeater delayed by several micro seconds. In vehicular scenarios, the pick-up antenna experiences a wide range of signal levels. Due to the limited output power of the repeater, the paths may arrive with similar power, which can be a problem for OFDM based systems like LTE. Drive tests in LTE have shown that uplink performance degrades, even if the delay stays within the cyclic prefix. In order to gain a better understanding, we conducted a laboratory experiment to measure, quantify, and analyze the impact of two-path channels, with different delays and power levels, onto a live LTE network. Results show uplink performance degradation and increased transmit power even for delays smaller than the cyclic prefix due to non-optimal time-synchronization. In the system under test this degradation already starts at a path power difference of 15 dB and reaches up to 50% at equal path power levels.

This paper introduces a simple measurement methodology that can identify extreme cases of multipath propagation for in-train repeater (ITR) deployments, which may cause performance impairments for users. Differential timing advance measurements allow to detect the relative power of multipath components in the range of -20...20 dB, as we have verified through laboratory measurements. The experimental results for an existing ITR deployment inside an express passenger train show that this method can, in fact, be used to pinpoint the specific locations for which the interplay between the base station configuration and the ITR is still not optimal. Therefore, it is possible to use the proposed method to continuously collect measurement points, e.g., by crowdsourcing from standard handsets running on Android. This enables to continually improve deployments with active components such as ITRs.

In this paper, we propose an optimization method for resource and numerology allocation in multi-user scenarios. We focus on the achievement of packet latency constraints in low latency communications as one of the fundamental service requirements in 5G. We consider two groups of users: low latency users that impose a latency constraint and conventional, non-low latency users that do not impose a latency constraint. Furthermore, we consider users with varying channel conditions characterized by their delay and Doppler spread. Additionally, our optimization includes the effects of channel estimation and inter-numerology interference. The optimization problem is formulated as a multiscenario max-min Knapsack problem and it is given by an integer programming solution.

The ability to perform systematic, automatized, and repeatable measurements over large areas is essential for analyzing the spatiotemporal properties of key performance indicators in cellular mobile networks. A systematic spatiotemporal analysis is required to verify 3D channel models and antenna patterns (e.g., massive MIMO or 3D beamforming in 5G). To this end, using unmanned aerial vehicles (UAVs) for measurements opens new possibilities for developers to perform measurements more efficiently, especially now that UAVs have become increasingly available and commercialized. Currently, researchers employing UAVs to investigate mobile network parameters rely on manual UAV control, which cannot perform repeatable flight routes. In this paper, we introduce our measurement setup using a hexacopter, and then present the initial results of our autonomous measurements. The stock drones, available on the market, do not fully match the requirements for repeatable measurements; thus we upgraded the drone with additional sensors. We give an overview of the different hardware and software components that we used in our experiment. Accordingly, our results will help other researchers to determine which UAV solutions will be useful in measuring KPIs and to identify the limitations of using this method.

Evaluating mobile broadband in an operational cellular network is a challenging measurement task that depends on many parameters. Throughput is an essential performance indicator in any data network. Particularly in cellular mobile networks, the throughput is very sensitive to the number of active users--the cell load. Nearly every publication discussing throughput measurements in live mobile networks cites that throughput strongly varies in time. However, none of the studies have attempted to remove the impact of the cell load to achieve a more predictable throughput behavior. In this paper, we propose a new throughput-correction technique. Since we have access to the information on the amount of resources scheduled to the measuring device, we can approximate the total throughput that the device would experience if all the resources would get allocated to it. We focus on LTE and analyze several repeated measurements conducted with user equipments mounted in a car. We measured in a live LTE network along a selected highway segment, at different times, on two different days. The results have revealed that the corrected throughput exhibits significantly smaller variations than the raw throughput does. Finally, we discuss possible limitations-especially frequency selective scheduling combined with a low share of resources-and outline further directions of research.

We show that the deterministic two-ray model and the statistical two-wave with diffuse power (TWDP) model are suitable channel models for millimeter wave train-to-infrastructure wireless communications. Both models make parsimonious use of model parameters and are therefore implementable with very low complexity. Once directive antennas are employed, such models turn out to be accurate. The wireless channel is mainly dominated by the line-of-sight component and a ground reflection then. These two components lead to an oscillatory interference pattern with instantaneous frequency defined by the geometry. We derive fading envelopes to find a simple parametrization of the TWDP model through the deterministic two-ray model. The effectiveness of our approach is demonstrated on ray-tracing data.

Grassmannian frames consist of unit-norm vectors with a maximum cross correlation between each other that is minimal. A property like that is desired in many applications, such as in wireless communications, sparse recovery, quantum information theory, and more. In this letter, we present an iterative algorithm targeting the construction of Grassmannian frames, based on a collision-based packing of equal-radius hyperspheres on the surface of a unit-norm hypersphere. Our results show that the algorithm is capable of producing frames with very low coherence, and at a fast convergence rate compared to other methods.

Millimeter waves represent a promising solution to provide a sufficiently wide spectrum to satisfy the future requirements on broadband services. In this paper we evaluate the capacity of a short-range indoor wireless 60 GHz channel affected by quadrature imperfections of a real RF transceiver with orthogonal division multiplex approach. In our investigation, we consider spatially-dependent channel transfer function coefficients obtained from real measurements.

Tariff limits in place are one of the major challenges for crowdsourced benchmarks, blocking us from determining the network performance. We propose a new, lightweight method for traffic shaping detection and compare it to an already existing algorithm that relies on the detection of a level shift in the achieved throughput. The level shift detection is suitable for online processing during a throughput measurement and allows for an estimation of additional parameters characterizing the traffic shaper. Our method has a lower computational complexity that offers us the opportunity to analyze large datasets. Another advantage of our method is that it requires only two parameters - threshold and bin size, whereas the level shift detection needs four tunable parameters. The level shift detection may fail in mobile networks where the throughput exhibits higher fluctuations. Based on reference LTE measurements, which we have made publicly available, we show that our method achieves better performance in terms of detection rates and false alarm rates. We propose a modification of the level shift detection algorithm to improve its performance. Finally, our method is validated in the context of mobile operator benchmarking. We apply the method to a large crowdsourced dataset containing approximately 150000 crowdsourced throughput measurements in the LTE networks of three major Austrian operators. After removing the tariff limited tests, the operator ranking changes dramatically and better reflects the quality of the network rather than the share of tariff limited users.

For real-world performance evaluation, channel models should be accurate in reflecting a realistic behavior between transmitter and receiver. A major concern with the actual standardized channel models, is that they do not consider the time evolution and are relevant only for drop based simulations. The need for spatial consistency of these channel models has been also acknowledged by standardization bodies, e.g., 3GPP, and alternative models are under discussion. In this paper, we compare the statistics generated with the 3GPP 3D channel model to those of ray tracing simulations and validate our method to achieve spatial consistency. Moreover, we estimate the correlation parameters such that the results obtained with our method match the ray-tracing results.

In current wireless cellular networks, the BS tries to provide the highest possible rate to the users that can reliably be decoded. The BS obtains this rate information as a feedback from the users. However, the biased CQI reporting by the users impacts the rate and hence the system performance. The bias is directly related to the estimation error of the predicted SNR obtained by the Gaussian Process Regression (GPR) method in which the spatial correlation of the wireless channel is used for wireless channel prediction, and in particular, for predictive resource allocation. In this paper, the objective is to optimize the predicted SNR-CQI mapping by introducing an offset, adapted to the estimation accuracy relative to the user density. Results show that the proposed CQI mapping enhances the system performance for different user densities.

Spatial wireless channel prediction is crucial for future wireless networks, and in particular, for predictive resource allocation. In this paper, we first predict the channel quality indicator (CQI) at an arbitrary test user based on the Gaussian process regression (GPR) method. Second, in order to limit the overall signalling overhead, we exploit the correlation property of the wireless propagation channel. The performance of the proposed method is evaluated by the Cram'er- Rao bound (CRB). Simulation results not only well demonstrate the potential of our proposed method, but also match with the theoretical analysis.

Channel quality prediction is an essential function for anticipatory and proactive radio resource allocation. In this paper, we propose a channel quality prediction method based on the concept of Gaussian Process Regression (GPR) in which the spatio-temporal correlation of the wireless channel is used for wireless channel prediction. The objective of the paper is to find the optimal channel quality prediction for non-static users. Furthermore, we propose our analytical optimization in the choice of users which enhances the spatio-temporal correlation of the wireless channel and results in performance improvements in terms of BLER and rate loss. Simulation results show the potential of our proposed method.

In order to meet the user requirements in the next generation radio access technology, the ongoing standardization within 3rd Generation Partnership Project (3GPP) offers flexibility in terms of employing mixed numerology. The term numerology refers to the parametrization of the multicarrier modulation scheme. In this paper we consider Single User Single-Input-Single-Output (SU-SISO) Orthogonal Frequency Division Multiplexing (OFDM) under the doubly-selective channel. We investigate the optimal pilot-symbol pattern, accounting not only for Intercarrier Interference (ICI) and Intersymbol Interference (ISI) caused by channel conditions, but also considering the channel estimation error as a consequence of imperfect channel estimation. As a cost function for our optimization we choose the upper bound of the constrained capacity and apply it for different numerology. We show the throughput results in terms of different numerology and compare it to an LTE-compliant scenario.

We propose a method to estimate doubly-selective channels based on the time and frequency correlation of scattered pilots. To reduce the interference at the pilot and data positions, we apply an iterative interference cancellation scheme. Our method is applicable to arbitrary linear modulation techniques, with Orthogonal Frequency Division Multiplexing (OFDM) and Filter Bank Multicarrier Modulation (FBMC), being special cases. Simulations over doubly-selective channels show that our channel estimation method comes close to having perfect channel knowledge available. A downloadable Matlab code supports reproducibility.

