Combining PHY experimental results and advanced network planning tool for accurate coverage estimation

After 5 years invested in developing accurate cost estimation tools, Rolls-Royce has learnt about the larger potential of the tools to shape many aspects of the business. A good estimation tool is a model of a project and is usually used to estimate cost and schedule, but it can also estimate and validate risks and opportunities. Estimation tools have unified engineering, project and business needs. The presence of good estimation tools has driven higher performance and stability in the business. It was evident we needed this capability to underpin decisions in our new Software Product Line strategy. The objective of this paper is twofold. First, we report the experiences gained in the past on the use of estimation tools. Second, we describe the current efforts and future directions on the development of an estimation tool for Software Product Lines. At the heart of the Product Line estimation tool is a simple representation of the product - represented as the number of Lines Of Code (LOC). The next generation of tool, will need to consider wider aspects of product quality in order to create more accurate estimates and support better decisions about our products.

The importance of efficient area and timing estimation techniques for hierarchical design methodology is well-established in High-Level Synthesis (HLS), since the estimation allows more realistic exploration of the design space, and hierarchical design methodology matches well with HLS paradigm. In this paper, we present ChipEst-FPGA, a chip level estimator for designs implemented using a hierarchical design methodology for Lookup Table Based FPGAs. In FPGAs, the wire delay may contribute to a significant portion of the overall design delay. ChipEst-FPGA uses a realistic model which takes the component area/delay as well as wiring effects into account. We tested our ChipEst-FPGA on several benchmarks and the results show that we can get accurate area and timing estimates efficiently.

We consider the problem of parameter estimation from real-valued multi-tone signals. Such problems arise frequently in spectral estimation. More recently, they have gained new importance in finite-rate-of-innovation signal sampling and reconstruction. The annihilating filter is a key tool for parameter estimation in these problems. The standard annihilating filter design has to be modified to result in accurate estimation when dealing with real sinusoids, particularly because the real-valued nature of the sinusoids must be factored into the annihilating filter design. We show that the constraint on the annihilating filter can be relaxed by making use of the Hilbert transform. We refer to this approach as the Hilbert annihilating filter approach. We show that accurate parameter estimation is possible by this approach. In the single-tone case, the mean-square error performance increases by 6 dB for signal-to-noise ratio (SNR) greater than 0 dB. We also present experimental results in the multi-tone case, which show that a significant improvement (about 6 dB) is obtained when the parameters are close to 0 or pi. In the mid-frequency range, the improvement is about 2 to 3 dB.

An accurate and fast estimation of the available bandwidth in a network with varying cross-traffic is a challenging task. The accepted probing tools, based on the fluid-flow model of a bottleneck link with first-in, first-out multiplexing, estimate the available bandwidth by measuring packet dispersions. The estimation becomes more difficult if packet dispersions deviate from the assumptions of the fluid-flow model in the presence of non-fluid bursty cross-traffic, multiple bottleneck links, and inaccurate time-stamping. This motivates us to explore the use of machine learning tools for available bandwidth estimation. Hence, we consider reinforcement learning and implement the single-state multi-armed bandit technique, which follows the ε-greedy algorithm to find the available bandwidth. Our measurements and tests reveal that our proposed method identifies the available bandwidth with high precision. Furthermore, our method converges to the available bandwidth under a variety of notoriously difficult conditions, such as heavy traffic burstiness, different cross-traffic intensities, multiple bottleneck links, and in networks where the tight link and the bottleneck link are not same. Compared to the piece-wise linear network a model based direct probing technique that employs a Kalman filter, our method shows more accurate estimates and faster convergence in certain network scenarios and does not require measurement noise statistics.

The timely knowledge of forage quality of grasslands is vital for matching the demands in animal feeding. Remote sensing (RS) is a promising tool for estimating field-scale forage quality compared with traditional methods, which usually do not provide equally detailed information. However, the applicability of RS prediction models depends on the variability of the underlying calibration data, which can be brought about by the inclusion of a multitude of grassland types and management practices in the model development. Major aims of this study were (i) to build forage quality estimation models for multiple grassland types based on an unmanned aerial vehicle (UAV)-borne imaging spectroscopy and (ii) to generate forage quality distribution maps using the best models obtained. The study examined data from eight grasslands in northern Hesse, Germany, which largely differed in terms of vegetation type and cutting regime. The UAV with a hyperspectral camera on board was utilised to acquire spectral images from the grasslands, and crude protein (CP) and acid detergent fibre (ADF) concentration of the forage was assessed at each cut. Five predictive modelling regression algorithms were applied to develop quality estimation models. Further, grassland forage quality distribution maps were created using the best models developed. The normalised spectral reflectance data showed the strongest relationship with both CP and ADF concentration. From all predictive algorithms, support vector regression provided the highest precision and accuracy for CP estimation (median normalised root mean square error prediction (nRMSEp) = 10.6%), while cubist regression model proved best for ADF estimation (median nRMSEp = 13.4%). The maps generated for both CP and ADF showed a distinct spatial variation in forage quality values for the different grasslands and cutting regimes. Overall, the results disclose that UAV-borne imaging spectroscopy, in combination with predictive modelling, provides a promising tool for accurate forage quality estimation of multiple grasslands.

