Imperial College London


Faculty of EngineeringDepartment of Computing

Vice-Dean (Research) for the Faculty of Engineering



+44 (0)20 7594 8375j.mccann Website




Miss Teresa Ng +44 (0)20 7594 8300




260ACE ExtensionSouth Kensington Campus





Publication Type

245 results found

Bouazizi Y, Benkhelifa F, ElSawy H, McCann JAet al., 2023, SF adaptation in duty-cycled LoRa networks: a spatiotemporal study, IEEE Wireless Communications and Networking Conference, WCNC, Publisher: IEEE, ISSN: 1525-3511

An analytical model joining stochastic geometry andqueuing theory is devised to study the performance of adaptive LoRa networks with dynamic Spreading Factor (SF) allocation. LoRa devices are perceived as interacting two-dimensional Discrete Time Markov Chains (DTMC)s. Each chain jointly tracks the number of packets in the buffer and the node’s protocol state while accounting for Duty Cycle (DC) restrictions and quantifying the imperfect orthogonality of SFs. The network performance is characterised in terms of coverage, delay and Pareto frontiers under different orthogonality assumptions and for various adaptation settings highlighting insights useful for the design of application-aware decentralised or semi-decentralisedSF adaptation schemes.

Conference paper

Breza M, Junejo A, McCann J, 2023, Threat modeling for communication security of IoT-enabled digital logistics, Sensors, Vol: 23, ISSN: 1424-8220

The modernization of logistics through the use of Wireless Sensor Network (WSN) Internet of Things (IoT) devices promises great efficiencies. Sensor devices can provide real-time or near real-time condition monitoring and location tracking of assets during the shipping process, helping to detect delays, prevent loss, and stop fraud. However, the integration of low-cost WSN/IoT systems into a pre-existing industry should first consider security within the context of the application environment. In the case of logistics, the sensors are mobile, unreachable during the deployment, and accessible in potentially uncontrolled environments. The risks to the sensors include physical damage, either malicious/intentional or unintentional due to accident or the environment, or physical attack on a sensor, or remote communication attack. The easiest attack against any sensor is against its communication. The use of IoT sensors for logistics involves the deployment conditions of mobility, inaccesibility, and uncontrolled environments. Any threat analysis needs to take these factors into consideration. This paper presents a threat model focused on an IoT-enabled asset tracking/monitoring system for smart logistics. A review of the current literature shows that no current IoT threat model highlights logistics-specific IoT security threats for the shipping of critical assets. A general tracking/monitoring system architecture is presented that describes the roles of the components. A logistics-specific threat model that considers the operational challenges of sensors used in logistics, both malicious and non-malicious threats, is then given. The threat model categorizes each threat and suggests a potential countermeasure.

Journal article

Yu H, Ren X, Zhao C, Yang S, McCann Jet al., 2023, Quantum-aided secure deep neural network inference on real quantum computers, Scientific Reports, Vol: 13, ISSN: 2045-2322

Deep neural networks (DNNs) are phenomenally successful machine learning methods broadly applied to many different disciplines. However, as complex two-party computations, DNN inference using classical cryptographic methods cannot achieve unconditional security, raising concern on security risks of DNNs' application to sensitive data in many domains. We overcome such a weakness by introducing a quantum-aided security approach. We build a quantum scheme for unconditionally secure DNN inference based on quantum oblivious transfer with an untrusted third party. Leveraging DNN's noise tolerance, our approach enables complex DNN inference on comparatively low-fidelity quantum systems with limited quantum capacity. We validated our method using various applications with a five-bit real quantum computer and a quantum simulator. Both theoretical analyses and experimental results demonstrate that our approach manages to operate on existing quantum computers and achieve unconditional security with a negligible accuracy loss. This may open up new possibilities of quantum security methods for deep learning.

Journal article

Surougi H, McCann J, 2023, Real-time optimisation-based path planning for visually impaired people in dynamic environments, 2023 International Conference on Computer Vision, Publisher: Computer Vision Foundation, Pages: 1839-1848

Most existing outdoor assistive mobility solutions notify Visually Impaired People (VIP) about potential collisions but fail to provide Optimal Local Collision-Free Path Planning (OLCFPP) to enable the VIP to get out of the way effectively. In this paper, we propose MinD, the first VIP OLCFPP scheme that notifies the VIP of the shortest path required to avoid Critical Moving Objects (CMOs), like cars, motorcycles, etc. This simultaneously accounts for the VIP's mobility constraints, the different CMO types and movement patterns, and predicted collision times, conducting a safety prediction trajectory analysis of the optimal path for the VIP to move in. We implement a real-world prototype to conduct extensive outdoor experiments that record the aforementioned parameters, and this populates our simulations for evaluation against the state-of-the-art. Experimental results demonstrate that MinD outperforms the Artificial Potential Field (APF) approach in effectively planning a short collision-free route, requiring only 1.69m of movement on average, shorter than APF by 90.23%, with a 0% collision rate; adapting to the VIP's mobility limitations and provides a high safe time separation (>5.35s on average compared to APF). MinD also shows near real-time performance, with decisions taking only 0.04s processing time on a standard off-the-shelf laptop.