Downlink throughput measurements in cellular mobile networks are of interest not only for researchers but also for mobile network operators and end-users. Conventional smartphone applications measure the throughput by downloading as much data as possible for a predefined duration, consuming tens of megabytes per measurement and blocking Internet connections for several seconds. Fast low-volume probing saves users' data. It also allows for more frequent measurements, leading to higher spatial resolution along a measurement path, e.g. in case of vehicular measurements. Self-driving cars require reliable connections; swift measurements would enable a constant monitoring without severe service interruptions. We analyze a possible reduction of data volume per test and compare the estimated throughput to a conventional, heavy load smartphone app. Measurements with a user equipment (UE) in live cellular mobile networks have a potential of characterizing real users' experience but suffer from limited repeatability (small-scale fading, quickly changing cell load, varying interference from neighbouring cells). To circumvent this restriction we performed measurements in a controlled LTE cell. We show that CRUSP (constant rate ultra short probing) achieves throughput estimates comparable to conventional apps in much shorter time (less than 50 ms) requiring much lower data volume (less than 2 MB). Whereas not an issue for high data volumes and long test durations, in the case of low volumes and short durations more careful processing - considering individual bursts of data - is crucial.

The performance of indoor wireless cellular networks depends on several parameters such as base station density or arrangement, as well as the blockage of the signal by walls. We investigate the dependence of the network performance in terms of coverage probability and also area spectral efficiency in order to demonstrate the influence of the base station density. We analytically derive expressions for both performance metrics including the blockage by walls and compare these results to those of extensive Monte-Carlo simulations. This is performed for different association strategies as well as for random and regular base station placements. It turns out that in general the addition of walls improves the network performance, and does so even more the smaller the base station density becomes.

We propose a novel modulation scheme which combines the advantages of filter bank multi-carrier (FBMC)-offset quadrature amplitude modulation and single-carrier frequency-division multiple access (SC-FDMA). On the top of a conventional FBMC system, we develop a novel precoding method based on a pruned discrete Fourier transform (DFT) in combination with one-tap scaling. The proposed technique has the same peak-to-average power ratio as SC-FDMA but does not require a cyclic prefix and has much lower out-of-band emissions. Furthermore, our method restores complex orthogonality, and the ramp-up and ramp-down period of FBMC is dramatically decreased, allowing low latency transmissions. Compared to pure SC-FDMA, the computational complexity of our scheme is only two times higher. Simulations over doubly selective channels validate our claims, further supported by a downloadable MATLAB code. Note that pruned DFT-spread FBMC can equivalently be interpreted as a modified SC-FDMA transmission scheme. In particular, the requirements on the prototype filter are less strict than in conventional FBMC systems.

In this paper, we present a mathematical framework to evaluate the performance of a typical outdoor user in an urban two-tier mobile network under a biased user association. The urban environment is abstracted by a Boolean scheme of blockages, which affects the deployment of outdoor base stations (BSs) and indoor small cell BSs, and the signal propagation characteristics of the wireless channels. A building is served by an indoor small cell BS with a certain probability, which we denote as occupation probability. Employing tools from stochastic geometry, we derive mathematical expressions for the coverage probability and the average spectral efficiency of a typical outdoor user. Numerical evaluations are carried out with the third generation partnership project compliant outdoor BS scenarios. Our results show that the increasing blockage density has a negative impact on the spectral efficiency in sparse outdoor BSs scenarios, but it has a positive impact on the performance of the network in dense outdoor BSs scenarios. We build our analysis on several assumptions, which are verified by extensive Monte Carlo simulations.

In LTE-Advanced (LTE-A) demodulation reference symbols are employed for pilot aided channel estimation to perform coherent detection. Since these reference symbols are allocated on the same time-frequency positions for all users in Multi-User MIMO (MU-MIMO) operation or for all spatial layers in Single-User MIMO (SU-MIMO), they are designed to be code-domain orthogonal in order to be separable at the receiver. This orthogonality is obtained by cyclically shifting a reference signal base sequence, where the specific cyclic shift values are signaled to each user. In this work we show formal equivalence of SU-MIMO and MU-MIMO for LTE-A uplink in the context of channel estimation. We propose a standard compliant mapping that assigns cyclic shifts to users such that SU-MIMO estimation methods are applicable also for MU-MIMO transmissions. Further we show the trade-off between the number of active users and the channel's frequency selectivity in MU-MIMO operation due to the residual channel estimation error.

The quality of service for an end user is defined by the quality of the network links connecting user and service provider. There are several parameters impacting network communication links, a very important one being the available end-to-end bandwidth. Bandwidth measurements in cellular wireless networks with traffic reactive nature, nomadic end users and highly dynamic shared resources is a challenging task. Current methods for such networks are very resource demanding and lack accuracy. In this paper we present a new approach for estimating the current mobile network link capacity and its implementation. It is specifically targeted on generating estimates of the available end-to-end capacity quickly and parsimoniously. The proposed solution combines packet rate and dispersion approaches putting the focus on mobile networks. We verified our technique with validation measurements in laboratory settings as well as in real world Internet access systems. In these measurements our new concept outperformed existing tools.

This paper presents a report on mutual information based on measured indoor millimeter-wave (mmWave) channels with multiple antennas at input and output (MIMO). We show that for fixed indoor communication systems, an optimal antenna element spacing can be found such that the measured mutual information almost reaches the capacity of a perfectly orthogonal (theoretical) MIMO channel (with the same number of receiver (RX) and transmitter (TX) antennas). Secondly, we present, that even though the measured channel transfer functions (CTFs) exhibit large fluctuations (i.e., temporal dispersion), the mutual information is mainly determined by the mean singular value of the line-of-sight (LOS) components. Due to their strong variations over frequency mmWave channels are tedious when describing them with classical methods in the frequency domain. An approximation by numerous flat subbands leads to an error in mutual information (MI) by 2bit/s/Hz (for 80% probability). In comparison, our proposed method in the time domain, however, offers a notably smaller error (1bit/s/Hz for 80% probability).

Today, crowdsourcing applications distribute the task of performance measurements in mobile networks over various mobile users to achieve an accurate benchmark. A common measurement methodology is an application on the customers' phones which injects probe packets into the network under test and derives estimates for selected network properties such as available bandwidth. Bandwidth measurements in cellular wireless networks, which are highly dynamic and offer only shared resources, are in general a resource demanding task. Current methods deploy tests of constant duration running data transfers over TCP. Their resource consumption scales directly with the available capacity of the link. In this work we take a closer look at the different chosen test durations and analyze the trade-off between accuracy and resources spent in various different configurations, e.g., numbers of parallel TCP connections, based on extensive measurements in real world operational LTE and UMTS networks conducted in 2016, Austria. In our experiments a minimum measurement duration of six seconds allows for estimating available bandwidth with an average error of less than 1%.

Massive MIMO, 3D beamforming and beam tracking strategies for mobile users are among the key strategies for the 5th generation of mobile cellular networks. Their investigation requires channel models that are spatially consistent and evolve smoothly over time. State-of-the-art channel models, such as, 3GPP SCM, WINNER and the 3GPP 3D channel model, do not consider the time evolution and are relevant only for drop based simulations. In this paper we propose two methods to model spatially correlated channels; first, by introducing spatial correlation to LOS/NLOS propagation and indoor/outdoor state of a user, and second, by introducing spatial correlation to the small scale parameters that represent short-term fading characteristics. Our results exhibit realistic behavior of users that move through the network, in terms of LOS/NLOS and indoor/outdoor states and preserve the statistics of the channel model.

The mobile cellular service coverage of nomadic users in the context of high speed vehicular scenarios, e.g., railroad trains, is challenging. Energy saving efforts of the past years have lead to substantially increased vehicle penetration losses, e.g., by metal shielded windows. As a result providing coverage into railroad cars is an emerging topic for optimization efforts, either based on network deployment or on railroad car materials. The quality of these efforts rely heavily on the understanding of the vehicle penetration loss and its angular dependency. In this paper we present a distributed measurement methodology and setup allowing for measuring the penetration loss of railroad cars. We validate the feasibility of the methodology in a real world measurement campaign in Austria. There, we compare two different car configurations, standard and prototypical, at 800-2600MHz for azimuthal angles of arrival of 0-60 degree. The comparison of the two configurations shows a substantial lower penetration loss of the prototypical setup due to windows with alternative material treatment.

Superior spectral properties of Filter Bank Multi-Carrier (FBMC) techniques make them an interesting choice for future wireless systems. In order to directly apply well known pilot-symbol aided channel estimation in FBMC, however, the intrinsic imaginary interference has to be canceled. Such cancellation method is usually designed for a doubly-flat channel. In this paper, we investigate the effects of doubly-selectivity, that is, time-selectivity and frequency-selectivity, on the channel estimation Mean Squared Error (MSE) in FBMC. The calculation of the MSE is based on a compact matrix description, allowing us also to find an MSE minimizing channel estimation method that takes the underlying waveform into account.

In an operational mobile cellular network, measuring the mean mobile broadband performance at a certain place in a building within a short period of time (e.g., several minutes) is a challenging task as fading is position dependent and the overall network load changes throughout a day. In this work we derive a measurement methodology that allows for swift and repeatable measurements of the mean network performance by combining three techniques: Firstly, the instantaneous IP-throughput is derived within one second by using a novel packet pattern method on an Android cell-phone with external antennas. Secondly, the long-term trend in the IP performance, that is the slowly changing average network load, is removed. Thirdly, the scenario mean of the remaining small scale fading scenario with overlaid fast network fluctuations is obtained by statistical inference, namely, spatial systematic-sampling with an XY-positioning table. To assess the validity of our approach, we present the results of a seven day long live-network measurement campaign in Vienna, Austria. We show that, once the long term network trend is removed, a one minute long measurement is sufficient to determine the mean IP-throughput performance over the whole week with 10% accuracy at 90% level of confidence.