The key importance of steamflood in a heavy oil reservoir is to provide heating at the right place. The provision of heating becomes a challenge in heterogeneous reservoirs, as it may create some unswept areas. Kangen Sand in Area X represents a heterogeneous reservoir that is indicated by high Vshale content, which leads to various times of steam breakthrough, injectivity issues and high-pressure patterns. This paper shows how Area X deploys a realistic steamflood conceptual model, which represents actual steamflood response corresponding to its reservoir heterogeneity. The study began with a quantitative approach to understand the reservoir static parameters that have the most influence with actual steamflood performance. A drawing of aerial steam distribution was done to understand the area or pattern with slow progress in steam distribution compared to other patterns, and to conduct a deep dive on relationship between slow heating and reservoir heterogeneity, a realistic steamflood conceptual model was then developed, taking into account the associated response and understanding of unforeseen reservoir heterogeneity. The updated steamflood conceptual model has led the team to develop an optimization plan by redirecting heating to unswept areas. Also to acquire new surveillance to update the conceptual model. Redirecting steam injection by performing additional perforations in the injectors has successfully improved heating direction to unswept areas, which is confirmed by a new trend of wellhead temperature in producers and temperature response in observation wells. The leading indicators show an increase in well injectivity, and spinner survey confirmed additional perforated injection. The lagging indicators also show a pattern of oil production increase and no subsurface high-pressure pattern has been recently recorded. This study shows that a conceptual model validated with actual condition can easily apprise unforeseen reservoir heterogeneity, and the optimization plan can be selected to the right place of reservoir target.

Chapter 3 - The Method Debate: A State-of-the-Art Analysis of PMI Estimation Techniques

Population pharmacokinetic studies of ciclosporin microemulsion are needed to identify the individual factors influencing ciclosporin pharmacokinetic variability in transplant patients and to design efficient tools for the accurate estimation of ciclosporin overall exposure (area under the plasma concentration-time curve from 0 to 12 hours [AUC12]). In the present retrospective study, a large database of heart, lung (with or without cystic fibrosis) and kidney (both adult and paediatric) transplant patients receiving ciclosporin microemulsion was analysed with the aims of (i) building a population pharmacokinetic model and finding the main covariates linked with ciclosporin microemulsion pharmacokinetic parameters; and (ii) developing a maximum a posteriori probability Bayesian estimator (MAP-BE) to estimate ciclosporin microemulsion pharmacokinetic parameters using a limited-sampling strategy. 3,072 concentration data from 147 patients (i.e. 309 full pharmacokinetic profiles) were analysed using the nonlinear mixed-effects model program NONMEM. The influence of numerous covariates was tested, and the final model was validated by data splitting. For Bayesian estimation, the best limited-sampling strategy was determined based on the D-optimality criterion, and validation performed in an independent group of 60 patients. The pharmacokinetics of ciclosporin microemulsion were accurately described by a two-compartment model with Erlang distribution for the absorption process. The type of graft and post-transplantation period were identified as significant sources of variability of the absorption parameter. Both apparent volume of the central compartment after oral administration (V1/F) and apparent oral clearance (CL/F) increased with bodyweight. The best limited-sampling strategy for Bayesian estimation was 0 hour, 1 hour and 3 hour post-dose, providing accurate estimation of ciclosporin microemulsion AUC12 in all patients of the test group, with a mean bias of 2.0 +/- 10.5% (range: -19.1% to -21.4% and 95% CI -0.6, +4.7). Population pharmacokinetic analysis of ciclosporin microemulsion in allograft transplants resulted in the design of a new pharmacokinetic model for ciclosporin microemulsion, identification of significant covariates and the design of an accurate MAP-BE based on three blood concentrations and these covariates.

The Tengiz reservoir is known for its substantial bitumen content, requiring careful differentiation during formation evaluation processes. To accurately estimate the effective porosity, wireline nuclear magnetic resonance (NMR) tools were used in conjunction with other logging tools. However, the increasing number of deviated wells drilled in Tengiz posed conveyance challenges for the wireline-based NMR tool due to high deviation angles. In light of this, TCO (Tengizchevroil) geoscientists sought opportunities to acquire NMR data using logging-while-drilling (LWD) technology in the reservoir section. Although the LWD NMR tool was not previously available for the required hole size, recent commercialization made it feasible. A team of geoscientists at TCO meticulously planned and executed a field trial of the LWD NMR tool to assess its potential future utilization in situations where the conveyance of wireline NMR tools may encounter difficulties. The LWD NMR tool was successfully deployed in a deviated well during drilling operations, followed by a subsequent measurement after drilling (MAD) pass to ensure log quality control. To verify and compare the acquired data, wireline-based NMR measurements were also conducted in the same well. The results of the field trial demonstrated the successful acquisition of high-quality LWD NMR data, corroborated by the MAD pass data and wireline log data. Furthermore, we revised and tested internal NMR data processing workflows, which conclusively established LWD NMR as a viable alternative to wireline-based NMR tools. Notably, the acquisition of NMR data while drilling offers additional benefits, as the near wellbore zone is not fully flushed with drilling mud, potentially enhancing fluid typing techniques (T1 and T2 distribution). This promising outcome opens up new avenues for more efficient and reliable estimation of effective porosity in deviated wells, as well as in wells with potential restrictions to run WL tools within the Tengiz reservoir and beyond, contributes to improved reservoir characterization and decision-making processes in asset development.

