Search

This article presents the research carried out in developing and targeting a novel real-time Dynamic Spectrum Access (DSA) Frequency Spread Filter Bank Multicarrier (FS-FBMC) transmitter prototype to programmable ‘ZynqSDR’ Software Defined Radio (SDR) hardware, and introduces a series of experiments used to validate the design’s ‘cognitive’ DSA capabilities. This transmitter is a proof of concept, that uses DSA techniques to enable Secondary Users (SUs) to access the band traditionally used for FM Radio broadcasting (88-108 MHz), and establish data communication channels in vacant parts of the FM Radio Primary User (PU) spectrum using a multicarrier modulation scheme with a Non Contiguous (NC) channel mask. Once implemented on the hardware, the transmitter is subjected to various FM Radio environments sampled from around Central Scotland, and it is demonstrated that it can dynamically adapt its NC transmitter mask in real time to protect the FM Radio signals it detects. A video is presented of this dynamic on-hardware spectral reconfiguration, and the reader is encouraged to view the video to appreciate the responsiveness of the design. An investigation into potential FBMC guardband sizes is carried out, with initial findings indicating a guardband of 200 kHz (either side of an FM Radio station) is required in order to prevent interference with the PUs. This article also demonstrates the capabilities of the MATLAB and Simulink Model Based Design Zynq-Based Radio workflow, and provides a case study and reference design that we feel other researchers working in this field can benefit from.

We explore the use of dependent types to address the disparity between the theory and the practical hardware description of DSP circuits. After discussing an approach to modeling synchronous circuit behaviour in Idris (a pure functional language with dependent types), two DSP case studies are introduced — an FIR filter with optimal wordlengths and a CIC decimator with register pruning. Both of these scenarios prove difficult to describe in a parameterised fashion using traditional HDLs and, as such, many implementations rely on ad hoc circuit generators which are challenging to test and evaluate. This work demonstrates that such circuits are readily described in an environment with dependent types. Dependent types can also encode various contracts between the IP designer and its user. These contracts are automatically verified by the Idris type checker before compilation, precluding many common mistakes in IP development and evaluation.

This demonstration will present a single chip spectrum analyser that has been developed using the Xilinx Zynq Radio Frequency System on Chip (RFSoC), and the PYNQ software framework. The design uses the RFSoC’s high speed Radio Frequency Analogue to Digital Converters (RF ADCs) to capture 2.048 GHz of instantaneous bandwidth and directly sample signals at frequencies up to 4.096 GHz using higher order Nyquist Zone techniques. All signal processing, software control, and graphical user interface functions are hosted entirely on the same RFSoC chip. As well as presenting the functionality achieved by the spectrum analyser, we will describe its underlying architecture and demonstrate its use in exploring the radio spectrum by showing examples of ambient signals.

This paper studies the capabilities of a proposed lossy, grouped-bin FFT quantisation compression method for targeting Off-The-Air (OTA) Radio Frequency (RF) signals. The bins within a 512-point Fast Fourier Transform (FFT) are split into groups of adjacent bins, and these groups are each quantised separately. Additional compression can be achieved by setting groups which are not deemed to contain significant information to zero, based on a pre-defined minimum magnitude threshold. In this paper, we propose two alternative methods for quantising the remaining groups. The first of these, Grouped-bin FFT Threshold Quantisation (GFTQ), involves allocating quantisation wordlengths based on several pre-defined magnitude thresholds. The second, Grouped-bin FFT Error Quantisation (GFEQ), involves incrementing the quantisation wordlength for each group until the calculated quantisation error falls below a minimum error threshold. Both algorithms were tested for a variety of signal types, including Digital Private Mobile Radio 446 MHz (dPMR446), which was considered as a case study. While GFTQ allowed for higher Compression Ratios (CR), the compression process resulted in added quantisation noise. The GFEQ algorithm achieved lower CRs, but also lower noise levels across all test signals.

Rural areas are often neglected while deploying newer mobile technologies. Hence, these place are digitally disconnected from the world. To overcome this challenge, 5G network slicing supporting multi-tenancy, also known as neutral host network, is studied to improve rural connectivity. The infrastructure provider (InP) deploys the rural 5G network and mobile network operators (MNOs) lease the slices from InP to serve their end-users. This aims to study the value network configuration (VNC) for the 5G network slicing architecture to understand the possible business model. As a result, three configurations are defined driven by micro-operator, MNO and community end-users respectively. The business models are constructed using SWOT analysis and business canvas models. The revenue streams for the proposed rural network are analyzed.

