Armstrong, N, Sutton, GJ & Hibbert, DB 2019, 'Estimating probability density functions using a combined maximum entropy moments and Bayesian method. Theory and numerical examples', Metrologia, vol. 56, no. 1.View/Download from: UTS OPUS or Publisher's site
© 2019 BIPM & IOP Publishing Ltd. Estimating the probability density function (pdf) from a limited sample of data is a challenging data analysis problem. Furthermore, determining which pdf best describes the available data involves an extra layer of complexity to the analysis, which if ignored, can have considerable consequences. We propose a combined maximum entropy (MaxEnt) moments and Bayesian model selection method to address this problem. The MaxEnt moments component is used to formulate a set of possible pdf models, each constrained by a different set of moments and parameterised by a set of Lagrangian multipliers. The Bayesian model selection component makes an inference about the most probable model, from the set of MaxEnt moment models. The structure of the prior pdf for the Lagrangian multipliers is determined from an expansion of the free energy functional for each MaxEnt model, and corresponding hyperparameters are calculated empirically. Numerical experiments were used to test the proposed method on samples taken from Gaussian and (more complex) non-Gaussian distributions, over a range of sample sizes. The results clearly demonstrate that the method can discriminate between simple and complex MaxEnt models for sample sizes approximately greater than 60. Our results demonstrate that MaxEnt and Bayesian methods are complementary. More critically, Bayesian inference is necessary when a set of competing MaxEnt models can be derived for a single dataset from a range of assumptions.
Zhou, W, Sutton, GJ, Zhang, JA, Liu, RP & Pan, S 2019, 'Delay-Guaranteed Admission Control for LAA Coexisting With WiFi', IEEE WIRELESS COMMUNICATIONS LETTERS, vol. 8, no. 4, pp. 1048-1051.View/Download from: UTS OPUS or Publisher's site
Sutton, GJ, Zeng, J, Liu, RP, Ni, W, Nguyen, DN, Jayawickrama, BA, Huang, X, Abolhasan, M, Zhang, Z, Dutkiewicz, E & Lv, T 2019, 'Enabling Technologies for Ultra-Reliable and Low Latency Communications: From PHY and MAC Layer Perspectives', IEEE Communications Surveys & Tutorials, pp. 1-1.View/Download from: UTS OPUS or Publisher's site
Sutton, G, Liu, RP & Guo, J 2019, 'Coexistence Performance and Limits of Frame-Based Listen-Before-Talk', IEEE Transactions on Mobile Computing, pp. 1-1.View/Download from: UTS OPUS or Publisher's site
Sutton, G, Zeng, J, Liu, R, Ni, W, Nguyen, D, Jayawickrama, B, Huang, X, Abolhasan, M & Zhang, Z 2018, 'Enabling Ultra-Reliable and Low Latency Communications through Unlicensed Spectrum', IEEE Network, vol. 32, no. 2, pp. 70-77.View/Download from: UTS OPUS or Publisher's site
In this article, we aim to address the question of how to exploit the unlicensed spectrum to achieve URLLC. Potential URLLC PHY mechanisms are reviewed and then compared via simulations to demonstrate their potential benefits to URLLC. Although a number of important PHY techniques help with URLLC, the PHY layer exhibits an intrinsic trade-off between latency and reliability, posed by limited and unstable wireless channels. We then explore MAC mechanisms and discuss multi-channel strategies for achieving low-latency LTE unlicensed band access. We demonstrate, via simulations, that the periods without access to the unlicensed band can be substantially reduced by maintaining channel access processes on multiple unlicensed channels, choosing the channels intelligently, and implementing RTS/CTS.
© 2013 IEEE. With the release of the 5 GHz unlicensed spectrum has emerged licensed-Assisted access, in which long-Term evolution (LTE) operators compete with Wi-Fi users for a share of the unlicensed spectrum so as to augment their licensed spectrum. Subsequently, there has been the need to develop a LTE channel access mechanism that enables harmonious coexistence between Wi-Fi and LTE. Load-based listen-before-Talk (LB-LBT) has been adopted as this LTE channel access mechanism by the 3rd Generation Partnership Project (3GPP). Theoretical modelling of LB-LBT schemes has focused on throughput and fair channel-Time sharing between Wi-Fi and LTE technologies. We explore a LB-LBT scheme that belongs to LBT category 4, as recommended by the 3GPP, and develop a model for the distribution of the medium access control (MAC) delays experienced by the Wi-Fi packets and LTE frames. The model, validated by simulations, reveals design insights that can be used to dynamically adjust the LB-LBT parameters not only to achieve channel-Time fairness, but also to guarantee MAC-delay bounds, with specified probability.