Conference paper

Awad H, Heggo M, Pang O, Kovac M, McCann Jet al., 2023, Multirotor motion enhancement using propeller speed measurements, The 2023 International Conference on Unmanned Aircraft Systems, Publisher: IEEE, Pages: 401-406

Multirotor autopilots often depend on open-loop control without the feedback of propeller speeds, although they are a critical factor in determining motion characteristics. This paper proposes a system that leverages actual propeller speeds as direct feedback to the autopilot to improve the state estimation and dynamics of the multirotor. Software-in-the-Loop (SITL) and Hardware-in-the-Loop (HITL) simulations with real data, in different scenarios, are conducted to demonstrate the impact of combining propeller speeds with typical drone sensors. The results show that the drone becomes more stable with lower trajectory errors. Further, a noticeable reduction in the vehicle position median error while following a trajectory is shown, and a considerable increase in the flying duration time before crashing in case of a motor fault. These results highlight the potential of adding propeller speed feedback to increase the autopilot’s controllability which enhances drone performance in sensitive applications.

Conference paper

Huang S, Chen P, McCann J, 2023, DiffAR: Adaptive conditional diffusion model for temporal-augmented human activity recognition, The 32nd International Joint Conference on Artificial Intelligence

Human activity recognition (HAR) is a fundamental sensing and analysis technique that supports diverse applications, such as smart homes and health-care. In device-free and non-intrusive HAR, WiFi channel state information (CSI) captures wireless signal variations caused by human interference without the need for video cameras or on-bodysensors. However, current CSI-based HAR performance is hampered by incomplete CSI recordings due to fixed window sizes in CSI collection and human/machine errors that incur missing values in CSI. To address these issues, we propose DiffAR, a temporal-augmented HAR approach thatimproves HAR performance by augmenting CSI. DiffAR devises a novel Adaptive Conditional Diffusion Model (ACDM) to synthesize augmented CSI, which tackles the issue of fixed windows by forecasting and handles missing values with imputation. Compared to existing diffusion models, ACDM improves the synthesis quality by guidingprogressive synthesis with step-specific conditions.DiffAR further exploits an ensemble classifier for activity recognition using both raw and augmented CSI. Extensive experiments on four public datasets show that DiffAR achieves the best synthesis quality of augmented CSI and outperforms state-of-the-art CSI-based HAR methods in terms of recognition performance. The source code of DiffAR is available at

Conference paper

Surougi HR, McCann JA, 2023, Real-Time Optimisation-Based Path Planning for Visually Impaired People in Dynamic Environments, Pages: 1831-1840

Most existing outdoor assistive mobility solutions notify Visually Impaired People (VIP) about potential collisions but fail to provide Optimal Local Collision-Free Path Planning (OLCFPP) to enable the VIP to get out of the way effectively. In this paper, we propose MinD, the first VIP OLCFPP scheme that notifies the VIP of the shortest path required to avoid Critical Moving Objects (CMOs), like cars, motorcycles, etc. This simultaneously accounts for the VIP's mobility constraints, the different CMO types and movement patterns, and predicted collision times, conducting a safety prediction trajectory analysis of the optimal path for the VIP to move in. We implement a real-world prototype to conduct extensive outdoor experiments that record the aforementioned parameters, and this populates our simulations for evaluation against the state-of-the-art. Experimental results demonstrate that MinD outperforms the Artificial Potential Field (APF) approach in effectively planning a short collision-free route, requiring only 1.69m of movement on average, shorter than APF by 90.23%, with a 0% collision rate; adapting to the VIP's mobility limitations and provides a high safe time separation (> 5.35s on average compared to APF). MinD also shows near real-time performance, with decisions taking only 0.04s processing time on a standard off-the-shelf laptop.

Conference paper

Wang L, Zhao C, Yang S, Yang X, Mccann Jet al., 2023, ACE: toward application-centric, edge-cloud, collaborative intelligence, Communications of the ACM, Vol: 66, Pages: 62-73, ISSN: 0001-0782

Constructing a unified platform for the scalable, reliable, robust, and efficient development and deployment of ECCI applications.

Journal article

Chen K, Benkhelifa F, Gao H, McCann J, Li Jet al., 2022, Minimizing age of information in multi-hop energy-harvesting wireless sensor network, IEEE Internet of Things Journal, Vol: 9, Pages: 25736-25751, ISSN: 2327-4662

Age of information (AoI), a metric measuring the information freshness, has drawn increased attention due to its importance in monitoring applications in which nodes send time-stamped status updates to interested recipients, and timely updates about phenomena are important. In this work, we consider the AoI minimization scheduling problem in multi-hop energy harvesting(EH) wireless sensor networks (WSNs). We design the generation time of updates for nodes and develop transmission schedules under both protocol and physical interference models, aiming at achieving minimum peak AoI and average AoI among all nodes for a given time duration. We prove that it is an NP-Hard problem and propose an energy-adaptive, distributed algorithm called MAoIG. We derive its theoretical upper bounds for the peak and average AoI and a lower bound for peak AoI. The numerical results validate that MAoIG outperforms all of the baseline schemes in all scenarios and that the experimental results tightly track the theoretical upper bound optimal solutions while the lower bound tightness decreases with the number of nodes.