The Line Hough Transform (LHT) is an effective line detection algorithm for digital images. To meet real-time requirements, Field Programmable Gate Arrays (FPGAs) are often chosen to accelerate the LHT. However, many LHT architectures select different design parameters and discretisation steps for the Hough Parameter Space (HPS). The work presented in this paper describes a novel evaluation platform for FPGA architectures of the LHT. Our system is named the Hough Evaluation Platform (HEP) and can be used to visualise and inspect the HPS produced from LHT architectures that use different design parameters. Architectures are compared by evaluating the HPS using a unique measurement named the Peak to Mean Vote Ratio (PMVR). Our system employs the PYNQ framework on a Xilinx Zynq MPSoC device for the visualisation of the HPS and also determines the processing time of a given LHT architecture. The HEP has been implemented on a XCZU7EV-2e device and can operate up to a target frequency of 250 MHz.

Neural networks are increasingly being moved to edge computing devices and smart sensors, to reduce latency and save bandwidth. Neural network compression such as quantization is necessary to fit trained neural networks into these resource constrained devices. At the same time, their use in safety-critical applications raises the need to verify properties of neural networks. Adversarial perturbations have potential to be used as an attack mechanism on neural networks, leading to "obviously wrong" misclassification. SMT solvers have been proposed to formally prove robustness guarantees against such adversarial perturbations. We investigate how well these robustness guarantees are preserved when the precision of a neural network is quantized. We also evaluate how effectively adversarial attacks transfer to quantized neural networks. Our results show that quantized neural networks are generally robust relative to their full precision counterpart (98.6%–99.7%), and the transfer of adversarial attacks decreases to as low as 52.05% when the subtlety of perturbation increases. These results show that quantization introduces resilience against transfer of adversarial attacks whilst causing negligible loss of robustness.

Floating point Convolutional Neural Networks(CNNs) are computationally expensive and deeper networks can be impractical to deploy on FPGAs – consuming a large number of resources and power, as well as having lengthy development times. Previous work has shown that CNNs can be quantised heavily using fixed point arithmetic to combat this without significant loss in classification accuracy. We aim to quantize an existing CNN architecture for radio modulation classification to 2-bit weights and activations, while retaining a level of accuracy close to the original paper, for deployment on a Zynq System on Chip (SoC). To improve the development time for hardware synthesisable CNNs, we make use of MATLAB System Objects and HDL Coder. The PYNQ framework is presented as a practical means for accessing the functionality of the CNN. Our preliminary results show a high classification accuracy even with 2-bit weights and activations.

This paper decomposes the algorithmic parameters that affect the accuracy and parallel run times of mean shift segmentation. Following Comaniciu and Meer [1], rather than perform calculations in the feature space of the image, the joint spatial-range domain is represented by the image space, with feature space information associated with each point. We report parallel speedup and segmentation accuracy using a standardised segmentation dataset and the Probabilistic Rand index (PRI) accuracy measure. Changes to the algorithmic parameters are analysed and a sweet spot between PRI and run time is found. Using a range window radius of 20, spatial window radius of 10 and threshold of 50, the PRI is improved by 0.17, an increase of 34% which is comparable to state of the art. Mean shift clustering run time is reduced by 97% with parallelism, a speedup of 32 on a 64-core CPU.

In the last few years Machine Learning (ML) has seen explosive growth in a wide range of research fields and industries. With the advancements in Software Defined Radio (SDR), which allows more intelligent, adaptive radio systems to be built, the wireless communications field has a number of opportunities to apply ML techniques. In this paper, a novel approach to demodulation using a Sequence to Sequence (Seq2Seq) model is proposed. This type of model is shown to work effectively with PSK data and also has a number of useful properties that are not present in other machine learning algorithms. A basic Seq2Seq implementation for BPSK and QPSK demodulation is presented in this paper, and learned properties such as Automatic Modulation Classification (AMC), and ability to adapt to different length input sequences, are demonstrated. This is an exciting new avenue of research that provides considerable potential for application in next generation 5G networks.

Visual attention modelling characterises the scene to segment regions of visual interest and is increasingly being used as a pre-processing step in many computer vision applications including surveillance and security. Smart camera architectures are an emerging technology and a foundation of security and safety frameworks in modern vision systems. In this paper, we present a dataflow design of a visual saliency based camera architecture targeting a heterogeneous CPU+FPGA platform to propose a smart camera network infrastructure. The proposed design flow encompasses image processing algorithm implementation, hardware & software integration and network connectivity through a unified model. By leveraging the properties of the dataflow paradigm, we iteratively refine the algorithm specification into a deployable solution, addressing distinct requirements at each design stage: from algorithm accuracy to hardware-software interactions, real-time execution and power consumption. Our\ud design achieved real-time run time performance and the power\ud consumption of the optimised asynchronous design is reported at only 0.25 Watt. The resource usages on a Xilinx Zynq platform remains significantly low.