In this paper, we investigate multicast transmit beamforming for vehicle to everything (V2X) communication employing LTE's multimedia broadcast single-frequency network (MBSFN) capabilities. LTE-MBSFN enables to broadcast/ multicast information in 4G networks, which can be utilized for efficient dissemination of road traffic related information to vehicles on highways/motorways. MBSFN is currently restricted to single antenna transmission within the LTE standard; this paper aims to present performance benefits obtained with the introduction of multiple transmit antennas. We propose an efficient multi-cell multicast beamforming technique that leads to significant SINR improvements, which enhances reliability, reduces transmission latency and facilitates achieving 5G requirements for vehicular communications.

We introduce a novel directional channel sounding concept where we sweep a horn antenna around its phase center. Directional channel measurements are thus carried out at a fixed coordinate in space. To verify our concept, we conducted multi-carrier measurements with 2 GHz of measurement bandwidth. The directional broadband channel was sampled uniformly within a cube of three wavelengths side length. In this contribution, we compare narrowband measurements with spatial averaging to traditional broadband channel sounding. We saw that, spatial filtering through directional antennas leads to a limited number of propagation paths in the channel. We show the difference of both approaches and explain the deviation by spatial correlation. The spatial correlation is evaluated at several two-dimensional slices. We observed wavelength-periodic correlations.

Future wireless systems will be characterized by a large range of possible uses cases. This requires a flexible allocation of the available time-frequency resources, which is difficult in conventional orthogonal frequency division multiplexing (OFDM). Thus, modifications of OFDM, such as windowing or filtering, become necessary. Alternatively, we can employ a different modulation scheme, such as filter bank multi-carrier (FBMC). In this paper, we provide a unifying framework, discussion, and performance evaluation of FBMC and compare it with OFDM-based schemes. Our investigations are not only based on simulations, but are substantiated by real-world testbed measurements and trials, where we show that multiple antennas and channel estimation, two of the main challenges associated with FBMC, can be efficiently dealt with. In addition, we derive closed-form solutions for the signal-to-interference ratio in doubly-selective channels and show that in many practical cases, one-tap equalizers are sufficient. A downloadable MATLAB code supports reproducibility of our results.

In this paper, we consider downlink multi-user MIMO transmission in massive MIMO systems, assuming massive amount of antennas at transmitter- as well as receiver-side. We investigate the performance of several linear MIMO transceiver architectures of comparatively low complexity, since non-linear schemes of high complexity (e.g., maximum likelihood detection) appear infeasible with large numbers of transmit and/or receive antennas. The focus of our investigation lies on the impact of spatial antenna correlation on the achievable rate of the considered schemes. We derive achievable rate upper bounds that facilitate interpretation and explanation of performance losses observed with increasing antenna correlation and with growing number of antennas and spatially multiplexed users. We identify favourable combinations of low complexity transmitter/receiver filters and show that the impact of antenna correlation strongly depends on the applied transceiver architecture.

Filter bank multicarrier modulation is considered as a possible candidate for 5G. In this paper, we consider pilot-symbol aided channel estimation and address the problem of canceling the imaginary interference at the pilot positions. We develop a matrix formulation for the transmission system which allows us to formulate general conditions on the auxiliary pilot symbols, capturing also the interde-pendency of closely spaced pilots and an arbitrary number of auxiliary pilot symbols. By using two auxiliary symbols per pilot instead of one, we are able to improve the peak-to-average power ratio as well as the achievable capacity for small to medium signal-to-noise ratios. The achievable capacity can further be increased by interference cancellation based on linear precoding for which we propose an algorithm to find the coding matrix required. Finally, we compare auxiliary pilot symbols and linear precoding in terms of complexity and performance.

Filter Bank MultiCarrier (FBMC) might replace Orthogonal Frequency Division Multiplexing (OFDM) in 5G. For such FBMC system, we derive closed-form expressions for the Bit Error Probability (BEP) including channel estimation, whereas we focus on the comparison of FBMC with OFDM. We assume additive white Gaussian noise and a Rayleigh fading channel with low delay spread and low Doppler spread, so that the channel induced interference can be neglected compared to the noise. Our channel estimation is based on pilot symbols whereby the imaginary interference, inherently caused in FBMC, is canceled at the pilot positions either by auxiliary symbols or through coding. Moreover, we propose an optimal power allocation between pilot symbols and data symbols to minimize the BEP.

The wireless channel behaves markedly different depending on whether a line-of-sight (LOS) between transmitter and receiver exists or not. State-of-the-art wireless channel models for mobile communications, such as, IST WINNER-II and the 3GPP 3D channel model specified in TR 36.873, define distinct macro- and microscopic fading parameters for LOS and non-LOS (NLOS) situations. Whether a user is in LOS/NLOS is randomly determined by a distance-dependent LOS probability; commonly, for each position in the network, the random realization of LOS/NLOS propagation is independently determined from this distance-dependent LOS probability. Since in this case random draws of neighboring positions are statistically independent, base station assignment regions in system-level simulations become highly irregular. To mitigate this problem, we propose an efficient method to generate spatially correlated LOS/NLOS channel maps that follow predefined distance-dependent LOS probability and additionally enable spatial clustering of LOS positions.

In this paper, we investigate the performance of typical indoor users in urban two-tier heterogeneous mobile networks with indoor-deployed small cell base stations (BSs) and outdoor BSs. The urban topology is modeled by a Boolean scheme of blockages, which is tied together with the deployment of outdoor and indoor BSs as well as the signal propagation characteristics. Each building is assumed to be served by an indoor small cell BS with a certain probability, which is denoted as occupation probability. Indoor and outdoor environments are partitioned by walls with a certain penetration loss. We separately account for outdoor BSs in line of sight (LOS) and non-line of sight (NLOS), as well as small cell BSs in neighboring and non-neighboring buildings. We derive expressions for the signal to interference ratio (SIR)-coverage probability and the average spectral efficiency of a typical indoor user. Numerical evaluations are performed with 3GPP compliant macro- and pico-BS scenarios. Our results exhibit that the indoor area coverage, which is a function of the building density and size, significantly affects the impact of the target building's wall penetration loss. The analysis is simplified by various assumptions, which are verified by extensive Monte Carlo simulations.

Scaling up the number of base stations per unit area and augmenting the heterogeneity of networks are two major trends in mobile cellular systems of the fourth (4G)- and fifth generation (5G), making it increasingly difficult to characterize aggregate interference statistics with system models that are characterized by a low number of parameters. This paper proposes a new circular interference model that aggregates given interferer topologies to power profiles along circles. We introduce a mapping procedure that preserves the aggregate interference statistics at arbitrary user locations within the associated cell with a certain desired accuracy. At the same time, our method identifies the amount of nodes in a given interferer topology that principally determine the shape of the interference distribution. To further refine the analysis of given power profiles, a new finite sum representation for the sum of Gamma random variables with integer-valued shape parameter is introduced. The approach allows decomposing the distribution of the aggregate interference into the contributions of the individual interferers. Such knowledge is particularly expedient for the application of base station coordination- and cooperation schemes. Moreover, the proposed method enables accurate prediction of the corresponding signal-to-interference-ratio- and rate statistics.

An open problem in slotted Aloha protocols is to optimally estimate the number of participants as such knowledge is crucial to select the optimal frame length. First results are known in literature based on observing the slot fill levels in case of empty slots and single occupancies (singleton slots). Advances in signal processing allow now also to decode successfully slots with higher fill levels, for example, due to multiple antennas. In this paper we derive the maximum likelihood estimator when arbitrary occupancies up to a maximal fill level R have been observed. Due to our novel approach, the derivation is rather simple and its implementation is of low complexity.

Although the LMS algorithm is often preferred in practice due to its numerous positive implementation properties, once the parameter space to estimate becomes large, the algorithm suffers of slow learning. Many ideas have been proposed to introduce some a-priori knowledge into the algorithm to speed up its learning rate. Recently also sparsity concepts have become of interest for such algorithms. In this contribution we follow a different path by focusing on the separability of linear operators, a typical property of interest when dealing with tensors. Once such separability property is given, a gradient type algorithm can be derived with significant increase in learning rate. Even if separability is only given to a certain extent, we show that the algorithm can still provide gains. We derive quality and quantity measures to describe the algorithmic behavior in such contexts and evaluate its properties by Monte Carlo simulations.

We consider downlink multi-user MIMO transmission based on block-diagonalization precoding with quantized channel state information at the transmitter, obtained through limited feedback. We assume that users are equipped with excess receive antennas, i.e., the number of receive antennas is larger than the number of data streams per user, and propose a novel receive antenna combining method that maximizes an estimate of the expected achievable user rate. By means of simulations, we validate our assumptions and demonstrate significant rate gains compared to existing combiners of similar complexity.

In high mobility orthogonal frequency division multiplexing systems, subcarriers are no longer orthogonal, causing Inter-Carrier Interference (ICI). Equalization then becomes more challenging and requires an accurate estimate of the time-variant channel. In this paper, we propose a novel Minimum Mean Squared Error (MMSE) estimation of the sampled time-variant transfer function. Based on such channel estimation, we propose an iterative three step ICI mitigation technique whereby each step increases the channel estimation accuracy and, consequently, the performance of our MMSE equalization. In Step 1, we consider ICI as an additional noise term. In Step 2, we reduce the ICI at pilot positions and finally, in Step 3, we treat all estimated data symbols, obtained from Step 2, as if they were pilot symbols. We evaluate the Bit Error Ratio (BER) of our ICI mitigation technique by means of simulation and testbed measurements (up to 400 km/h). In both cases, we achieve a BER close to perfect channel knowledge and zero ICI.