Journal article

Bouazizi Y, Benkhelifa F, ElSawy H, McCann JAet al., 2022, On the scalability of duty-cycled LoRa networks with imperfect SF orthogonality, IEEE Wireless Communications Letters, Vol: 11, Pages: 2310-2314, ISSN: 2162-2337

This papers uses stochastic geometry and queuingtheory to study he scalability of long-range (LoRa) networks,accounting for duty cycling restrictions and imperfect spreadingfactor (SFs) orthogonality. The scalability is characterised by thejoint boundaries of device density and traffic intensity per device.Novel cross-correlation factors are used to quantify imperfect SForthogonality. Our results show that a proper characterisationof LoRa orthogonality extends the scalability of the network.They also highlight that for low/medium densities decreasingthe SF extends the spanned spectrum of sensing applicationscharacterised by their traffic requirements (i.e. sensing rate).However, for high density (> 104 nodes/Km2), the Pareto frontiersconverge to a stability limit governed by the SF allocation schemeand the predefined capture thresholds. The results further evincethe importance of capturing threshold distribution among the SFsto mitigate the unfair latency.

Journal article

Benkhelifa F, Bouazizi Y, McCann J, 2022, How orthogonal is LoRa modulation?, IEEE Internet of Things Journal, Vol: 9, Pages: 19928-19944, ISSN: 2327-4662

In this paper, we provide, for the first time, acomprehensive understanding of LoRa waveform theory in order to quantify its orthogonality. We present LoRa waveformexpressions in continuous and discrete time domains, and analyzemeasures of orthogonality between different LoRa spreadingfactors (SFs) through cross-correlation functions. The crosscorrelation functions are analytically expressed in a generalform and they account for diverse configuration parameters(bandwidth, SF, etc.) and different cases of signal displacements(time delay shift, frequency shift, etc.). We quantify their meanand maximum in all time domains. We highlight the impact ofthe temporal displacement and different bandwidths. The generalresult is that LoRa modulation is non-orthogonal. Firstly, weobserve that for same bandwidths the largest maximum crosscorrelation happens for same SF and is equal to 100% due to samesymbols; whereas for different bandwidths, the largest maximumcross-correlation is no longer observed at the same SF. Secondly,the maximum cross-correlation is less than 26% between differentSFs, is higher for closer SFs and decreases as the differencebetween SFs increases. After downchirping, the maximum crosscorrelation increases and the mean decreases compared to thosebefore downchirping. Moreover, the maximum cross-correlationis insignificantly impacted by the temporal delay which makes itvalid to adopt for the performance analysis of both synchronousand asynchronous systems. Finally, we analyze by simulation thebit error probability statistics for different bandwidth ratios andhighlight their correlated behaviour with the insights obtainedfrom the maximum cross-correlation expressions.

Journal article

Yu W, McCann J, Zhang C, Ferhatosmanoglu Het al., 2022, Scaling high-quality pairwise link-based similarity retrieval on billion-edge graphs, ACM Transactions on Information Systems, Vol: 40, Pages: 1-45, ISSN: 1046-8188

SimRank is an attractive link-based similarity measure used in fertile fields of Web search and sociometry. However, the existing deterministic method by Kusumoto et al. [24] for retrieving SimRank does not always produce high-quality similarity results, as it fails to accurately obtain diagonal correction matrix D. Moreover, SimRank has a “connectivity trait” problem: increasing the number of paths between a pair of nodes would decrease its similarity score. The best-known remedy, SimRank++ [1], cannot completely fix this problem, since its score would still be zero if there are no common in-neighbors between two nodes.In this article, we study fast high-quality link-based similarity search on billion-scale graphs. (1) We first devise a “varied-D” method to accurately compute SimRank in linear memory. We also aggregate duplicate computations, which reduces the time of [24] from quadratic to linear in the number of iterations. (2) We propose a novel “cosine-based” SimRank model to circumvent the “connectivity trait” problem. (3) To substantially speed up the partial-pairs “cosine-based” SimRank search on large graphs, we devise an efficient dimensionality reduction algorithm, PSR#, with guaranteed accuracy. (4) We give mathematical insights to the semantic difference between SimRank and its variant, and correct an argument in [24] that “if D is replaced by a scaled identity matrix (1-Ɣ)I, their top-K rankings will not be affected much”. (5) We propose a novel method that can accurately convert from Li et al. SimRank ~{S} to Jeh and Widom’s SimRank S. (6) We propose GSR#, a generalisation of our “cosine-based” SimRank model, to quantify pairwise similarities across two distinct graphs, unlike SimRank that would assess nodes across two graphs as completely dissimilar. Extensive experiments on various datasets demonstrate the superiority of our proposed approaches in terms of hig

Journal article

Heggo M, Bhatia L, McCann J, 2022, RFTacho: Non-intrusive RF monitoring of rotating machines, The International Conference on Information Processing in Sensor Networks (IPSN) 2022, Publisher: IEEE, Pages: 403-414

