FSFrank Suits

Frank Suits

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Frank Suits has a total of 230 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Dynamic Edge Fabric EnvironmenT: Seamless and Automatic Switching among Resources at the Edge of IoT Network and Cloud

    by Fatemeh Jalali, Timothy Lynar, Olivia J. Smith, Ramachandra Rao Kolluri, Claire V. Hardgrove, Nick Waywood, Frank Suits

    The number of IoT devices at the edge of the network is increasing rapidly. Data from IoT devices can be analysed locally at the edge, or they may be sent to the cloud. Currently, the decision to deploy a task for execution at the edge or in the cloud is not decided as the data are received. Instead the decision is usually based on pre-defined system design and corresponding assumptions about locality and connectivity. However, the mobile environment has rapid, sometimes unpredictable changes and requires a system that can dynamically adapt to these changes. An intelligent platform is required that can discover available resources (both nearby and in the cloud) and autonomously orchestrate a seamless and transparent task allocation at runtime to help the IoT devices achieve their best performance given the available resources. We propose a new platform, DEFT (Dynamic Edge-Fabric environmenT), that can automatically learn where best to execute each task based on real-time system status and task requirements, along with learned behavior from past performance of the available resources. The task allocation decision in this platform is powered by machine learning techniques such as regression models (linear, ridge, Lasso) and ensemble models (random forest, extra trees). We have implemented this platform on heterogeneous devices and run various IoT tasks on the devices. The results reveal that choosing proper machine learning approaches based on the tasks properties and priorities can significantly improve the overall performance of selecting resources (either from the edge or cloud) dynamically at runtime.


  • Optimal power purchase agreements in PV-rich communities

    by Nihan Cicek van der Heijden, Tansu Alpcan, Eduardo Alejandro Martinez-Cesena, Frank Suits

    This paper studies small PV-rich communities that wish to purchase their electricity directly from generators through long-term power purchase agreements (PPA) and the spot market if necessary. Net demand, defined as the difference between total demand and solar generation, is satisfied through PPA and the spot market. Due to the random nature of solar generation, net demand is random and hard to estimate. This randomness may lead to a mismatch between the actual net demand and PPA. When net demand exceeds PPA, electricity is purchased at the spot market prices causing increased operational costs. To avoid this mismatch and utilise solar potential efficiently, we propose a probabilistic optimisation approach using the statistical properties of solar generation. We model solar energy generation as a random variable for each daylight hour at a chosen location using historical solar irradiance data. By transforming random variables, we find the probability distribution of the net demand which is used to propose a market exposure probability limiting optimisation problem. The overall objective of the optimisation problem is to minimise operational costs of communities. Unlike forecasting and Monte Carlo Simulation, our methodology enables market exposure risk analysis, fine-tuning PPA and a good understanding of statistical properties of solar generation.


  • Open Access

    Cognitive IoT Gateways

    by Fatemeh Jalali, Olivia J. Smith, Timothy Lynar, Frank Suits

    Fog computing, also known as Edge computing, is an emerging computational paradigm, increasingly utilized in Internet of Things (IoT) applications, particularly those that cannot be served efficiently using Cloud computing due to limitations such as bandwidth, latency, Internet connectivity. At present, the norm is the static allocation of tasks by developers of an application, where some IoT applications are allocated to be performed on the Cloud, some on the Fog, and some on a hybrid Cloud-Fog. The applications are pre-programmed and predefined to be run on a platform, and this is unchangeable at run-time. IoT gateways, which are devices that bridge the IoT local network and the Internet, are in a position to make dynamic adjustments and allocation decision between platforms based upon real-time conditions such as an IoT applications' performance. However, currently there is no (or very little) intelligence embedded into IoT gateways. This paper proposes cognitive IoT gateways powered by cognitive analytics and machine learning to improve the performance of IoT applications. These IoT devices are able to automatically learn and decide when and where to run an application, be that on the Cloud or on the Fog. The dynamic task sharing and platform interchanging will enable the IoT applications to be optimized for multiple objectives including task performance.


  • Greening IoT with Fog: A Survey

    by Fatemeh Jalali, Safieh Khodadustan, Chrispin Gray, Kerry Hinton, Frank Suits

    The current growth in Internet services, mobile devices, and machine-to-machine (M2M) technologies is providing the building blocks for the Internet of Things (IoT) as it is being applied across all industry sectors. With ongoing proliferation of IoT applications, a new platform called Fog/edge computing, in addition to Cloud computing, is being developed to address requirements such as bandwidth, latency and location awareness. As with previous many telecommunication systems, energy consumption concerns in IoT have been deferred to the point that it may become a bottleneck in the future. This work conducts a survey of existing literature addressing IoT energy consumption growth. We firstly highlight the factors and technologies in the system design, application layer and network virtualizations which lead to higher or lower energy consumption of an IoT service. Furthermore, we report strategies that can help to alleviate power consumption of IoT applications and services using Fog computing. Our objective is to provide a survey for network designers and policy makers who wish to gain an insight into deploying energy-efficient IoT applications.