One of the key challenges for state of the art radio systems is enabling efficient utilisation of the Radio Frequency (RF) spectrum. Licensed frequency bands are often under-utilised in both time and geographical location and thus the opportunity exists for secondary users to transmit in these bands, provided that they do not interfere significantly with the operation of the primary licensed user. A proposed method for exploiting this opportunity is Cognitive Radio (CR) wherein the secondary user is able to modify its transmissions based on observation of the operating RF environment. Orthogonal Frequency Division Multiplexing (OFDM) is the enabling technology for many modern communications standards such as IEEE 802.11a (WiFi) and 4G Long Term Evolution (LTE). Therefore, facilitating robust and cost effective detection of OFDM signals is a key problem for the design of secondary user CR systems. In this paper, we derive and assess the performance of a low complexity detection scheme that exploits the inherent cyclostationarity of OFDM signals. We then present details of its implementation on a Xilinx Artix 7 FPGA and compare the resource cost of the proposed detector with another low complexity detection algorithm found in the literature.

The ubiquitous deployment of IEEE 802.11 based Wireless Mesh Networks (WMNs) has facilitated non-real-time applications effectively to date. However the advancement in IEEE 802.15.4 Wireless Sensor Network (WSN) technology has dramatically increased the volume and range of applications available under the umbrella of the Internet of Things (IoTs). Such applications cross the divide of simple environmental data collectors to highly specific multimedia real-time applications demanding high Quality of Service (QoS) guarantees. However delivering QoS guarantees is difficult to achieve when deployed in a hostile environment such as that found in the Precision Agriculture (PA) industry. Difficulties in wireless propagation attenuation delivers an uncertainty to the network design phase. Therefore the motivation of this research is a requirement for real-time characterisation and derivation of suitable propagation models to support network design and deployment planning. In this work, empirical results were collected from a heterogeneous wireless deployment (the real-world ECOMESH test-bed) in a native woodland environment. We derive two empirical models which can be utilised in predicting attenuation network performance in two different scenarios. These models deliver the first phase of a Dynamic Network Management System for maintaining QoS guarantees in heterogeneous wireless networks.

The Internet of Things (IoT) network is typified by deployments such as a Wireless Sensor Network (WSN). These WSN's are commonly deployed within a target area for a specific task such as farm monitoring and asset movement control. A group of sensors are typically led by a Personal Area Network Coordinator (PANC). In some situations like in large scale greenhouse monitoring, mobile PANC are required to eliminate energy-hole issues and hence improve the performance of the WSN. The energy-hole issue is a problem caused by the depleted cluster head nodes. In this paper we analyse the performance of a WSN with a mobile PANC. Using inputs from previous empirical studies we designed a scenario of a mobile PAN coordinator using a simulation package. The performance was evaluated. The mobile PANC speed and contact time is monitored and the average network throughput and delay are measured. Based on an analysis of these results a comprehensive set of guidelines are provided for operators of IoT networks. These guidelines optimise the network performance and ultimately lead to cost savings for network providers.

In order to deliver transportation project on time and within budget, project planners and decision makers need to understand the desirable outcomes of projects or programs to make intelligent decisions. In addition, they also need effective tools and techniques to guide those decisions. In development of a guide that identifies available tools and techniques to foster useable and improved practices for key stakeholders of transportation projects, this paper as a preliminary study assessed current utilization of project delivery value improvement programs such as value engineering (VE), risk analysis/management (RA/RM) and constructability reviews (CR) for transportation projects. To determine whether states have adapted any of these programs or other innovative project delivery techniques, state DOT initiatives were studied through a review of past research and reports. Also, the collection of documents relevant to DOT practices that have been used as a literature review to identify and design the survey questions. Value Management Strategic (VMS), academic advisors at Portland State University (PSU), and project delivery experts were consulted to advice and validate the survey questions before and after the distribution stage. Then, three surveys of VE, RA, and CR were carried out with state DOT staff. Finally, the survey results were discussed in-depth in this paper to address the main issues of these programs and the level of its conducting within the project delivery processes related to transportation construction projects.