Measuring rotation speed is essential to many engineering applications; it elicits faults undetectable by vibration monitoring alone and enhances the vibration signal analysis of rotating machines. Optical, magnetic or mechanical Tachometers are currently state-of-art. Their limitations are they require line-of-sight, direct access to the rotating object. This paper proposes RFTacho, a rotation speed measurement system that leverages novel hardware and signal processing algorithms to produce highly accurate readings conveniently. RFTacho uses RF Orbital Angular Momentum (OAM) waves to measure rotation speed of multiple machines simultaneously with no requirements from the machine’s properties. OAM antennas allow it to operate in high-scattering environments, commonly found in industries, as they are resilient to de-polarization compared to linearly polarized antennas. RFTacho achieves this by using two novel signal processing algorithms to extract therotation speed of several rotating objects simultaneously amidst noise arising from high-scattering environments, non-line-of-sight scenarios and dynamic environmental conditions with a resolution of 1𝑟𝑝𝑚. We test RFTacho on several real-world machines like fans, motors, air conditioners. Results show that RFTacho has avg. error of < 0.5% compared to ground truth. We demonstrate RFTacho’s simultaneous multiple-object measurement capability that other tachometers do not have. Initial experiments show that RFTacho can measure speeds as high as 7000 rpm (theoretically 60000 rpm) with high resiliency at different coverage distances and orientation angles, requiring only 150 mW transmit power while operating in the 5 GHz license-exempt band. RFTacho is the first RF-based sensing system that combines OAM waves and novel processing approaches to measure the rotation speed of multiple machines simultaneously in a non-intrusive way.

Conference paper

Junejo AK, Benkhelifa F, Wong B, McCann JAet al., 2022, LoRa-LiSK: a lightweight shared secret key generation scheme for LoRa Networks, IEEE Internet of Things Journal, Vol: 9, Pages: 4110-4124, ISSN: 2327-4662

Physical layer security (PLS) schemes use the randomness of the channel parameters, namely, channel state information (CSI) and received signal strength indicator (RSSI), to generate the secret keys. There has been limited work in PLS schemes in long-range (LoRa) wide area networks (Lo- RaWANs), hindering their widespread application. Limitations observed in existing studies include the requirement of having a high correlation between channel parameter measurements and the evaluation in either fully indoor or outdoor environments. The real-world wireless sensor networks (WSNs) and LoRa use cases might not meet both requirements, thus making the current PLS schemes inappropriate for these systems. This paper proposes LoRA-LiSK, a practical and efficient shared secret key generation scheme for LoRa networks to address the limitations of existing PLS schemes. Our proposed LoRa-LiSK scheme consists of several preprocessing techniques (timestamp matching, two sample Kolmogorov Smirnov tests, and a Savitzky- Golay filter), multi-level quantization, information reconciliation using Bose-Chaudhuri-Hocquenghem (BCH) codes, and finally, privacy amplification using secure hash algorithm SHA-2. The LoRa-LiSK scheme is extensively evaluated on real WSN/IoT devices in practical application scenarios: 1) indoor to outdoor and 2) long range static and mobile outdoor links. It outperforms existing schemes by generating keys with channel parameter measurements of low correlation values (0:2 to 0:6) while still achieving high key generation rates and low key disagreement rates (10%–20%). The scheme updates a key in one hour approximately using an application profile with a high transmission rate compared to three hours reported by existing works while still respecting the duty cycle regulation. It also incurs less communication overhead compared to the existing works.

Journal article

Bezerra P, Chen P-Y, McCann JA, Yu Wet al., 2022, Adaptive monitor placement for near real-time node failure localisation in wireless sensor networks, ACM Transactions on Sensor Networks, Vol: 18, ISSN: 1550-4859

As sensor-based networks become more prevalent, scaling to unmanageable numbers or deployed in difficult to reach areas, real-time failure localisation is becoming essential for continued operation. Network tomography, a system and application-independent approach, has been successful in localising complex failures (i.e., observable by end-to-end global analysis) in traditional networks.Applying network tomography to wireless sensor networks (WSNs), however, is challenging. First, WSN topology changes due to environmental interactions (e.g., interference). Additionally, the selection of devices for running network monitoring processes (monitors) is an NP-hard problem. Monitors observe end-to-end in-network properties to identify failures, with their placement impacting the number of identifiable failures. Since monitoring consumes more in-node resources, it is essential to minimise their number while maintaining network tomography’s effectiveness. Unfortunately, state-of-the-art solutions solve this optimisation problem using time-consuming greedy heuristics.In this article, we propose two solutions for efficiently applying Network Tomography in WSNs: a graph compression scheme, enabling faster monitor placement by reducing the number of edges in the network, and an adaptive monitor placement algorithm for recovering the monitor placement given topology changes. The experiments show that our solution is at least 1,000× faster than the state-of-the-art approaches and efficiently copes with topology variations in large-scale WSNs.

Journal article

Yang S, Wang X, Adeel U, Zhao C, Hu J, Yang X, McCann Jet al., 2021, The design of user-centric mobile crowdsensing with cooperative D2D communications, IEEE Wireless Communications, Vol: 29, Pages: 1-9, ISSN: 1536-1284

Ubiquitous sensor-rich smartphones have promoted MCS, an emerging people-centric sensing paradigm for urban IoT. However, using cellular networks to transmit the big data collected by MCS would incur expensive financial costs for both participating phone users and the MCS organizer. A promising solution to this is to integrate the low-cost D2D communication into the MCS design. However, using D2D in MCS will require cooperative interactions among self-interested and strategic participating phone users, which significantly complicates the "human-in-the-loop" MCS design in both the digital and human dimensions, and brings a branch of new challenges: data communication and networking become more complex; D2D connections among opportunistically encountered phone users must be secure, fast, and user-transparent; and participating phone users need to be properly incentivized. In dealing with these challenges, this article covers recent developments of D2D-enabled MCS from both the theoretical and practical perspectives. Future research questions in this rapidly growing field are also discussed.