Li, B, Chen, S, Sutton, GJ, Shi, Y & Liu, RP 2017, 'MAC Performance Analysis for Drive-Thru Internet Networks with Rayleigh Capture', IEEE Access, vol. 5, pp. 10649-10661.View/Download from: UTS OPUS or Publisher's site
© 2013 IEEE. In practical radio transmissions, channel capture is a dominating factor that affects wireless network performance. The capture effect can occur in wireless network when packets arrive with different powers. Packets with high power can effectively swamp low power packets, such that they are received successfully, when otherwise a collision would have occurred. We present a vehicular network performance-prediction model for a Rayleigh capture channel in Drive-thru Internet scenario. The model incorporates the capture effect into a 2-D Markov chain modeling the high-node mobility and distributed coordination function broadcast scheme. The performance-prediction model unveils the impacts of mobility velocity and number of vehicles on the throughput in a Rayleigh capture channel. We use a vehicular traffic flow model to predict vehicular movement on road by aggregating all vehicles into a flow. Simulation results confirm that our performance-prediction model accurately predicts the performance of traveling vehicles with Rayleigh capture channel in the Drive-thru Internet scenario. We demonstrate that using our performance-prediction model, we can obtain optimal contention window value, by which the best system throughput can be reached without wasting contention time. This is also proved by Anastasi et al.
Li, B, Sutton, GJ, Hu, B, Liu, RP & Chen, S 2017, 'Modeling and QoS analysis of the IEEE 802.11p broadcast scheme in vehicular ad hoc networks', Journal of Communications and Networks, vol. 19, no. 2, pp. 169-179.View/Download from: UTS OPUS or Publisher's site
© 2011 KICS. Quality of service (QoS) and queue management are critical issues for the broadcast scheme of IEEE 802.11p systems in vehicular ad hoc networks (VANETs). However, existing 1-dimensional (1-D) Markov chain models of 802.11p systems are unable to capture the complete QoS performance and queuing behavior due to the lack of an adequate finite buffer model. We present a 2-dimensional (2-D) Markov chain that integrates the broadcast scheme of the 802.11p system and the queuing process into one model. The extra dimension, which models the queue length, allows us to accurately capture the important QoS measures, delay and loss, plus throughput and queue length, for realistic 802.11p systems with finite buffer under finite load. We derive a simplified method to solve the steady state probabilities of the 2-D Markov chain. Our 2-D Markov chain model is the first finite buffer model defined and solved for the broadcast scheme of 802.11p systems. The 2-D model solutions are validated by extensive simulations. Our analyses reveal that the lack of binary exponential backoff and retransmission in the 802.11p system results in poor QoS performance during heavy traffic load, particularly for large VANETs. We demonstrate that our model provides traffic control guidelines to maintain good QoS performance for VANETs.
Wu, G, Liu, RP, Ni, W, Sutton, GJ & Xu, P 2017, 'Modeling Channel Switching and Contention Control in Vehicular Networks', IEEE Access, vol. 5, pp. 23102-23110.View/Download from: UTS OPUS or Publisher's site
© 2017 IEEE. In vehicular networks, multi-channel operation standard IEEE 1609.4 is designed for vehicular communications across multiple channels. It has been revealed that such multi-channel operations may result in high contention in vehicular communications. However, existing analytical models are unable to capture the dynamic characteristic during channel switching. We develop a novel Markov model that takes into account the dynamic contention behavior during channel switching. In particular, our model reveals the high contention caused by the burst arrivals, which are the results of multi-channel operations. To combat such performance decline, we propose two solutions, a centralized equal-spaced algorithm and a distributed random-spaced algorithm. The key idea is to disperse the burst packet arrivals across the available timeframe in order to alleviate contention. Our model, validated by simulations, accurately characterizes the high contention caused by multi-channel operations. Our results demonstrate our proposed solutions can effectively mitigate packet collision, enhance reliability, and improve system throughput during the multi-channel operations.