  • Interconnecting Fog computing and microgrids for greening IoT

    by Fatemeh Jalali, Arun Vishwanath, Julian de Hoog, Frank Suits

    The Internet of things (IoT) is hailed as the next big phase in the evolution of the Internet. IoT devices are rapidly proliferating, owing to widespread adoption by industries in various sectors. Unlike security and privacy concerns, the energy consumption of the IoT and its applications has received little attention by the research community. This paper explores different options for deploying energy-efficient IoT applications. Specifically, we evaluate the use of a combination of Fog computing and microgrids for reducing the energy consumption of IoT applications. First, we study the energy consumption of different types of IoT applications (such as IoT applications with differing traffic generation or computation) running from both Fog and Cloud computing. Next, we consider the role of local renewable energy provided by microgrids, and local weather forecasting, along with Fog computing. To evaluate our proposal, energy consumption modeling, practical experiments and measurements were performed. The results indicate that the type of IoT application, the availability of local renewable energy, and weather forecasting all influence how a system makes dynamic decisions in terms of saving energy.


  • Proteogenomics: Key Driver for Clinical Discovery and Personalized Medicine

    by Ruggero Barbieri, Victor Guryev, Corry-Anke Brandsma, Frank Suits, Rainer Bischoff, Peter Horvatovich

    Proteogenomics is a multi-omics research field that has the aim to efficiently integrate genomics, transcriptomics and proteomics. With this approach it is possible to identify new patient-specific proteoforms that may have implications in disease development, specifically in cancer. Understanding the impact of a large number of mutations detected at the genomics level is needed to assess the effects at the proteome level. Proteogenomics data integration would help in identifying molecular changes that are persistent across multiple molecular layers and enable better interpretation of molecular mechanisms of disease, such as the causal relationship between single nucleotide polymorphisms (SNPs) and the expression of transcripts and translation of proteins compared to mainstream proteomics approaches. Identifying patient-specific protein forms and getting a better picture of molecular mechanisms of disease opens the avenue for precision and personalized medicine. Proteogenomics is, however, a challenging interdisciplinary science that requires the understanding of sample preparation, data acquisition and processing for genomics, transcriptomics and proteomics. This chapter aims to guide the reader through the technology and bioinformatics aspects of these multi-omics approaches, illustrated with proteogenomics applications having clinical or biological relevance.


  • Optimal Operation of Energy Storage Systems Considering Forecasts and Battery Degradation

    by Khalid Abdulla, Julian de Hoog, Valentin Muenzel, Frank Suits, Kent Steer, Andrew Wirth, Saman Halgamuge

    Energy storage systems have the potential to deliver value in multiple ways, and these must be traded off against one another. An operational strategy that aims to maximize the returned value of such a system can often be significantly improved with the use of forecasting - of demand, generation, and pricing - but consideration of battery degradation is important too. This paper proposes a stochastic dynamic programming approach to optimally operate an energy storage system across a receding horizon. The method operates an energy storage asset to deliver maximal lifetime value, by using available forecasts and by applying a multi-factor battery degradation model that takes into account operational impacts on system degradation. Applying the method to a dataset of a residential Australian customer base demonstrates that an optimally operated system returns a lifetime value which is 160% more, on average, than that of the same system operated using a set-point-based method applied in many settings today.


  • Identification of Analytical Factors Affecting Complex Proteomics Profiles Acquired in a Factorial Design Study with Analysis of Variance: Simultaneous Component Analysis

    by Vikram Mitra, Natalia Govorukhina, Gooitzen Zwanenburg, Huub Hoefsloot, Inge Westra, Age Klaas Smilde, Theo H. Reijmers, Ate G. J. van der Zee, Frank Suits, Rainer Bischoff, et al.

    Complex shotgun proteomics peptide profiles obtained in quantitative differential protein expression studies, such as in biomarker discovery, may be affected by multiple experimental factors. These pre-analytical factors may affect the measured protein abundances which in turn influence the outcome of the associated statistical analysis and validation. It is therefore important to determine which factors influence the abundance of peptides in a complex proteomics experiment, and to identify those peptides that are most influenced by these factors. In the current study we analysed depleted human serum samples to evaluate experimental factors that may influence the resulting peptide profile such as the residence time in the autosampler at 4˚C, stopping or not stopping the trypsin digestion with acid, the type of blood collection tube, different haemolysis levels, differences in clotting times, the number of freeze-thaw cycles and different trypsin/protein ratios. To this end we used a two-level fractional factorial design of resolution IV (27-3IV). The design required analysis of 16 samples in which the main effects were not confounded by two-factor interactions. Data pre-processing using the Threshold Avoiding Proteomics Pipeline1 produced a data-matrix containing quantitative information on 2,559 peaks. The intensity of the peaks was log-transformed, and peaks having intensities of a low t-test significance (p-value > 0.05) and a low absolute fold ratio (< 2) between the two levels of each factor were removed. The remaining peaks were subjected to ANOVA – simultaneous component analysis (ASCA)2. Permutation tests were used to identify which of the pre-analytical factors influenced the abundance of the measured peptides most significantly. The most important pre-analytical factors affecting peptide intensity were (1) the haemolysis level, (2) stopping trypsin digestion with acid and (3) the trypsin/protein ratio. This provides guidelines for the experimentalist to keep the ratio of trypsin/protein constant and to control the trypsin reaction by stopping it with acid at an accurately set pH. The haemolysis level cannot be controlled tightly as it depends on the status of a patient’s blood (e.g. red blood cells are more fragile in patients undergoing chemotherapy) and the care with which blood was sampled (e.g. by avoiding shear stress). However, its level can be determined with a simple UV spectrophotometric measurement and samples with extreme levels or the peaks affected by haemolysis can be discarded from further analysis. The loadings of the ASCA model led to peptide peaks that were most affected by a given factor, for example to haemoglobin-derived peptides in the case of the haemolysis level. Peak intensity differences for these peptides were assessed by means of extracted ion chromatograms confirming the results of the ASCA model.