Journal article

Wang H, Zhou G, Xu J, Liu Z, Yan X, Mccann Jet al., 2021, A simplified historical-infomation-based SOC prediction method for supercapacitors, IEEE Transactions on Industrial Electronics, Vol: 69, Pages: 13090-13098, ISSN: 0278-0046

Range anxiety has become an important issue for the application of electric vehicles (EVs). Drivers need information on whether they can reach their destinations and what the remaining capacity would be before starting a trip. In order to satisfy the needs and save computing resources for computing-intense applications in vehicles, we propose a simplified historical-information-based State of Charge (SOC) prediction (SHSP) algorithm. First, definitions of SOC, historical average power, and equivalent current are given. Based on these definitions, Rint-based models of supercapacitors, under constant power and constant current loading, are established respectively. Then, a relationship between the historical average power and the predicted SOC is derived with the help of the equivalent current as a bridge. The experimental results demonstrate that the 35-step-forward SOC prediction error of the driving-behavior-based SOC prediction (SHSP) is close to the driving-behavior-based SOC prediction method (DBSP) and lower than Long-Short-Term-Memory-based SOC prediction method (LSTM). Importantly, the time of running SHSP code is less than that of running DBSP code, and much less than that of running LSTM code.

Journal article

Heggo M, Bhatia L, McCann J, 2021, Cognisense: A contactless rotation speed measurement system, The 19th ACM Conference on Embedded Networked Sensor Systems, Pages: 341-348

Several engineering applications require reliable rotation speed measurement for their correct functioning. The rotation speed measurements can be used to enhance the machines’ vibration signalanalysis and can also elicit faults undetectable by vibration monitoringalone. The current state of the art sensors for rotation speed measurement are optical, magnetic and mechanical tachometers. Thesesensors require line-of-sight and direct access to the machine which limits their use-cases. In this demo, we showcase Cognisense, an RF-based hardware-software sensing system that uses Orbital AngularMomentum (OAM) waves to accurately measure a machine’s rotation speed. Cognisense uses a novel compact patch antenna in a monostatic radar configuration capable of transmitting and receiving OAM waves in the 5GHz license-exempt band. The demo will show Cognisense working on machines with varied numbers of blades, sizes and materials. We will also present how Cognisense operates reliably in non-line-of-sight scenarios where traditional tachometers fail. We demonstrate how Cognisense works well in high-scattering scenarios and is not impacted by the material of rotor blades. Unlike optical tachometers that require one to face the machine head-on, Our demo will also show Cognisense performing reliably in the presence of a tilt angle between the system and the machine which is not possible with optical tachometers.

Conference paper

Zhao C, Sun X, Yang S, Ren X, Zhao P, McCann Jet al., 2021, Exploration across small silos: federated few-shot learning on network edge, IEEE Network: the magazine of global information exchange, Vol: 36, ISSN: 0890-8044

Federated Learning (FL) has been drawing significant attention from both academia and industry working on distributed machine learning. In practice, learning over mutually isolated datasets residing at the network edge, also known as silos, FL clients can suffer from a lack of samples, due to many reasons (e.g., expensive annotation), and this has potentially significant negative impact on FL performance. Few-Shot Learning (FSL) has been considered as a promising solution, but unfortunately cannot be directly applied to practical Cross-Silo Federated Learning (CSFL) systems. In this article, as far as we know, we conduct the first systematic discussion of the specific challenges of FSL in CSFL systems. We extract essential design issues found in Federated Few-Shot Learning (FFSL), and develop a new FFSL method based on Model-Agnostic Meta Learning (MAML). Through experiments using real-world federated datasets, we comprehensively demonstrate our method's advantages over existing FL and FSL methods in different practical CSFL scenarios where hitherto FL and FSL methods failed. We also highlight some promising future research directions.

Journal article

Bhatia L, Chen P-Y, Breza M, Zhao C, McCann Jet al., 2021, IRONWAN: increasing reliability of overlapping networks in LoRaWAN, IEEE Internet of Things Journal, Vol: 9, Pages: 10763-10776, ISSN: 2327-4662