In this paper, we investigate downlink physical layer multicast transmit beamforming in wireless cellular networks, considering interference between multiple independent multicast transmitters (base stations). Transmit beamforming can exploit multiple antennas at the transmitter to direct the multicast signal towards the intended users, while minimizing the interference leakage caused to other users of the network. We propose a multicast beamformer optimization problem that maximizes the achievable multicast transmission rate while restricting the interference leakage caused to other users, by applying a semidefinite relaxation to approximate this NP-hard problem with a convex optimization problem that can be solved efficiently. Furthermore, we consider multiple receive antennas at the users and propose an antenna combiner that maximizes the achievable user rate. Finally, we combine the proposed beamforming and receive antenna combining methods via alternating optimization and evaluate the performance using Monte-Carlo simulations.

In this paper, we derive closed-form capacity expressions for a low complexity bit-interleaved coded modulation system with uniform inputs in a Rayleigh fading channel with additive white Gaussian noise. Additionally, we include pilot-symbol assisted channel estimation in our considerations. Finding a closed-form solution is enabled by assuming quantization and that the decoder has no channel state information. The effects of these assumptions on the capacity are investigated separately. To verify our closed-form expressions and to show the applicability of our system model to real world physical channels, we perform measurements using the Vienna Wireless Testbed.

In this paper, we investigate the impact of various remote unit collaboration schemes on the transmission performance to passengers, traveling in high speed trains. We assume that the user equipments directly communicate with the remote units, i.e., the trains are not equipped with relay nodes. We introduce a theoretical model based on Maximum Ratio Transmission and Gamma distributed fading, and analyze a representative segment of a railroad track. Results from both, theory and simulations by the Vienna LTE-A System Level Simulator are provided in terms of train average spectral efficiency and allow us to discuss the suitability of the moving cell concept for trains without relay support. Our theoretical results exhibit an offset from the simulations. Investigating its origin reveals the limitations of codebook-based precoding and allows us to adapt our theoretical model for accurate performance predictions.

Long Term Evolution (LTE) is expected to substitute the Global System for Mobile Communications (GSM) as the radio access technology for railway communications. Recently, especial attention has been devoted to high-speed trains since this particular environment poses challenging problems in terms of performance simulation and measurement. In order to severely decrease the cost and complexity of high-speed measurement campaigns, we have proposed a technique to induce effects caused by highly-time varying channels on Orthogonal Frequency-Division Multiplexing (OFDM) signals while conducting measurements at low speeds. In this work, we evaluate this technique by comparing the results of LTE measurements at different velocities as well as by simulations. Additionally, we use this technique to show the performance of LTE for high-speed train scenarios up to 600 km/h. To accomplish this, we use both a controlled high-speed measurement setup as well as a channel model developed according to the guidelines of the ITU Radiocommunication Sector (ITU-R) for the evaluation of radio interface technologies for IMT- Advanced systems.

In this letter, we consider time-varying complex-valued n × m matrices H[k] (m ≤ n) and propose a predictive quantizer for the eigenvectors of the Gramian H[k]H[k]H, which operates on the associated compact Stiefel manifold. The proposed quantizer exploits the temporal correlation of the source signal to provide high-fidelity representations with significantly reduced quantization codebook size compared to memoryless schemes. We apply the quantizer to channel state information quantization for limited feedback based multi-user MIMO, employing regularized block-diagonalization precoding. We demonstrate significant rate gains compared to block-diagonalization precoding using Grassmannian predictive feedback.

In this paper, we consider multiple-input single-output (MISO) physical layer multicasting, where several multiantenna transmitters each simultaneously multicast a common message to a distinct set of single-antenna receivers, causing interference between different multicast messages. To optimize the achievable rate of this MISO multicast interference channel, we propose iterative distributed transmit optimization algorithms that are based on interference leakage control, requiring local channel state information at each transmitter and leakage information exchange among transmitters. We, furthermore, propose extensions of existing coordinated multipoint transmission schemes that have been developed for unicast interference channels, such as signal to leakage and noise ratio beamforming, to the considered multicast system. Such methods are of importance, e.g., for future releases of LTE that will support multiantenna multicasting using MBMS/MBSFN. Numerical simulations confirm the potential of the proposed distributed transmit optimization algorithm.

We propose an iterative extension of the covariance intersection (CI) algorithm for distributed data fusion. Our iterative CI (ICI) algorithm is able to disseminate local information throughout the network. We show that the ICI algorithm converges asymptotically to a consensus across all network nodes. We furthermore apply the ICI algorithm to distributed sequential Bayesian estimation and propose an ICI-based distributed particle filter (DPF). This DPF allows for spatially correlated measurement noises with unknown cross-correlations and does not require knowledge of the network size. The performance of the proposed DPF is assessed experimentally for a target tracking problem.

We provide an analysis of a consensus-type algorithm with weights dependent only on the received data. Differently from previous approaches that require a global knowledge of the network, we consider general weights inferred only from local data which can be modified by local functions on each node. We provide convergence conditions of such algorithms for general weight functions and derive analytical steady states in some selected cases.

In this work, we investigate downlink co-channel interference in wireless cellular networks. Our main target is to facilitate statistical analysis in networks with regular grid layout. In particular, we focus on the interference statistics outside the center of a grid scenario. First, a novel circular interference model is introduced. The key idea is to spread the power of the interferers uniformly along the circumcircle of the grid-shaping polygon. We then propose to model the aggregate interference statistics by a single Gamma random variate. The corresponding shape- and scale parameters are determined in closed form by employing this circular model. The analysis yields key insights on the distribution's formative components. We verify the accuracy of the Gamma approximation by qualitative-and quantitative measures. A basic guideline for applying and extending our approach indicates that it considerably improves analytic accessibility of regular grid models.

For an orthogonal frequency-division multiplexing system with pilot-symbol aided channel estimation, we compare the measured bit error ratio to the theoretical bit error probability. In order to measure mobile physical systems, we utilize the Vienna Wireless Testbed which has been augmented by a rotation wheel unit. The analytical solution assumes Rayleigh fading, additive Gaussian noise, and an arbitrary linear interpolation method to estimate the unknown channel taps. Our measurements confirm our assumptions and demonstrate convincingly that our theoretical expressions accurately model the true physical behavior, even for speeds of up to 100\,km/h.

In this article, we investigate the performance of cellular networks employing remote radio units to enable distributed multi-user MIMO transmission using block-diagonalization precoding. We consider the downlink of an OFDM-based LTE compliant multi-carrier system in which the users provide channel state information (CSI) to the base station via limited capacity feedback links. With limited feedback, residual interference between the spatially multiplexed users cannot be avoided, causing an interference-limitation of the achievable downlink throughput. Efficient CSI quantization is therefore central in such systems to achieve a performance gain over single-user MIMO. Building on Grassmannian CSI quantization concepts, we propose an effective CSI feedback clustering approach in this work to exploit the correlation of the wireless channel in the frequency domain. We furthermore derive the corresponding channel quality feedback and propose a practical greedy multi-user scheduler.

The Proportionate Normalized Least Mean Squares (PNLMS) algorithm has been quite successful in combining higher convergence rates with low to moderate complexity that at the same time avoids numerical difficulties in fixed-point implementations. While the algorithm is stable in the mean square and l2-sense for time-invariant matrices, the treatment of time-variant matrices requires additional approximations. These approximations are discarded in this paper which allows us to analyse the robustness in terms of l2-stability for actually time-variant matrix step-sizes. This provides important results, as the algorithm in its variants also occurs in other fields of adaptive filtering such as cascaded filter structures. By simulations as well as by theoretical analysis, we demonstrate that in general, even small variations of the matrix step-size are sufficient for the algorithm to loose its robustness. Only in special cases, where specific constraints are imposed additionally, robustness can be guaranteed.

We propose an adaptive quantization algorithm for subspace tracking on the Grassmann-manifold of p-dimensional subspaces in the n-dimensional Euclidean space. This quantization problem arises naturally in limited feedback based wireless communication systems, which apply precoding for interference cancellation and alignment. The proposed algorithm exploits the differential geometry associated with the Grassmann-manifold for efficient differential and predictive quantization. The algorithm is applied to channel state information quantization in a multi-user block-diagonalization based wireless communication system, demonstrating large throughput gains compared to memoryless quantization.

In order to estimate a transmission channel, a typical wireless communications system multiplexes a fixed pattern of symbols known at the receiver within the data-symbols. In Long Term Evolution (LTE), these reference symbols can consume up to 14.2% of the total transmission bandwidth. In this work, we investigate an optimal pilot-symbol pattern design using a post-equalization Signal to Interference and Noise Ratio (SINR) framework. We show how to choose the distance between adjacent pilot-symbols close to optimally and how to distribute the available power among pilot- and data-symbols at the same time under time-variant channels. We confirm our analytical solution by means of simulation. Compared to a system with a fixed distance between pilot-symbols and unit power distribution, the proposed system configuration yields significant gains in terms of capacity.

This paper introduces a highly efficient waterfilling-based strategy for optimal use of the channel in vehicular or personal ad-hoc communications. The approach provides near optimum allocation of resources and enables small communications devices to establish connections when the energy efficiency is at its best. Instead of calculating the transmit power through conventional waterfilling, the proposed algorithm calculates the waterlevel, thus providing a decision threshold and a strategy for optimum use of the time-variant channel at the same time. The algorithm adapts to the changing average channel quality by applying an exponentially-weighted moving-average (EWMA) trigger to re-calculate the waterlevel. The new algorithm is compared to an efficient non-iterative algorithm that directly calculates the transmit powers in every time-slot. It is shown that the new strategy reduces computation time by approximately 90% compared to the classic approach without compromising performance measures such as transmitted information or energy. Practical implementation is briefly discussed to demonstrate suitability of the algorithm for integration into tomorrow's communication devices.