Sutton, GJ, Liu, RP & Jay Guo, Y 2017, 'Harmonising Coexistence of Machine Type Communications with Wi-Fi Data Traffic under Frame-Based LBT', IEEE Transactions on Communications, vol. 65, no. 9, pp. 4000-4011.View/Download from: UTS OPUS or Publisher's site
© 1972-2012 IEEE. The existence of relatively long LTE data blocks within the licensed-assisted access (LAA) framework results in bursty machine-type communications (MTC) packet arrivals, which cause system performance degradation and present new challenges in Markov modeling. We develop an embedded Markov chain to characterize the dynamic behavior of the contention arising from bursty MTC and Wi-Fi data traffic in the LAA framework. Our theoretical model reveals a high-contention phenomenon caused by the bursty MTC traffic, and quantifies the resulting performance degradation for both MTC and Wi-Fi data traffic. The Markov model is further developed to evaluate three potential solutions aiming to alleviate the contention. Our analysis shows that simply expanding the contention window, although successful in reducing congestion, may cause unacceptable MTC data loss. A TDMA scheme instead achieves better MTC packet delivery and overall throughput, but requires centralized coordination. We propose a distributed scheme that randomly spreads the MTC access processes through the available time period. Our model results, validated by simulations, demonstrate that the random spreading solution achieves a near TDMA performance, while preserving the distributed nature of the Wi-Fi protocol. It alleviates the MTC traffic contention and improves the overall throughput by up to 10%.
Gray, DJ, Noble, RRP, Reid, N, Sutton, GJ & Pirlo, MC 2016, 'Regional scale hydrogeochemical mapping of the northern Yilgarn Craton, Western Australia: a new technology for exploration in arid Australia', GEOCHEMISTRY-EXPLORATION ENVIRONMENT ANALYSIS, vol. 16, no. 1, pp. 100-115.View/Download from: Publisher's site
Khoury, RR, Sutton, GJ, Ebrahimi, D & Hibbert, DB 2014, 'Formation Constants of Copper(II) Complexes with Tripeptides Containing Glu, Gly, and His: Potentiometric Measurements and Modeling by Generalized Multiplicative Analysis of Variance', INORGANIC CHEMISTRY, vol. 53, no. 3, pp. 1278-1287.View/Download from: Publisher's site
Sutton, GJ, Cawley, A, Murphy, C, Lau, ML & Hibbert, DB 2014, 'Factors influencing total carbon dioxide concentrations in plasma of thoroughbred and standardbred racehorses', DRUG TESTING AND ANALYSIS, vol. 6, no. 9, pp. 936-943.View/Download from: Publisher's site
Liu, RP, Sutton, GJ & Collings, IB 2014, 'WLAN Power Save with Offset Listen Interval for Machine-to-Machine Communications', IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, vol. 13, no. 5, pp. 2552-2562.View/Download from: UTS OPUS or Publisher's site
Khoury, RR, Sutton, GJ, Hibbert, DB & Ebrahimi, D 2013, 'Measurement and modeling of acid dissociation constants of tri-peptides containing Glu, Gly, and His using potentiometry and generalized multiplicative analysis of variance', DALTON TRANSACTIONS, vol. 42, no. 8, pp. 2940-2947.View/Download from: Publisher's site
Liu, RP, Sutton, GJ & Collings, IB 2013, 'A New Queueing Model for QoS Analysis of IEEE 802.11 DCF with Finite Buffer and Load (vol 9, pg 2664, 2010)', IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, vol. 12, no. 10, pp. 5374-5374.View/Download from: Publisher's site
Salouros, H, Sutton, GJ, Howes, JM, Hibbert, DB & Collins, M 2013, 'Measurement of stable isotope ratios in methylamphetamine: A link to its precursor source', Analytical Chemistry, vol. 85, no. 19, pp. 9400-9408.View/Download from: UTS OPUS or Publisher's site