LoRaWAN deployments follow an ad-hoc deployment model that has organically led to overlapping communication networks, sharing the wireless spectrum, and completely unaware of each other. LoRaWAN uses ALOHA-style communication where it is almost impossible to schedule transmission between networks belonging to different owners properly. The inability to schedule overlapping networks will cause inter-network interference, which will increase node-to-gateway message losses and gateway-to-node acknowledgement failures. This problem is likely to get worse as the number of LoRaWAN networks increase. In response to this problem, we propose IRONWAN, a wireless overlay network that shares communication resources without modifications to underlying protocols. It utilises the broadcast nature of radio communication and enables gateway-to-gateway communication to facilitate the search for failed messages and transmit failed acknowledgements already received and cached in overlapping network’s gateways. IRONWAN uses two novel algorithms, a Real-time Message Inter-arrival Predictor, to highlight when a server has not received an expected uplink message. The Interference Predictor ensures that extra gateway-to-gateway communication does not negatively impact communication bandwidth. We evaluate IRONWAN on a 1000-node simulator with up to ten gateways and a 10-node testbed with 2-gateways. Results show that IRONWAN can achieve up to 12% higher packet delivery ratio (PDR) and total messages received per node while increasing the minimum PDR by up to 28%. These improvements save up to 50% node’s energy. Finally, we demonstrate that IRONWAN has comparable performance to an optimal solution (wired, centralised) but with 2-32 times lower communication costs. IRONWAN also has up to 14% better PDR when compared to FLIP, a wired-distributed gateway-to-gateway protocol in certain scenarios.

Journal article

Wong B, McCann J, 2021, Failure detection methods for pipeline networks: from acoustic sensing to cyber-physical systems, Sensors, Vol: 21, ISSN: 1424-8220

Pipeline networks have been widely utilised in the transportation of water, natural gases, oil and waste materials efficiently and safely over varying distances with minimal human intervention. In order to optimise the spatial use of the pipeline infrastructure, pipelines are either buried underground, or located in submarine environments. Due to the continuous expansion of pipeline networks in locations that are inaccessible to maintenance personnel, research efforts have been ongoing to introduce and develop reliable detection methods for pipeline failures, such as blockages, leakages, cracks, corrosion and weld defects. In this paper, a taxonomy of existing pipeline failure detection techniques and technologies was created to comparatively analyse their respective advantages, drawbacks and limitations. This effort has effectively illuminated various unaddressed research challenges that are still present among a wide array of the state-of-the-art detection methods that have been employed in various pipeline domains. These challenges include the extension of the lifetime of a pipeline network for the reduction of maintenance costs, and the prevention of disruptive pipeline failures for the minimisation of downtime. Our taxonomy of various pipeline failure detection methods is also presented in the form of a look-up table to illustrate the suitability, key aspects and data or signal processing techniques of each individual method. We have also quantitatively evaluated the industrial relevance and practicality of each of the methods in the taxonomy in terms of their respective deployability, generality and computational cost. The outcome of the evaluation made in the taxonomy will contribute to our future works involving the utilisation of sensor fusion and data-centric frameworks to develop efficient, accurate and reliable failure detection solutions.

Journal article

Fu A, McCann JA, 2021, Dynamic Decentralized Periodic Event-Triggered Control for Wireless Cyber-Physical Systems, IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY, Vol: 29, Pages: 1783-1790, ISSN: 1063-6536

This brief adds dynamic variables to decentralized periodic event-triggered control (DPETC), denoting dynamic DPETC (DDPETC) for wireless cyber-physical systems whose sensors are distributed and bandwidths are limited, and studies the stability and L 2 -gain. In DDPETC, dynamic variables that require only local information are designed and added to the event-triggered mechanisms (ETMs). This dynamic variable can guarantee that the ETMs still be localized which does not require information from other nodes. For implementations with DDPETC, at each periodic sampling time the ETMs decide events based on new measurements from local sensors. Those measurements that trigger the events are transmitted to controllers, and after network transmission delays, they are received and the controller computes the control actions. The transmissions of these control actions from controllers to actuators are also decided by ETMs. The feasibility of the approach is shown via two examples. Numerical results show that, with the same stability and performance, DDPETC can save more than 40% events, making it more suitable for wireless cyber-physical systems.

Journal article

Junejo AK, Komninos N, McCann JA, 2021, A secure integrated framework for Fog-assisted Internet-of-Things systems, IEEE Internet of Things Journal, Vol: 8, Pages: 6840-6852, ISSN: 2327-4662

Fog-Assisted Internet of Things (Fog-IoT) systems are deployed in remote and unprotected environments, making them vulnerable to security, privacy, and trust challenges. Existing studies propose security schemes and trust models for these systems. However, mitigation of insider attacks, namely blackhole, sinkhole, sybil, collusion, self-promotion, and privilege escalation, has always been a challenge and mostly carried out by the legitimate nodes. Compared to other studies, this paper proposes a framework featuring attribute-based access control and trust-based behavioural monitoring to address the challenges mentioned above. The proposed framework consists of two components, the security component (SC) and the trust management component (TMC). SC ensures data confidentiality, integrity, authentication, and authorization. TMC evaluates Fog-IoT entities’ performance using a trust model based on a set of QoS and network communication features. Subsequently, trust is embedded as an attribute within SC’s access control policies, ensuring that only trusted entities are granted access to fog resources. Several attacking scenarios, namely DoS, DDoS, probing, and data theft are designed to elaborate on how the change in trust triggers the change in access rights and, therefore, validates the proposed integrated framework’s design principles. The framework is evaluated on a Raspberry Pi 3 Model B to benchmark its performance in terms of time and memory complexity. Our results show that both SC and TMC are lightweight and suitable for resource-constrained devices.

