Bryan Ng

A survey of routing and channel assignment in multi-channel multi-radio WMNs

Wireless mesh networks provide cost-effective coverage and high network performance by utilising advanced radio frequency technology such as multiple channel multiple radio (MCMR). In spite of the advantages in MCMR wireless mesh networks, some wireless mesh networks still suffer from performance issues such as throughput degradation and unacceptable latency when network size increases. Existing studies focused on solving the performance issues from a single aspect such as routing or channel assignment. A major limitation of these studies is its failure to consider the interaction between routing and channel assignment which is a key factor influencing network performance.Different from the existing studies, we investigate routing and channel assignment in tandem. A new taxonomy is developed for evaluating the capabilities of current MCMR wireless mesh networks and used to organise past and ongoing research on routing and channel assignment algorithms. Our work provides a comprehensive analysis of existing studies from the interactive factors between routing and channel assignment. We aim to stimulate and guide the future research on joint design between routing and channel assignment in MCMR wireless mesh networks that can optimise network performance better than the individual solutions. Graphical abstractDisplay Omitted HighlightsSurvey existing routing and channel assignment algorithms (RCAA).Construct a new taxonomy for joint optimisation of RCAA.Analyse weakness and strengths of RCAA in the taxonomy.Suggest directions for optimisation in joint RCAA.

Forwarding Schemes for EM-based Wireless Nanosensor Networks in the Terahertz Band

Electromagnetic-based Wireless NanoSensor Networks (EM-WNSNs) operating in the TeraHertz (THz) band (0.1 THz--10 THz) has been in focus recently because of potential applications in nano-scale scenarios. However, one major hurdle for advancing nano-scale communications is the lack of suitable networking protocols to address current and future needs of nanonetworks. Working together with routing that finds the path from a source to destination, forwarding is a networking task of sending a packet to the next-hop along its path to the destination. While forwarding has been straightforward in traditional wired networks, forwarding schemes now play a vital role in determining wireless network performance. In this paper, we propose a channel-aware forwarding scheme and compare it against traditional forwarding schemes for wireless sensor networks. To fit the peculiarity of EM-WNSNs, the channel-aware forwarding scheme makes forwarding decision considering the frequency selective pecularities of the THz channel which are undesirable from a networking perspective. It is shown through simulation that the proposed channel-aware forwarding scheme outperforms traditional forwarding schemes in terms of the end-to-end capacity while maintaining comparable performance for delay.

Developing a traffic classification platform for enterprise networks with SDN: Experiences & lessons learned

Software Defined Networking (SDN) is an innovative approach to networking architecture that opens up avenues to create a whole new class of networking functionality. While data centre networks are steadily adopting the SDN approach with considerable success, other areas of networking such as network access control, load balancing and traffic classification remain nascent. Traffic classification in SDN is relatively experimental and attempts for SDN traffic classification to become a viable solution for enterprise networks require additional investigation. This paper reports on the practical experiences and lessons learned while developing an SDN based traffic classification platform for an enterprise network. We use the platform to demonstrate the feasibility of SDN based traffic classifiers by evaluating against a set of desired outcomes. We make note of the design choices using the currently available technologies that may be helpful to networks operators considering deploying their own solution. We conclude the paper with suggested changes to better address limitations for software traffic classification that will remove the need for workarounds with future versions of OpenFlow.

Scheduling deadline constrained scientific workflows on dynamically provisioned cloud resources

Abstract Commercial cloud computing resources are rapidly becoming the target platform on which to perform scientific computation, due to the massive leverage possible and elastic pay-as-you-go pricing model. The cloud allows researchers and institutions to only provision compute when required, and to scale seamlessly as needed. The cloud computing paradigm therefore presents a low capital, low barrier to operating dedicated HPC eScience infrastructure. However, there are still significant technical hurdles associated with obtaining sufficient execution performance while limiting the financial cost, in particular, a naive scheduling algorithm may increase the cost of computation to the point that using cloud resources is no longer a viable option. The work in this article concentrates on the problem of scheduling deadline constrained scientific workloads on dynamically provisioned cloud resources, while reducing the cost of computation. Specifically we present two algorithms, Proportional Deadline Constrained (PDC) and Deadline Constrained Critical Path (DCCP) that address the workflow scheduling problem on such dynamically provisioned cloud resources. These algorithms are additionally extended to refine their operation in task prioritization and backfilling respectively. The results in this article indicate that both PDC and DCCP algorithms achieve higher cost efficiencies and success rates when compared to existing algorithms.

Network health and e-Science in commercial clouds

This paper explores the potential for improving the performance of e-Science applications on commercial clouds through the detailed examination, and characterization, of the underlying cloud network using network tomography. Commercial cloud providers are increasingly offering high performance and GPU-enabled resources that are ideal for many e-Science applications. However, the opacity of the clouds internal network, while a necessity for elasticity, limits the options for e-Science programmers to build efficient and high performance codes. We introduce health indicators, markers, metrics, and score as part of a network health system that provides a model for describing the overall network health of an e-Science application. We then explore the suitability of a range of tomographic techniques to act as health indicators using two testbeds-the second of which spanned one hundred AWS instances. Finally, we evaluate our work using a real-world medical image reconstruction application. We identify and characterize network performance in commercial clouds.An overall health system is constructed using tomographic probes to establish and compare an instances network performance.We deploy the health system over a testbed of 100 AWS instances and explore its ability to scale.We apply the health system to a medical imaging e-Science application and demonstrate performance benefits.