The illicit drug methylamphetamine is often prepared from the precursor ephedrine or pseudoephedrine, which in turn are obtained by three processes: extraction from the Ephedra plant (natural), via fermentation of sugars (semi-synthetic), and by a fully synthetic route from propiophenone. We report the first method to differentiate between the three industrial routes used to produce the precursors ephedrine and pseudoephedrine by measurement of stable isotope ratios of nitrogen (ä15N), hydrogen (ä2H), and carbon (ä13C). Analysis of 782 samples of seized methylamphetamine allowed classification into three groups using k-means clustering or the expectation-maximization algorithm applied to a Gaussian mixture model. By preparation of 30 samples of ephedrine by the fully synthetic industrial process and measuring their ä15N, ä2H, and ä13C values, we observed that 15N becomes significantly depleted compared to the methylamine starting material. Conversion of ten ephedrine samples to methylamphetamine showed that this depletion is maintained in the final drug product, of which the ä15N, ä13C, and ä2H values were distinct from those of ephedrine and methylamphetamine samples of a semi-synthetic (fermentation pathway) origin. Combining modeling analysis with the new experiments and published information on the values of ä2H gave a definitive assignment of the three model groups, and equations to obtain probabilities for the precursor origin of any new sample. A simple rule of thumb is also presented. Making an assignment using delta values is particularly useful when no other chemical profiling information is available.
Liu, RP, Sutton, GJ & Collings, IB 2013, 'Erratum: Errata to the paper a new queueing model for QoS analysis of IEEE 802.11 DCF with Finite Buffer and Load (IEEETrans. Wireless Commun) (2010) 9:8 (2664-2675))', IEEE Transactions on Wireless Communications, vol. 12, no. 10, p. 5374.View/Download from: Publisher's site
Sutton, GJ, Liu, RP & Collings, IB 2013, 'Modelling IEEE 802.11 DCF Heterogeneous Networks with Rayleigh Fading and Capture', IEEE TRANSACTIONS ON COMMUNICATIONS, vol. 61, no. 8, pp. 3336-3348.View/Download from: UTS OPUS or Publisher's site
Sparks, RS, Sutton, G, Toscas, P & Ormerod, JT 2011, 'Modelling inverse Gaussian data with censored response values: EM versus MCMC', Advances in Decision Sciences, vol. 2011.View/Download from: Publisher's site
Low detection limits are common in measure environmental variables. Building models using data containing low or high detection limits without adjusting for the censoring produces biased models. This paper offers approaches to estimate an inverse Gaussian distribution when some of the data used are censored because of low or high detection limits. Adjustments for the censoring can be made if there is between 2% and 20% censoring using either the EM algorithm or MCMC. This paper compares these approaches. Copyright © 2011 R. S. Sparks et al.
Liu, RP, Sutton, GJ & Collings, IB 2010, 'A New Queueing Model for QoS Analysis of IEEE 802.11 DCF with Finite Buffer and Load', IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, vol. 9, no. 8, pp. 2664-2675.View/Download from: Publisher's site
Sutton, GJ & Bitmead, RR 1998, 'Experiences with Model Predictive Control applied to a nonlinear constrained submarine', PROCEEDINGS OF THE 37TH IEEE CONFERENCE ON DECISION AND CONTROL, VOLS 1-4, pp. 1370-1375.
Sparks, R, Sutton, GJ, Toscas, P & Innes, RM 2013, 'Estimates of likelihood and risk associated with Sydney drinking water supply from reservoirs, local dams and feed rivers' in Water Quality: Indicators, Human Impact and Environmental Health, pp. 31-61.