Journal article

Benkhelifa F, Qin Z, McCann JA, 2021, User fairness in energy harvesting-based LoRa networks with imperfect SF orthogonality, IEEE Transactions on Communications, Vol: 69, Pages: 4319-4334, ISSN: 0090-6778

Long range (LoRa) demonstrates high potential in supporting massive Internet-of-Things (IoT) applications. In this paper, we study the resource allocation in energy harvesting (EH)-enabled LoRa networks with imperfect spreading factor (SF) orthogonality. We maximize the user fairness in terms of the minimum time-averaged throughput while jointly optimizing the SF assignment, the EH time duration, and the transmit power of all LoRa users. First, we provide a general expression of the packet collision time between LoRa users which depends on the SFs and EH duration requirements of each user. Then, we develop two SF allocation schemes that either assure fairness or not for the LoRa users. Within this, we optimize the EH time and the power allocation for single and multiple uplink transmission attempts. For the single uplink transmission attempt, the optimal power allocation is obtained using bisection method. For the multiple uplink transmission attempts, the suboptimal power allocation is derived using concave-convex procedure (CCCP). Our results unearth new findings. Firstly, we demonstrate that the unfair SF allocation algorithm outperforms the others in terms of the minimum data rate. Additionally, we observe that co-SF interference is the main limitation in the throughput performance, and not really energy scarcity.

Journal article

Mijic A, Whyte J, Fisk D, Angeloudis P, Ochieng W, Cardin M-A, Mosca L, Simpson C, McCann J, Stoianov I, Myers R, Stettler Met al., 2021, The Centre for Systems Engineering and Innovation – 2030 vision and 10-year celebration

The 2030 vision of the Centre is to bring Systems Engineering and Innovation to Civil Infrastructure by changing how cross-sector infrastructure challenges are addressedin an integrated way using principles of systems engineering to maximise resilience, safety and sustainability in an increasingly complex world.We want to better understand the environmental and societal impacts of infrastructure interventions under uncertainty. This requires a change in current approaches to infrastructure systems engineering: starting from the natural environmentand its resources, encompassing societaluse of infrastructure and the supporting infrastructure assets and services.We argue for modelling that brings natural as well as built environments within the system boundaries to better understand infrastructure and to better assess sustainability. We seethe work as relevant to both the academic community and to a wide range of industry and policy applications that are working on infrastructure transition pathways towards fair, safe and sustainable society.This vision was developed through discussions between academics in preparation for the Centre for Systems Engineering and Innovation (CSEI) 10 years celebration. These rich discussions about the future of the Centre were inspired by developing themes for a celebration event, through which we have summarised the first 10 years of the Centre’s work and our vision for the future and identified six emerging research areas.


Chen P-Y, Bhatia L, Kolcun R, Boyle D, McCann Jet al., 2021, Contact-aware opportunistic data forwarding in disconnected LoRaWAN mobile networks, 40th IEEE International Conference on Distributed Computing Systems, Publisher: IEEE, Pages: 574-583

LoRaWAN is one of the leading Low Power WideArea Network (LPWAN) architectures. It was originally designedfor systems consisting of static sensor or Internet of Things (IoT)devices and static gateways. It was recently updated to introducenew features such as nano-second timestamps which open upapplications to enable LoRaWAN to be adopted for mobile devicetracking and localisation. In such mobile scenarios, devices couldtemporarily lose communication with the gateways because ofinterference from obstacles or deep fading, causing throughputreduction and delays in data transmission. To overcome thisproblem, we propose a new data forwarding scheme. Instead ofholding the data until the next contact with gateways, devices canforward their data to nearby devices that have a higher probabil-ity of being in contact with gateways. We propose a new networkmetric called Real-Time Contact-Aware Expected TransmissionCount (RCA-ETX) to model this contact probability in real-time. Without making any assumption on mobility models, thismetric exploits data transmission delays to model complex devicemobility. We also extend RCA-ETX with a throughput-optimalstochastic backpressure routing scheme and propose Real-TimeOpportunistic Backpressure Collection (ROBC), a protocol tocounter the stochastic behaviours resulting from the dynamicsassociated with mobility. To apply our approaches seamlesslyto LoRaWAN-enabled devices, we further propose two newLaRaWAN classes, namely Modified Class-C and Queue-basedClass-A. Both of them are compatible with LoRaWAN Class-Adevices. Our data-driven experiments, based on the London busnetwork, show that our approaches can reduce data transmissiondelays up to25%and provide a53%throughput improvementin data transfer performance.

Conference paper

Zhao C, Yang S, McCann JA, 2021, On the data quality in privacy-preserving mobile crowdsensing systems with untruthful reporting, IEEE Transactions on Mobile Computing, Vol: 20, Pages: 647-661, ISSN: 1536-1233

The proliferation of mobile smart devices with ever improving sensing capacities means that human-centric Mobile Crowdsensing Systems (MCSs) can economically provide a large scale and flexible sensing solution. The use of personal mobile devices is a sensitive issue, therefore it is mandatory for practical MCSs to preserve private information (the user's true identity, precise location, etc.) while collecting the required sensing data. However, well intentioned privacy protection techniques also conceal autonomous, or even malicious, behaviors of device owners (termed as self-interested), where the objectivity and accuracy of crowdsensing data can therefore be severely threatened. The issue of data quality due to untruthful reporting in privacy-preserving MCSs has been yet to produce solutions. Bringing together game theory, algorithmic mechanism design, and truth discovery, we develop a mechanism to guarantee and enhance the quality of crowdsensing data without jeopardizing the privacy of MCS participants. Together with solid theoretical justifications, we evaluate the performance of our proposal with extensive real-world MCS trace-driven simulations. Experimental results demonstrate the effectiveness of our mechanism on both enhancing the quality of the crowdsensing data and eliminating the motivation of MCS participants, even when their privacy is well protected, to report untruthfully.