Nowadays, most wireless communication systems utilize coherent detection which implies the necessity of multiplexing reference symbols between data-symbols for the purpose of channel estimation. In Long Term Evolution (LTE), these reference symbols can consume up to 14.2% of the total bandwidth. In this work, we search for an optimal pilot-symbol pattern design using a post-equalization Signal to Interference and Noise Ratio (SINR) framework. We show how close to optimally choose the distance between adjacent pilot-symbols and how to distribute the available power between pilot and data symbol at the same time. We confirm the performance of our analytical solution by means of simulation. Compared to a system with a fixed distance between pilot-symbol and unit power distribution, the proposed system configuration yields a gain of around 30% in terms of capacity.

We present novel distributed algorithms for estimating the number of nodes in a wireless sensor network without any a-priori knowledge or node preferences. The algorithms originate from distributed forms of Gram-Schmidt orthogonalization algorithms where the goal is to distributively find a set of orthogonal vectors. Using concepts from linear algebra, by finding the number of independent (orthogonal) vectors, we also find the number of nodes in a network.

Spatial multiplexing of multiple users in a multi-antenna broadcast system, i.e., multi-user MIMO, is a promising technique for improving the spectral efficiency of cellular networks. Still, practical implementations struggle with the difficulty of obtaining sufficiently accurate channel state information at the transmitter (CSIT), due to restrictions on the capacity of the CSI feedback links from the users. Without accurate CSIT, residual multi-user interference diminishes the potential performance gain of multi-user MIMO, challenging its value for practical realizations. In this work we show that the CSI feedback overhead can be significantly reduced if the number of spatially multiplexed data streams per user is less than the number of receive antennas, and the antenna outputs are appropriately combined. We propose an interference-unaware antenna combining algorithm that minimizes the CSI quantization error, which has the effect of minimizing multi-user interference at the price of reducing the received signal power. We mathematically analyze the performance of the proposed algorithm, substantiating its value for limited feedback systems. Numerical simulations confirm the potential of the proposed algorithm in comparison to a traditional technique that maximizes the received signal power.

Recent standards for cellular transmission systems offer a lot of flexibility, such as the choice of transmission modes, modulation alphabets, coding rates, and precoding matrices. Despite this trend, pilot-symbol patterns in today's standards remain fixed, although such an approach is suboptimal. In this paper, we show how to design optimal pilot-symbol patterns by maximizing an upper bound of a constrained capacity that takes channel estimation errors and Inter Carrier Interference into account. Furthermore, we propose adaptive pilot-symbol patterns that follow changing channel statistics. As a proof of concept, we present throughput simulation results of two competitive systems, a transmission system compliant with the Long Term Evolution (LTE)-standard and an improved system utilizing the proposed adaptive pilot patterns. The transmission system utilizing adaptive pilot patterns outperforms an LTE-standard compliant system in all considered scenarios. The throughput gain for a single input single output system ranges between 3% and 80%. For a 4 × 4 transmission system, the performance gain is significantly higher and can reach up to 850% compared to a conventional LTE system.

This paper presents an SINR prediction model for LTE systems with the aim of improving the accuracy of physical layer models used for higher-than-physical-level simulations. This physical layer SINR abstraction for zero-forcing receivers takes channel estimation errors into account. It allows for the calculation of the post-equalization receive SINR of an LTE MIMO transmission on a per-layer and subcarrier basis. This estimate is further used as input for a link performance model to accurately predict the receiver throughput. The accuracy of the model is validated by simulations for 2×2 and 4×4 MIMO antenna configurations with different levels of spatial multiplexing employing also the precoding matrices defined in the LTE standard. Memory requirements and complexity for the model are quantified as well. The whole simulation environment together with the fully reproducible results of this paper are made available for download from our homepage.

In this paper, we present algorithms for the quantization of a correlated unit norm random vector process. Such algorithms are important, e.g., for channel vector quantization in wireless communication systems. Starting from a quantization codebook that uniformly quantizes the unit sphere, we propose to iteratively adapt the codebook to the channel statistics, in order to improve the quantization accuracy. This is achieved by increasing the density of the quantization code vectors in that area of the unit sphere where the next realization of the random process is expected to lie, without increasing the codebook size. Additionally, we propose spherical prediction algorithms for the considered random process. Combining the codebook adaptation techniques with this prediction leads to improved quantization accuracy. The performance of the algorithms is demonstrated by employing them in an LTE system for providing accurate transmitter channel knowledge used for multiuser beamforming.

Many candidate technologies for future wireless communication systems (e.g., multi-user MIMO, interference alignment), rely on accurate Channel State Information (CSI) at the Transmitter (CSIT) to determine the appropriate multi-antenna pre-processing steps. Thus, in frequency division duplexing systems, a dedicated feedback link has to be provided to each attached user for CSI reporting. In order to keep the feedback rate moderate, CSI quantization is required. Additionally, CSI is typically only provided for a subset of all available time/frequency resources, implicating the need for an interpolation algorithm at the base station. Furthermore, to compensate for the processing delay of the feedback link, CSI prediction is required as well. In this work, we treat these problems by incorporating our previously presented CSI feedback algorithm in a 3GPP LTE-A compliant wideband OFDM simulation environment, and augmenting it with appropriate interpolators and predictors. As an example, the performance of multi-user zero-forcing beamfoming is investigated, under practical feedback delay- and time-frequency granularity-constraints, demonstrating substantial throughput gains, in case of low to moderate user mobility.

A continuous challenge in the field of network traffic modeling is to map recorded traffic onto parameters of random processes, in order to enable simulations of the respective traffic. A key element thereof is a convenient model which is simple, yet, captures the most relevant statistics.

Current Multiple Input Multiple Output (MIMO) Orthogonal Frequency Division Multiplexing (OFDM) based systems for wireless communications enable to adjust power radiated at the pilot symbols. Under time-invariant channels, a power increase at the pilot symbols results in improved quality of the channel estimate. However, under time-variant channels, this is not necessarily the case. Due to the decreased temporal channel correlation, the channel estimation performance becomes saturated. If under time-variant channels more power is assigned to the data symbols, such approach increases inter layer interference due to the imperfect channel knowledge at the receiver. In this paper, we show how to distribute power among data and pilot symbols in a power efficient way. Using our proposed method, up to 50% of the transmit power can be saved in a 4x4 Long Term Evolution (LTE) downlink transmission at a Signal to Noise Ratio (SNR) of 20 dB.

We present a large-scale cell based measurement analysis of the user behavior in a live operational HSDPA network. The motivations are: first, to understand the statistical properties of users in cells for refining network planning procedures; and second, to provide realistic traffic models for simulations of cellular packet-oriented networks. We provide an analysis of mean cell load over daytime, as well as models for short-term cell load in terms of user activity and throughput. Furthermore, an evaluation of user-sessions with respect to duration and mean throughput is given. Our findings lead to four different models reflecting realistic user-traffic load of cellular networks. To verify the concept we present respective simulations which investigate multiuser-scheduling in an LTE-network. They show that conventional simulation settings can lead to an overestimation of performance.

The authors take a look at the existing 3G systems in service and investigate the capabilities of 4G, and while the theoretical throughput of these cellular systems is expected to be high, the future promises to offer more technological improvements and innovations.

The Myth: Sixty years of research following Shannon's pioneering paper has led to telecommunications solutions operating arbitrarily close to the channel capacity—“flawless telepresence” with zero error is available to anyone, anywhere, anytime across the globe.

We propose a distributed implementation of the Gaussian particle filter (GPF) for use in a wireless sensor network. Each sensor runs a local GPF that computes a global state estimate. The updating of the particle weights at each sensor uses the joint likelihood function, which is calculated in a distributed way, using only local communications, via the recently proposed likelihood consensus scheme. A significant reduction of the number of particles can be achieved by means of another consensus algorithm. The performance of the proposed distributed GPF is demonstrated for a target tracking problem.

In this contribution we provide a thorough stability analysis of gradient type algorithms with non-symmetric matrix step-sizes. We hereby extend existing analyses for symmetric matrix step-sizes and present several methods to derive step-size bounds. Although we can guarantee the I2- stability for such algorithms only under very restrictive conditions, we are able to proof convergence in the mean square sense under much more general conditions. Some of the derived step-size bounds turn out to very tight and allow for accurate algorithmic design.

Many models of wireless sensor networks (WSNs) assume a perfect synchronization along the graph of such network as a simplifying assumption. In our contribution we base our investigations of distributed algorithms solving consensus problems on more realistic, asynchronous networks in which nodes randomly transmit to their neighborhood. Following a linear algebraic approach we show conditions for convergence to a consensus and derive convergence properties in the mean and mean square sense.

This paper presents a link error prediction model capable of accurately predicting the block error ratio in OFDM systems with hybrid ARQ. The modeling of both the channel coding and the combining of the retransmissions solves many of the issues with present models. Our model is based on accumulated conditional mutual information, mutual information effective SNR averaging, and simulated AWGN performance curves. As a consequence, our model is fading-insensitive and takes into account the non-ideality of rate matching. The model can be applied to any OFDM system that utilizes rate matching as a means to achieve adaptive modulation and coding and HARQ. We demonstrate the application of our model in LTE and simulate its accuracy. The whole simulation environment together with the results of this paper are made available for download at our homepage.

Accurate models for error characteristics of radio channels are essential to estimate performance of data transmission. This paper analyzes error-gap and error-burst statistics of the WCDMA dedicated channel (DCH) and provides respective models. We start with an analytical study of the outer-loop power control mechanism (OLPC), which gives directions to this work, revealing that the OLPC strongly influence error behavior. We prove this by large-scale measurements at the Iub-interface of a live UMTS network. Beside statistical evaluation of the measurement data we provide a simple generative hidden Markov model for emulation of DCH errors. It is able to characterize any arbitrary DCH by only two parameters, with an accuracy below 1%, in terms of Kullback-Leibler divergence. This novel model presents an accurate light-weight alternative to complex tools for simulation of WCDMA DCH connections. It also covers all other communication technologies deploying a similar power control mechanism.