As the main supplier of potable water to Sydney, Sydney Catchment Authority aims to supply water that is in accordance with Australian drinking water quality standards. The standard specifies thresholds for acceptable ranges for over 80 analytes and attributes. When the quality thresholds are exceeded, the water needs to be treated, so that the treated water complies with the standard. The total expected cost of treating the water is the risk to be calculated. The main objective of the chapter is to describe a water quality risk assessment process that is free from sampling biases. The risk assessment process involves modelling each analyte based on historical data so that estimates can be obtained for the likelihood of exceeding the water quality thresholds. These likelihood estimates are then combined with associated estimated water treatment costs to produce an annual water treatment cost estimate, or equivalently, an annual risk estimate. Costs of exceeding thresholds are in their early stages of development and, at best, costs have been developed for each threshold exceedance separately. Some analytes have one-sided minor and major thresholds, while others have two-sided thresholds. Risk values (equal to the integral of exceedance likelihood estimates times the treatment cost associated with such exceedances) have been explicitly calculated for 2006. A future aim is to repeat this process for all past years as a way of monitoring the year-to-year variation in risk for each analyte. Since the process is to be to be repeated for each year - the chapter works towards providing a robust approach to estimating the risk, which is:1. Flexible in distributional assumptions; and 2. Capable of selecting variables that are important for estimating risk. The application datasets come from the Sydney Catchment Authority and comprise approximately 80 analytes sampled at 35 different locations. Given the number of analyte-site combinations and the aim of yearly model fitti...
Sutton, GJ & Bitmead, RR 2000, 'Performance and computational implementation of nonlinear model predictive control on a submarine' in Nonlinear Model Predictive Control, Birkhäuser Basel, pp. 461-472.
Liu, RP, Sutton, GJ & Collings, IB 2013, 'Power save with Offset Listen Interval for IEEE 802.11ah Smart Grid communications', IEEE International Conference on Communications, IEEE International Conference on Communications, IEEE, Budapest, Hungary, pp. 4488-4492.View/Download from: UTS OPUS or Publisher's site
Communication is an enabling technology for the efficient control and management of next-generation Smart Grids. Energy conservation of the communication devices is essential for future large scale deployment of Smart Grid communication networks. However, existing power save protocols experience high contention in Smart Grid communication networks that have a large number of nodes and periodic traffic. We design a new energy conservation protocol, Power Save with Offset Listen Interval (PS-OLi), to address such contention problems. PS-OLi avoids message collisions by controlling the station wake up time with a calculated offset. A new analytical model is developed to characterize the power save performance of networks with periodic traffic. Simulation results show that our analytical model accurately predicts the collision probability and packet delay. We use our model to evaluate the energy efficiency of PS-OLi and standard power save protocols. Our results show that PS-OLi extends the lifetime of a Smart Grid communication network by more than 10%. © 2013 IEEE.
Liu, RP, Sutton, GJ, Yang, X & Collings, IB 2011, 'Modelling QoS performance of IEEE 802.11 DCF under practical channel fading conditions', IEEE International Conference on Communications, IEEE International Conference on Communications, IEEE, Kyoto, Japan.View/Download from: UTS OPUS or Publisher's site
We consider the impacts of channel fading on the quality of service (QoS) performance of the IEEE 802.11 system. Traditional 2-D Markov chain models, while suitable for throughput analysis, are unable to capture the QoS performance due to the lack of a proper queueing model. We present a 3-D Markov chain queueing model that incorporates channel fading effects and solve the Markov chain efficiently with our Collapsed Transition onto Basis approach. In doing so, we are able to investigate important QoS measures, packet delay and loss, as well as throughput, for a 802.11 system under practical channel fading conditions. The analytical results are validated by extensive simulations. Our 3-D model offers new insights in channel fading effects on system capacity and QoS performance. We demonstrate that our 3-D model can also be used to provides guidelines for traffic control. © 2011 IEEE.
Sutton, GJ, Liu, RP, Yang, X & Collings, IB 2010, 'Modelling Capture Effect for 802.11 DCF under Rayleigh Fading', 2010 IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS, 2010 IEEE International Conference on Communications, IEEE, Cape Town, SOUTH AFRICA.
Liu, RP, Sutton, G & Collings, IB 2009, 'A 3-D Markov Chain Queueing Model of IEEE 802.11 DCF with Finite Buffer and Load', 2009 IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS, VOLS 1-8, IEEE International Conference on Communications (ICC 2009), IEEE, Dresden, GERMANY, pp. 5320-+.
Sutton, GJ & Bitmead, RR 1997, 'Robust stability theorems for nonlinear predictive control', PROCEEDINGS OF THE 36TH IEEE CONFERENCE ON DECISION AND CONTROL, VOLS 1-5, 36th IEEE Conference on Decision and Control, IEEE, SAN DIEGO, CA, pp. 4886-4891.