Journal article

Bouazizi Y, Benkhelifa F, McCann J, 2021, Spatiotemporal modelling of multi-gateway LoRa networks with imperfect SF orthogonality, 2020 IEEE Global Communications Conference (also virtual), Publisher: IEEE, Pages: 1-7

Meticulous modelling and performance analysis ofLow-Power Wide-Area (LPWA) networks are essential for largescale dense Internet-of-Things (IoT) deployments. As Long Range(LoRa) is currently one of the most prominent LPWA tech-nologies, we propose in this paper a stochastic-geometry-basedframework to analyse the uplink transmission performance ofa multi-gateway LoRa network modelled by a Matern ClusterProcess (MCP). The proposed model is first to consider alltogether the multi-cell topology, imperfect spreading factor (SF)orthogonality, random start times, and geometric data arrivalrates. Accounting for all of these factors, we initially develop theSF-dependent collision overlap time function for any start timedistribution. Then, we analyse the Laplace transforms of intra-cluster and inter-cluster interference, and formulate the uplinktransmission success probability. Through simulation results, wehighlight the vulnerability of each SF to interference, illustratethe impact of parameters such as the network density, and thepower allocation scheme on the network performance. Uniquely,our results shed light on when it is better to activate adaptivepower mechanisms, as we show that an SF-based power allocationthat approximates LoRa ADR, negatively impacts nodes nearthe cluster head. Moreover, we show that the interfering SFsdegrading the performance the most depend on the decodingthreshold range and the power allocation scheme.

Conference paper

Bhatia L, Tomic I, Fu A, Breza M, McCann Jet al., 2021, Control communication co-design for wide area cyber-physical systems, ACM Transactions on Cyber-Physical Systems, Vol: 5, Pages: 1-27, ISSN: 2378-962X

Wide Area Cyber-Physical Systems (WA-CPSs) are a class of control systems that integrate low-powered sensors, heterogeneous actuators and computer controllers into large infrastructure that span multi-kilometre distances. Current wireless communication technologies are incapable of meeting the communication requirements of range and bounded delays needed for the control of WA-CPSs. To solve this problem, we use a Control-Communication Co-design approach for WA-CPSs, that we refer to as the C3 approach, to design a novel Low-Power Wide Area (LPWA) MAC protocol called Ctrl-MAC and its associated event-triggered controller that can guarantee the closed-loop stability of a WA-CPS. This is the first paper to show that LPWA wireless communication technologies can support the control of WA-CPSs. LPWA technologies are designed to support one-way communication for monitoring and are not appropriate for control. We present this work using an example of a water distribution network application which we evaluate both through a co-simulator(modelling both physical and cyber subsystems) and test bed deployments. Our evaluation demonstrates full control stability, with up to 50% better packet delivery ratios and 80% less average end-to-end delays when compared to a state of the art LPWA technology. We also evaluate our scheme against an idealised, wired,centralised, control architecture and show that the controller maintains stability and the overshoots remain within bounds.

Journal article

Benkhelifa F, McCann JA, 2020, Resource allocation for NOMA-based LPWA networks powered by energyharvesting, IEEE Wireless Communications and Networking Conference (WCNC), Publisher: arXiv

In this paper, we explore perpetual, scalable, Low-powered Wide-area networks (LPWA). Specifically we focus on the uplink transmissions of non-orthogonal multiple access (NOMA)-based LPWA networks consisting of multiple self-powered nodes and a NOMA-based single gateway. The self-powered LPWA nodes use the "harvest-then-transmit" protocol where they harvest energy from ambient sources (solar and radio frequency signals), then transmit their signals. The main features of the studied LPWA network are different transmission times-on-air, multiple uplink transmission attempts, and duty cycle restrictions. The aim of this work is to maximize the time-averaged sum of the uplink transmission rates by optimizing the transmission time-on-air allocation, the energy harvesting time allocation and the power allocation; subject to a maximum transmit power and to the availability of the harvested energy. We propose a low complex solution which decouples the optimization problem into three sub-problems: weassign the LPWA node transmission times (using either the fair or unfairapproaches), we optimize the energy harvesting (EH) times using aone-dimensional search method, and optimize the transmit powers using aconcave-convex (CCCP) procedure. In the simulation results, we focus on Long Range (LoRa) networks as a practical example LPWA network. We validate our proposed solution and we observe a $15\%$ performance improvement when using NOMA.

Conference paper

This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.

Request URL: Request URI: /respub/WEB-INF/jsp/search-html.jsp Query String: respub-action=search.html&id=00333341&limit=30&person=true