We address the problem of downlink multiuser scheduling in practical wireless networks under a desired fairness constraint. Wireless networks such as LTE, WiMAX and WiFi provide partial channel knowledge at the base station/access point by means of quantized user equipment feedback. Specifically in 3GPP's LTE, the Channel Quality Indicator (CQI) feedback provides time-frequency selective information on achievable rates. This knowledge enables the scheduler to achieve multiuser diversity gains by assigning resources to users with favorable channel conditions. However, focusing only on the possible diversity gains leads to unfair treatment of the individual users. To overcome this situation we propose a method for multiuser scheduling that operates on the boundary of the achievable multiuser rate region while guaranteeing a desired long term average fairness. Our method is based on a sum utility maximization of the α-fair utility functions. To obtain a given fairness, quantified with Jain's fairness index, it is necessary to find an appropriate α, which we obtain from the observed CQI probability mass function.

A novel methodology for QoE evaluation in the social network environment is proposed. It provides high applicability for subjective testing of the multimedia services with respect to real usage scenarios. The environment of social networks provides also significant demographic data and ability to contact extremely many test subjects while allows to focus on or filter specific social groups. QoE results for HD internet video services are presented and followed by discussion on their statistical significance.

In scenarios with time-varying channels such as intelligent traffic systems or high speed trains, the orthogonality between subcarriers in orthogonal frequency division multiplexing (OFDM) is destroyed leading to inter-carrier interference (ICI). In the literature, ICI equalization algorithms have been proposed; however, they assume perfect channel knowledge at sample level. Unfortunately, existing channel estimation algorithms do not provide accurate channel estimates at high Doppler spreads, prohibiting data transmission with high spectral efficiency. In this paper, we propose an algorithm for ICI estimation that can be applied to OFDM systems with an arbitrary pilot structure. Thus, our algorithm can be applied to any already standardized OFDM system. Our ICI estimator models the channel variation by means of a basis expansion model (BEM). The performance of the estimator and of the subsequent equalization is evaluated in an UMTS long term evolution (LTE) link level simulator. In a Rayleigh fading scenario, the proposed algorithm allows a velocity increase of 150 km/h without throughput degradation. The gain in terms of the post-equalization signal to interference and noise ratio (SINR) is about 3.7 dB at a user speed of 300 km/h.

The UMTS Long Term Evolution (LTE) allows the pilot symbol power to be adjusted with respect to that of the data symbols. Such power increase at the pilot symbols results in a more accurate channel estimate, but in turn reduces the amount of power available for the data transmission. In this paper, we derive optimal pilot symbol power allocation based on maximization of the post-equalization Signal to Interference and Noise Ratio (SINR) under imperfect channel knowledge. Simulation validates our analytical mode for optimal pilot symbol power allocation.

This paper deals with the symbol timing issue of an OFDM system in frequency selective fading scenarios. A modified symbol timing offset estimator based on an energy transition metric is proposed. This metric can also be applied to estimate the maximum channel delay spread. Benefiting from the knowledge of the maximum channel delay spread, the modified estimator shows its advantage in a relatively low SNR regime. Compared to the original, the modified algorithm is able to acquire the symbol timing correctly in a relatively low SNR region.

In the context of digital pre-distortion, a typical requirement is to identify the power amplifier with stringently low computational complexity. Accordingly, we consider a simple gradient method which is used to adaptively fit a simplified Wiener model, i.e., a cascade of a linear filter followed by a memoryless nonlinearity, to a dispersive and saturating reference system which represents the power amplifier. For adaptation, the gradient method only relies on the difference between the output of the reference system and the Wiener model. We show that such a structure can be formulated as a proportionate normalised least mean squares (PNLMS) algorithm. As a consequence, conditions for stability in the mean square sense can be deduced. Although not proven in a strict sense, simulation results allow to conjecture robustness.

The major contribution of this paper is the presentation of a general unifying description of distributed algorithms allowing to map local, node-based algorithms onto a single global, network-based form. As a first consequence the new description offers to analyze their learning and steady-state behavior by classical methods. A further consequence is the analysis of implementation issues as they appear due to quantization in computing and communication links. Exemplarily, we apply the new method on several different averaging algorithms: the Push-Sum protocol, average consensus as well as its quantized form and furthermore examine the effects of quantization noise which is introduced by the bandwidth limited communication links and finite precision computation ability of every node. Statistical properties of these quantization noises are provided and verified by simulations.

This paper has two major contributions. First, we present measured frequency selective wireless channels and show that there is only a small difference between mutual information and capacity. For single-antenna channels the maximum difference is about 0.15 bit/s/Hz, but typical differences are much smaller. These differences are increased with the number of transmit and receive antennas, achieving a maximum of about 0.4 bit/s/Hz for 2×2 MIMO channels and about 1.8 bit/s/Hz for 4×4 MIMO channels. The second contribution is a simple two-parameter channel model for frequency selective wireless channels, exhibiting the identical behavior in terms of capacity and mutual information than the measured channels. Due to its simplicity, this so-called poor-rich model is especially suitable for fast simulations and for analytical research in the field of information theory.

In this paper, we investigate carrier frequency synchronization in the downlink of 3GPP Long Term Evolution (LTE). A complete carrier frequency offset estimation and compensation scheme based on standardized synchronization signals and reference symbols is presented. The estimation performance in terms of mean square error is derived analytically and compared to simulation results. The impact of estimation error on the system performance is shown in terms of uncoded bit error ratio and physical layer coded throughput. Compared to perfect synchronization, the presented maximum likelihood estimator shows hardly any performance loss, even when the most sophisticated MIMO schemes of LTE are employed.

In order to evaluate the performance of new mobile network technologies, system level simulations are crucial. They aim at determining whether, and at which level predicted link level gains impact network performance. In this paper we present a MATLAB computationally efficient LTE system level simulator. The simulator is offered for free under an academic, noncommercial use license, a first to the authors' knowledge. The simulator is capable of evaluating the performance of the Downlink Shared Channel of LTE SISO and MIMO networks using Open Loop Spatial Multiplexing and Transmission Diversity transmit modes. The physical layer model is based on the postequalization SINR and provides the simulation pre-calculated "fading parameters" representing each of the individual interference terms. This structure allows the fading parameters to be pregenerated offline, vastly reducing computational complexity at run-time.

Arbitration and scheduling of multiple tags in state-of-the-art Radio Frequency Identification (RFID) systems is accomplished on the medium access control layer. Currently, only answers of a single tag can be decoded in such a system. If multiple tags respond simultaneously, a collision occurs. In that case, conventional systems discard the physical layer information and a retransmission is executed. This work shows how to recover from such collisions on the physical layer and successfully read the data. The contributions of the paper are: 1) An analysis of the achievable throughput increase of a system, that can recover from collisions at a physical layer is given. 2) A model for a description of collisions on the physical layer is presented. 3) Based on this model, we propose a channel estimation method and two types of receiver structures for separating the signals of a collision of two tags: first, single antenna receivers that discriminate the sources of the two tags in the I/Q plane, and second, multiple antenna receivers which exploit the different spatial signatures of both tags. 4) The functionality of the proposed receiver structures is verified with measurement data of two colliding tags. Eventually, a performance analysis of the receivers is provided.

Framed Slotted Aloha (FSA) is a popular anticollision technique in state-of-the-art RF-ID systems, as in ISO/IEC CD 18000-6 for 900MHz or the EPCglobal HF Gen 2 draft for 13.56MHz. In many applications the number of tags entering and leaving the detection range of the reader is subject to a strong fluctuation and usually unknown. The current number of tags in the field is a crucial parameter to operate the FSA anti-collision in an optimal manner. Therefore, a lot of effort is spent on the estimation of this parameter and a range of different estimation techniques exist. The contributions of this paper are: 1) a closed formula for the probability of any observed event defined by the number of empty, singleton, and collision slots in the observed frame is developed and empirically verified. 2) This formula is then modified to compute the probability for partly observed frames as well which is of great interest as the referred standards allow for the in-frame adjustment of the frame size without quitting the interrogation round. 3) Then, a maximum likelihood estimator is formulated to yield the estimated number of tags on a slot-wise basis. 4) Its superior estimation performance is compared to the known best estimators over the complete parameter set. While its performance is strongly superior compared to Schoute¿s estimate, compared to Vogt¿s MSE estimator only marginally improvement is obtained1.

It is known that Wideband Code-Division Multiple Access (W-CDMA) networks are limited by interference more than by any other single effect. Due to the frequency selectivity of the wireless channel, intra-cell interference composed of selfand other-user interference occurs. The intra-cell interference as a result of Transmit Antenna Array (TxAA) HSDPA multi-user transmissions, however, shows a special structure that can be exploited to allow for an efficient interference suppression. The contributions of this article are the following: we (a) propose a suitable system model to derive an intra-cell interference aware Minimum Mean Squared Error (MMSE) equalizer with moderate complexity and investigate its capabilities to suppress the multi-user intra-cell interference, (b) propose solutions to estimate the pre-coding state of the cell both with and without available training, and (c) investigate the resulting throughput performance with physical layer as well as system-level simulations. Our proposed solution can be interpreted as a multi-user extension of the classical MMSE equalizer for HSDPA systems without decoding the undesired users, resulting in only slightly increased complexity.

We provide a time domain analysis of the robustness and stability performance for coupled adaptive algorithms of gradient type. The considered coupling may occur inherently as well as by desire of the designer. The presented analyses focus on system identification. Examples are presented to investigate convergence and steady-state behaviour by simulations which are compared to theory. In particular, the presented approach allows for a deeper understanding of cascaded adaptive filters in terms of robustness and l2-stability.

In this paper we discuss possible benefits of transmitting SI frames as an error resilience tool in UMTS video streaming. SP and SI frames can be used to stop temporal error propagation, as an alternative to the regular insertion of I frames. Since their application relays on feedback information from the mobile equipment, different delay scenarios have been investigated. The performed transmission simulations show enhancement of the rate-distortion behaviour when using SP and SI frames with reduced feedback times. The error traces were measured at transport block level in a live UMTS mobile network.