UbiTouch: ubiquitous smartphone touchpads using built-in proximity and ambient light sensors

Smart devices are increasingly shrinking in size, which results in new challenges for user-mobile interaction through minuscule touchscreens. Existing works to explore alternative interaction technologies mainly rely on external devices which degrade portability. In this paper, we propose UbiTouch, a novel system that extends smartphones with virtual touchpads on desktops using built-in smartphone sensors. It senses a users finger movement with a proximity and ambient light sensor whose raw sensory data from underlying hardware are strongly dependent on the fingers locations. UbiTouch maps the raw data into the fingers positions by utilizing Curvilinear Component Analysis and improve tracking accuracy via a particle filter. We have evaluate our system in three scenarios with different lighting conditions by five users. The results show that UbiTouch achieves centimetre-level localization accuracy and poses no significant impact on the battery life. We envisage that UbiTouch could support applications such as text-writing and drawing.

Queueing Analysis of Software Defined Network with Realistic OpenFlow–Based Switch Model

Software Defined Networking (SDN) is the latest network architecture that does for networking what virtualisation did for servers in data centres. In SDN, separation of the control plane from the data plane brought about new flexibility in the routing of flows through the network. Closely associated with SDN is OpenFlow, the most widely used protocol governing the information exchange between the data plane (switching devices) and the control plane (controller). The ease of implementing and testing new schemes in SDN has prompted many researchers to adopt the experimental and prototyping approach to validate their ideas. Consequently, there has been very little work done to evaluate the performance of SDN and/or OpenFlow-based networks analytically. While the experimentation approach in validation has merits, analytical modelling provides valuable insights by making explicit the dependence of SDN performance on chosen parameters. In this paper: (i) we propose a queueing model of an OpenFlow-based SDN that takes into account classful treatment of packets arriving at a switch and (ii) derive an exact analysis of the proposed queueing model.

Heavy Hitter Detection and Identification in Software Defined Networking

In a large network, it is often important to be able to detect high-volume traffic in near real-time. Existing work on the detection and identification of such high volume traffic (so-called heavy hitters) is typically delegated to individual nodes and often relies on deep packet inspection and/or packet sampling. However, these techniques have well known limitations in terms of its ability to scale with network size. Inspired by the capabilities of Software Defined Networking (SDN), we explore a novel heavy hitter detection solution based on understanding connections between traffic statistics and OpenFlow rules. Our approach relies on mining traffic statistics (e.g. port bitrate) and forwarding table entry (FTE) to improve heavy hitter detection. The rationale behind this approach are (i) the information is readily available with minimal overheads, thus it scales better with increasing network size; and (ii) the FTEs and traffic statistics provide different vantage for detection and identification of heavy hitters. We evaluate the effectiveness and accuracy of our proposed heavy hitter detection algorithm on a test bed as a proof-of-concept. The test results show that our approach to heavy hitter detection simultaneously achieves considerable accuracy and good scalability.

An Automated Tool Profiling Service for the Cloud

Cloud providers offer a diverse set of instance types with varying resource capacities, designed to meet the needs of a broad range of user requirements. While this flexibility is a major benefit of the cloud computing model, it also creates challenges when selecting the most suitable instance type for a given application. Sub-optimal instance selection can result in poor performance and/or increased cost, with significant impacts when applications are executed repeatedly. Yet selecting an optimal instance type is challenging, as each instance type can be configured differently, application performance is dependent on input data and configuration, and instance types and applications are frequently updated. We present a service that supports automatic profiling of application performance on different instance types to create rich application profiles that can be used for comparison, provisioning, and scheduling. This service can dynamically provision cloud instances, automatically deploy and contextualize applications, transfer input datasets, monitor execution performance, and create a composite profile with fine grained resource usage information. We use real usage data from four production genomics gateways and estimate the use of profiles in autonomic provisioning systems can decrease execution time by up to 15.7% and cost by up to 86.6%.

A deadline constrained critical path heuristic for cost-effectively scheduling workflows

Effective use of elastic heterogeneous cloud resources represents a unique multi-objective scheduling challenge with respect to cost and time constraints. In this paper we introduce a novel deadline constrained scheduling algorithm, Deadline Constrained Critical Path (DCCP), that manages the scheduling of workloads on dynamically provisioned cloud resources. The DCCP algorithm consists of two stages: (i) task prioritization, and (ii) task assignment, and builds upon the concept of Constrained Critical Paths to execute a set of tasks on the same instance in order to fulfil our goal of reducing data movement between instances. We evaluated the normalized cost and success rate of DCCP and compared these results with IC-PCP. Overall, DCCP schedules with lower cost and exhibits a higher success rate in meeting deadline constraints.