An explicit variance reduction expression for the Rao-Blackwellised particle filter

In simulating acoustic source localization by matched‐field processing, it is desirable to use a realistic ambient noise model. Two general classes of noise model were considered: a planar isotropic noise source distribution and a line source distribution. The former was implemented based on the work of Kuperman and Ingenito [J. Acoust. Soc. Am. 67, 1988–1996 (1980)]. Their expressions for the correlation integrals were adapted to compute the correlation function for both simple and realistic propagation models, through the use of analytical and numerical Green's functions, respectively. The line source model involved the assumption of uncorrelated noise sources along an infinite line and was designed to model a shear zone in a shallow ice‐covered ocean. Using the corresponding Green's functions for a waveguide, the line model correlation function could be reduced to a summation over all mode pairs, each term of which involved an integration over wavenumber. In this paper, the line source correlation model is presented and correlation functions given by both noise models are illustrated. The effect of using these models on the localization of a discrete source by a grid search‐optimization procedure is then examined.

Accurate estimation of aggregate interference is crucial for understanding the potential capacity gains from densification of cells. In this paper we study the impact of three-dimensionality of femtocell deployments on both cross-tier and co-tier aggregate interference using realistic building data and accurate propagation models. Our results show that the three-dimensional character of urban femtocell deployments can result in significantly higher interference effects compared to typically used two-dimensional evaluation scenarios. Further, we also show that heterogeneity in building heights plays an important role in the amount of aggregate interference, and should be considered more carefully when specifying standard topologies for interference studies.

We study the problem of joint user association and intercell interference (ICI) mitigation in heterogeneous long-term evolution advanced (LTE-A) networks, presuming the absence of the precise global channel state information (CSI). We confront this problem from a contract-theoretic perspective and model the considered network as a labor market where the LTE evolved NodeBs (eNBs) are regarded as the employers offering a contract to the employees (cellular users). We explore a scenario where the wireless channels are ordered into different types according to their link gains and the cost of power consumption. It is assumed that each user knows the type of its channel but is not aware of the types of the channels belonging to other users. Hence, the channel type is regarded as the private information which is known only by the corresponding user. Under the presence of such asymmetric information, the eNBs formulate the optimal contract given by a set of the contract items (the combinations of user association, allocated spectrum, and power) and pass these items to the users, who then select the best contract items according to their channel types. The performance of a proposed contract-based resource allocation mechanism has been evaluated using the OPNET-based simulations. It has been demonstrated that the efficiency of a contract-based mechanism with asymmetric information is close to the optimal method (where the network resources are allocated based on the precise CSI). It has also been shown that the contract-based method outperforms the other relevant resource allocation techniques in terms of the total network service rate and signal-to-noise ratio.

This paper introduces GLAAPE, a novel graph learning-assisted neural network model that enables fast, accurate, and transferable estimation of average power in gate-level combinational designs from Register Transfer Level (RTL) simulation. GLAAPE learns to propagate the average toggle rates by effectively embedding the structural information and logical function feature values as vectors on each logic cell of the netlist file. GLAAPE predicts the output toggle rates of combinational logic cells and hence the average power of the benchmark circuits from inputs obtained through RTL simulation without the requirement of gate-level simulation. We evaluate GLAAPE with the commercial gate-level power estimation tool for inference throughput, which is, on average, 15.69X faster. GLAAPE outperforms the commercial RTL Power estimation tool for average power estimation by achieving a mean improvement of 25.28%. We also compare GLAAPE with the state-of-the-art model GRANNITE to predict the output toggle rates to demonstrate the transferability and achieve better accuracy with a mean improvement of 13.95%. We consider a good range of benchmark circuits for analysis, and our experimental results show the generalization and efficacy capability of GLAAPE.

Power spectral density (PSD) of the long-term, 24-hour heart rate variability (LHRV) is widely used for risk stratification and noninvasive assessment of autonomic balance. The method can correctly identify the total power changes, but spectral estimates over narrow frequency ranges are inaccurate because of nonstationary character of the signal. An energy-frequency distribution (EFD) of LHRV based on a wavelet transform is proposed to overcome the nonstationarity problem and to obtain accurate spectral estimates over narrow frequency ranges. Holter ECGs from 54 patients with chronic autonomic imbalance and 27 normal subjects were used to derive the LHRV-signals. Mean EFD showed a characteristic peak that was shifted toward lower frequencies in patients compared to normal subjects. The slope of EFD at the high frequency ranges (>0.0329 Hz) was smaller in patients than in healthy subjects. EFD of 24-hour heart rate variability might be used as a more accurate tool for spectral estimation of the nonstationary frequency components than the power spectral integration over empirical frequency ranges.