Modern wireless communication systems employ MIMO and feedback-two properties that make it especially difficult to measure the performance of such systems with reasonable effort in actual outdoor scenarios. In this paper, we will present a time, cost, and manpower efficient measurement approach to evaluate the throughput achieved by such systems. Summer/winter, large distance outdoor-toindoor/outdoor, urban/alpine measurements have been carried out to successfully test this approach. Exemplarily, we report on the throughput gains (over TX power and base-station-antennarotation) of standard compliant 2 × 2 MIMO HSDPA and IEEE 802.16-2004 WiMAX when for example employing improved channel coding methods.

In this work, we provide a new throughput-based antenna selection criterion that considers the entire system design rather than only the channel state information. The performance evaluation was carried out in a measurement campaign in a common outdoor urban scenario. We present our results in terms of physical layer throughput over transmit power. In addition to diversity gains, we also see gains due to the selection of differently polarized antennas. In terms of SNR, the results show a gain of several decibels depending on the number of antennas selected.

In this work, we report on results of MIMO High Speed Downlink Packet Access (HSDPA) throughput measurements. These measurements were carried out in an urban scenario. Besides the standard compliant SISO and two transmit antenna schemes, we defined and measured also a four antenna scheme. In all throughput measurements we used link adaptation, that is, adaptation of the channel coding rate and adaptive preceding. The measured throughput is compared to a so-called "achievable throughput" that is calculated based on the mutual information of the channel measured.

In this paper, a tap-wise LMMSE channel estimator for MIMO W-CDMA is derived. Descrambling operations applied to delayed versions of the received signal whiten the input signal. The descrambling process thus breaks up the full channel autocorrelation matrix (including spatial and temporal correlation) into several independent, small autocorrelation matrices that include only spatial correlation. The problem of estimating the channel autocorrelation matrices is tackled by an "instantaneous" autocorrelation estimator that uses only one channel estimate (obtained by a very low complexity channel estimator) as input. Simulation results of a MIMO HSDPA system show that the tap-wise LMMSE estimator with instantaneous channel autocorrelation estimation yields performance gains of up to 0.85 dB over a correlation-based channel estimator. Even larger gains can be achieved with improved channel autocorrelation estimators.

In this paper, we present a novel framework for resampling and complexity reduction of tapped delay line channel models. In contrast to related algorithms in this field, our framework guarantees that the total impulse response power, the mean delay, as well as the RMS delay spread of every input impulse response remains unchanged if a solution is found. The phase angles of the output impulse response are chosen such that the magnitude error of the channel realization in the frequency domain is minimized as well.

In this article we propose an optimization method for encoding low resolution soccer video sequences. The considered video codec is the state of the art H.264/AVC, as recommended by the 3GPP specifications. The proposed optimization aims at maximizing the user perceived quality, considering the characteristic features of soccer videos. In the typical wide angle shots, about one fourth of each frame contains the grandstands. Even if, from the observerpsilas point of view, they can be considered as side information, the performed analysis showed that a prominent fraction of the resulting bandwidth is allocated to them. Our approach consists of a segmentation of the picture intended to recognize three components: the field, the audience and the remaining objects: the ball, the players and the field lines. The resulting map will be then utilized while encoding to optimize the bitrate assigned to each macroblock group, targeting the best resulting subjective perceived quality.

Various publications have drawn a precise picture on how correlation and capacity change when decreasing the receive-antenna distance of a 2 x 2 MIMO transmission to less than 0.1 lambda; however, little is known on how today's commonly employed coding schemes actually perform under such harsh conditions. In this paper, we evaluate - by measurements with a testbed - the uncoded BER over antenna distance performance of: spatial multiplexing, a linear dispersion code, the Golden Code, as well as Alamouti coding. We show, that even at very small antenna distances, a significant gain compared to an ordinary 1 x 2 SIMO transmission can be measured.

One of the advantages in an existing WCDMA network is that HSDPA (high speed downlink packet access) can be operated within an existing Release 4 (Release '99) 5 MHz band, thus allowing for sharing the power amplifier and spreading codes at the Node-B between the HS-DSCH (high speed downlink shared channel) and the Release 4 downlink channels (DCHs). In this paper, we investigate the throughput performance of HSDPA in a network, in which also Release 4 traffic is present by utilizing a system level modeling of the HSDPA transmission if a Rake receiver is used, together with a coarse modeling of the Release 4 performance. By numerical simulations, we are able to deduce the optimum power split between HSDPA and Release 4 to maximize the overall cell throughput. The obtained results offer a more sophisticated understanding of the HSDPA network performance, and can be used by network operators for cell operation planning.

The quality of end-to-end services over wireless communication networks highly depends on the error characteristics of the wireless links. By utilizing the link error characteristics for optimizing the higher layer protocols in a so called cross-layer approach, high quality improvements can be achieved, especially when considering the predictability of errors. In this document it is shown that wireless communication systems with quality based power control mechanisms entail such predictive nature of the link error characteristics. A method for obtaining estimated values of future link error probabilities in the UMTS (universal mobile telecommunications system) DCH (Dedicated CHannel) as an example is presented in this document. Furthermore, it is shown that various well-known and often used error models based on Markov chains are not capable of representing the predictive nature of the channel, whereas a two-state alternating Weibull renewal process describes the characteristics properly.

The loss of UDP/IP packets has a high impact on the quality of the low-rate video streaming, since one UDP/IP packet typically represents a considerable part of a picture, which is discarded and concealed if an error was detected by a UDP checksum. In this work we propose a method that allows for the utilization of error-free parts of such UDP packets. The detection of errors is facilitated by the use of CRC (cyclic redundancy check) information of the smaller link layer packets. Resynchronization of the variable length code at the beginning of the link layer packets is provided by using side information. To keep the additional rate small and to have low distortion at the same time, we propose sending the side information depending on the prediction of the channel quality. Experimental results show a considerable improvement in video quality at the receiver.

The popularity of MMOGs (massively multiplayer online games) has grown fast. This new type of online games uses a different type of protocol and has other limitations to delay and packet loss than the classical well established games like FPSs (first person shooters). This paper focuses on the traffic patterns generated by a MMOG named WoW (World of Warcraft). We analyzed IP layer information like packet size and rate from network traces and user level information like session durations. Finally we present a simple synthetic traffic generator for WoW traffic clients.

Approximation of Toeplitz matrices with cir-culant matrices is a well-known approach to reduce the computational complexity of linear equalizers. This paper presents a novel technique to compute linear equalizer coefficients in the frequency domain. It is shown how a regularization term can help to reduce the error caused by the frequency domain approximation. A corresponding VLSI implementation provides reference for the true silicon complexity and for the complexity increase associated with the proposed algorithm.

The quality of video streaming in an error-prone environment suffers from packet loss. Since the loss of a packet typically affects a rather large picture area, the performance of error concealment is limited, too. In order to avoid the huge overhead caused by using smaller packet sizes, in this paper we propose and analyze a scheme utilizing the residual redundancy of the encoded video stream. This scheme has two components -(i) syntax analysis implemented at the decoder, allowing exacter localization of errors within the packet and (ii) entropy code resynchronization mechanism based on the out-of-band signalized length indicators. We show that the proposed approach provides substantial improvement in PSNR for the same rate, compared to the standard packet size reduction.

Traffic models are to provision the increasing load of packet switched traffic in mobile 3G networks. However, most traffic models were designed to fit the need of low delay wired connections. In this paper we propose a new traffic model for POP3 e-mail traffic generation in 3G mobile networks scenarios. The model is based on traces which were extracted in a live 3G network of a mobile operator. The main improvement of the model is the explicit modeling of the login process that takes place between the client and the server. We implemented the model in the network simulation tool ns-2 and based on this we simulated various scenarios to visualize the impact of a leading initialisation phase in high RTT environments. The results proof that the extension can reproduce the effect from the login in a more precise way. Finally a comparison between the measured and the simulated service footprints were compared. The new model performs better than common models in this scenario.

The scope of this work is the estimation of video quality for low resolution video sequences typical in (mobile) video streaming applications. Since the video quality experienced by users depends considerably on the spatial (edges, colors, ...) and temporal (movement speed, direction, ...) features of the video sequence, this paper presents a two-step approach to quality estimation. Firstly, shots between two scene changes are analyzed and their content class is found. Secondly, based on the content class, frame rate and bitrate, an estimation of quality is carried out. In this paper, the design of the content classifier as well as an appropriate choice of the content classes and their characteristics is discussed. Moreover, the design of quality metric is presented, based on the mean opinion score obtained by a survey. The performance of the proposed method is evaluated and compared to several common methods. The results show that the proposed approach provides powerful means of estimating the video quality experienced by users for low resolution video streaming services.

In this paper we present results from a measurement campaign evaluating the web performance of different mobile Internet access technologies with and without caching. We measured in GPRS, EDGE, UMTS and HSDPA radio access networks of a mobile operator in Austria. The operator has implemented a web-proxy for performance improvement. To evaluate the impact on the different technologies we ran the setup with and without the web-proxy. We defined the download time of web pages as the performance index for our work. We developed a tool to capture this parameter and derived the index for a set of web sites frequently used by the customers in the network. Beginning with an over-all comparison we will present detailed analysis for special cases. We evaluated the impact of the proxy system to different site configurations using the performance index. Finally we compared the performance numbers for HTTP with results collected from FTP transfers.

We propose an iterative receiver for horizontally encoded MIMO systems like the PARC proposal for MIMO-HSDPA. Our receiver uses a K-best-MAP demodulator to provide the channel decoder with approximate log-likelihood ratios. The complexity of K-best-MAP is reduced by exploiting the CRC present in each encoded layer. We show that K-best-MAP with list update is superior to demodulation using a fixed candidate list. Finally, our numerical results demonstrate significant SNR gains as compared to a receiver using MMSE-based successive interference cancelation.

In this contribution we derive a union bound of the bit error ratio (BER) in wireless multiple input multiple output (MIMO) data transmission systems using maximum likelihood (ML) receivers. A novel channel model is used to take into account channel correlation. Relying on the Euclidean distance between different transmit signals, we show that the error performance is governed by a few so-called error types (ET). Simulation results are shown for a 4x4 MIMO system with BPSK modulation and for a 2x2 MIMO system with 16QAM modulation for several channel correlations. Our union bound becomes tight for low BER for all MIMO systems under investigation.

We present a high performance implementation of a belief propagation decoder for decoding low-density parity-check (LDPC) codes on a fixed point digital signal processor. A simplified decoding algorithm was used and a stopping criteria for the iterative decoder was implemented to reduce the average number of required iterations. This leads to an implementation with increased throughput compared to other implementations of LDPC codes or turbo codes. This decoder is able to decode at 5.4 Mbps on a Texas Instruments TMS320C64xx DSP running at 600 MHz.

Recently many space-time coding schemes for multiple antenna systems (MIMO) have been proposed in order to achieve high data rates when transmitting over wireless channels. However, most of such schemes rely on maximum likelihood (ML) detection, which can become quite complex when many antennas are involved and higher modulation schemes are utilized. On the other hand, the high diversity gains of MIMO channels are easily lost when low-complexity receivers like ZF or MMSE are applied. It is thus of utmost importance to look for low-complexity receivers which achieve almost ML performance. In this paper, a new scheme for approximate ML detection for typical flat Rayleigh fading channels is proposed, achieving ML performance in the area of bit error ratio between 10/sup -4/ and 0.1 as it is of interest in wireless communications. The proposed scheme allows the transmission of 64QAM schemes on 4/spl times/4 antenna schemes with a detection complexity of only 1% of a brute force ML receiver.

In this paper we present a family of quasi-orthogonal space-time codes for correlated MIMO (multi-input/multi-output) channels using four and eight transmit antennas when only partial channel state information is sent back to the transmitter. Transmit diversity is improved up to nearly the optimum diversity value. This high diversity can be exploited with zero forcing receivers as well as with maximum-likelihood receivers.

Multiple-input-multiple-output (MIMO) systems are under intense investigation for wireless transmission systems. Space-time coding is an efficient approach to exploit the huge diversity offered by the MIMO channel. One of the main problems of space-time block coding (STBC) is that the offered diversity can only be exploited by the optimal ML receiver which has exponential complexity. In contrast, linear receivers like the ZF-receiver are less complex, but achieve in general only a fraction of the offered diversity. In this paper, we show that it is also possible to achieve high diversity with linear receivers utilizing STBCs with a specific structure. We investigate the diversity properties of cyclic codes and extended Alamouti codes. It is shown, that the diversity order D increases asymptotically linear with the number of receive antennas (n/sub R/) and that the extended Alamouti code achieves already almost full diversity utilizing linear receivers and only small n/sub R/.

In this paper, new space-time coding schemes for multiple transmit and receive antenna systems (MIMO) are proposed. By applying a specific block coding scheme for transmission, the MIMO channel is structured in an advantageous manner. Diversity can be gained simultaneously with savings in receiver complexity. In a second step, such structured schemes and BLAST schemes are combined, offering and continuous trade-off between quality of service (QoS) and choice of data rate. Due to the simplicity of the coding schemes, only very moderate modifications in the existing UMTS standard are required in order to support them. If the transmitter knows the numbers of antenna elements at both sides of the radio link then this is sufficient for selecting the most appropriate scheme.

Based on power amplifier measurements a low-complex nonlinear model with memory, consisting of a memoryless, but parameterized nonlinearity, sandwiched between two linear FIR-filter is developed (L-N-L structure). Although the signal bandwidth is with 2 MHz rather small, memory effects can be noticed. Furthermore, an adaptive algorithm for continuous identification of the model parameters is developed. Simulations show that the algorithm converges in the mean-square sense to the global minimum, given only little knowledge regarding the region where this minimum lies.

In this paper, a multiple-input-multiple-output (MIMO) extension for a third-generation (3G) wireless system is described. The integration of MIMO concepts within the existing UMTS standard and the associated space-time RAKE receiver are explained. An analysis is followed by a description of an actual experimental MIMO transmitter and receiver architecture, both realized on digital signal processors (DSPs) and FPGAs within a precommercial OneBTS base station. It uses four transmit and four receive antennas to achieve downlink data rates up to 1 Mb/s per user with a spreading factor of 32 and the UMTS chip rate of 3.84 MHz. Furthermore, different MIMO detectors are evaluated, comparing their performance and complexity. System performance is evaluated through simulations and indoor over-the-air measurements. Capacity and bit-error rate measurement results are presented.

In state-of-the-art digital communication systems, channel estimation and/or equalization is an indispensable part of the system. Due to limited resources, only relatively simple algorithms with low complexity can be applied for channel estimation and equalization in high data-rate systems. In fact, the complexity of these algorithms can become prohibitive for very high bit-rate applications, for example, several 100 Msymbols-per-second. We explore the signal constellation structure of practical digital communication systems and describe an efficient computation technique that not only eliminates multiplications but also minimizes the number of addition operations required to implement channel estimation algorithms such as the least mean squares (LMS) estimation and equalization algorithms such as the maximum likelihood (ML) sequence detection using the Viterbi algorithm. This new technique preserves the numerical precision of the algorithms while it reduces their complexity dramatically.

This paper studies the behavior of the error sequence of stop-anti-go variants of two adaptive blind equalizers, namely CMA2-2 and Sato's (1975) algorithm. It is shown that for transmitted signals with constant modulus /spl gamma/, the equalizer output can be made to lie within the circle of radius /spl gamma//spl radic/c infinitely often, for some value of c that is only slightly larger than one.

This paper proposes two modifications of the filtered-x least mean squares (FxLMS) algorithm with improved convergence behavior albeit at the same computational cost of 2M operations per time step as the original FxLMS update. The paper further introduces a generalized FxLMS recursion and establishes that the various algorithms are all of filtered-error form. A choice of the stepsize parameter that guarantees faster convergence and conditions for robustness are also derived. Several simulation results are included to illustrate the discussions.

Optimum detection in randomly time-varying channels requires an efficient adaptive receiver structure. The complexity of signal processing in the receiver is limited by the amount of processing power and power consumption of the receiver. Therefore, efficiency and convergence of adaptive matched filtering and equalization techniques are very important. We extend common results in tracking theory for system identification to the equalization case for systems with short impulse response. Unlike in system identification where a steady-state error energy is minimized, the optimization criterion here is the minimum of the BER. However, they are related by a monotone function and therefore minimizing the BER is equivalent to minimizing the steady-state-error energy. Optimum parameters for LMS as well as RLS algorithms are derived and simulation results indicate that under the conditions defined in the TDMA standards and small delay spread, the performance of the two methods is comparable.

Recently, two simple gradient-based algorithms for unbiased IIR system identification in the presence of zero-mean correlated output noise were derived and shown to perform well in simulation. In this paper, we study the stability and robustness of these two adaptive filters, deriving strictly positive real (SPR) conditions on the overall unknown-plus-adaptive systems to guarantee convergence of the coefficients to their optimum values. Unlike other algorithms for unbiased IIR adaptive filtering, the stability of each of these algorithms depends on the initial values of the filter coefficients. However, near the optimum coefficient solutions, both algorithms are locally-stable, irrespective of the unknown system. Simulations verify the results of our analyses.

This paper provides a time-domain feedback analysis of the perceptron learning algorithm and of training schemes for dynamic networks with output feedback. It studies the robustness performance of the algorithms in the presence of uncertainties that might be due to noisy perturbations in the reference signals or due to modeling mismatch. In particular, bounds are established on the step-size parameters in order to guarantee that the resulting algorithms will behave as robust filters. The paper also establishes that an intrinsic feedback structure can be associated with the training schemes. The feedback configuration is motivated via energy arguments and is shown to consist of two major blocks: a time-variant lossless (i.e., energy preserving) feedforward path and a time-variant feedback path. The stability of the feedback structure is then analyzed via the small gain theorem, and choices for the step-size parameter in order to guarantee faster convergence are deduced by using the mean-value theorem. Simulation results are included to demonstrate the findings.

Gradient-descent adaptive algorithms are among the most widely used in current practice, with many different variants that generally fit into two major groups: one group includes algorithms that are especially suited for FIR (or finite-impulse-response) modeling, while the other group includes algorithms that are tailored for IIR (or infinite-impulse-response) modeling. In the first group, the regression (or data) vectors do not depend on the unknown parameters, which leads to convenient linear models that often facilitate the analysis of the algorithms. In the second category, on the other hand, the regression vectors are dependent on the unknown parameters, thus giving rise to nonlinear functionals and to a richer structure that requires a more thorough analysis. This paper focuses on a widely used adaptive IIR algorithm, the so-called Feintuch's (1976) algorithm, and provides a study of its robustness, stability, and convergence properties in a deterministic framework.

A comparison of adaptive IIR (infinite impulse response) filters with gradient-based adaptation algorithms is presented. The following algorithms were investigated: series parallel LMS (SP-LMS), equation error formulation LMS (EEF-LMS), bias-remedy LMS (BRLMS) alternate filtering mode (AFM), simple hyperstable adaptive recursive filter (SHARF), and normalized LMS (NLMS), which served as an FIR comparison case. The classification structure employed clearly illustrates the relationships of the algorithms to each other; additionally, other feasible filter methodologies for further investigation were revealed. All algorithms were implemented on a Motorola 56001. Correct normalization of the adaptation stepsize played a critical role in the results, which were obtained by real-time measurements. Only the SHARFS and BRLMS algorithms fulfil the requirements of a low-cost hybrid echo canceller.>