Wireless Sensor Networks (WSNs) are formed with multiple sensor nodes that use the IEEE 802.15.4 protocol. The WSN can operate in 16 different channels in the unlicensed Industrial, Scientific and Manufacturing (ISM) 2.4 GHz spectrum. The WLAN or IEEE 802.11g/n protocol also operates in the same spectrum. Additional problems arise when the WSN is mobile. Mobile WSN's are more efficient in scenarios where mobility is needed to collect data from remote target areas. Moving nodes may encounter overlapping frequencies from WLAN coexistence in its journey. Therefore, WSN channels have to be allocated intelligently to avoid channels overlapping with the WLAN. However with mobile WSN the WLAN networks encountered may have random spectrum allocation depending on the WLAN technology in place. A good resource allocation approach is needed for the sensor nodes to dynamically change their working channel to avoid the overlapping channels which cause interference to the network (either WSN or WLAN) which impacts the service delivery. This paper focuses on the interference when the mobile WSN and WLAN both exist in the same building. The analysis presented is based on the percentage of packets dropped and percentage of packet re-transmitted for each sensor node in the experiments. The outcome of this research will be valuable to research and development of WSN applications in terms of reliable service delivery.

Profile guided optimisation is a common technique used by compilers and runtime systems to shorten execution runtimes and to optimise locality aware scheduling and memory access on heterogeneous hardware platforms. Some profiling tools trace the execution of low level code, whilst others are designed for abstract models of computation to provide rich domain-specific context in profiling reports. We have implemented mean shift, a computer vision tracking algorithm, in the RVC-CAL dataflow language and use both dynamic runtime and static dataflow profiling mechanisms to identify and eliminate bottlenecks in our naive initial version. We use these profiling reports to tune the CPU scheduler reducing runtime by 88%, and to optimise our dataflow implementation that reduces runtime by a further 43% — an overall runtime reduction of 93%. We also assess the portability of our mean shift optimisations by trading off CPU runtime against resource utilisation on FPGAs. Applying all dataflow optimisations reduces FPGA design space significantly, requiring fewer slice LUTs and less block memory.

Transcranial Sonography is a clinical procedure for monitoring cerebral blood flow in patients suspected of suffering from severe blood flow disruption in the brain. The Transcranial Doppler (TCD) clinicians are currently only guided by their experience and the temporal bone surface to position the Doppler probe on a suitable site for TCD. This site is known as an acoustic window. Neither experience nor the bone surface can sufficiently improve TCD procedure time for new patients. A suitable acoustic window for transcranial Doppler ultrasound can be difficult to find, and in some patients, no such acoustic window is possible. It is hypothesized that the visibility of the temporal bone in an acousto-spectroscopic image suggests an appropriate transcranial Doppler ultrasound acoustic window. The aim of this research is to investigate the ultrasonic detection of such a pathological acoustic window. It was anticipated that an acoustic window would best be detected by its Acousto-spectroscopic properties in an attenuation-velocity image. These images contrast hard tissue from soft tissue based on their acousto-spectroscopic properties. Identifying where the temporal bone is the thinnest could potentially indicate the best location for performing a Transcranial Doppler procedure. Experimental investigations on a specially designed layered transcranial phantom were carried out to detect the region where the temporal bone is the thinnest and hence suggest an adequate acoustic window. It was shown that a plausible transcranial acoustic window is visible in an ultrasound image, a previously unreported and very promising result. Attenuation-velocity images showed greater boundary contrast between hard and soft tissue. However, hard and soft tissue regions appear similar in comparison to B-Mode images. It was determined that nonuniform scattering from inhomogeneous life-like tissues causes attenuation-velocity images to display less media contrast such as bone versus soft tissue contrast. This research is the first to report that B-Mode images are currently the best method for visually detecting a plausible potential acoustic window for Transcranial Doppler.

This paper investigates the problem of robot visual homing — the navigation to a goal location by a mobile robot using visual sensory input. The visual homing approach taken is to consider the flow vectors between a robot's current view and a desired milestone view. The flow vectors can be used to determine an angular velocity command that attempts to align the two views under a constant forward speed. Experiments with a mobile robot have been conducted following the teach-replay approach. By using a sequence of milestone images taken successively along a path, preliminary results show that a robot can successfully repeat the path and navigate to its goal autonomously. The method should be useful for route following and other applications involving visual navigation.

Communications devices operating in the TV white space (TVWS) spectrum will be strictly regulated, requiring compliance with spectral masks to protect incumbent users and sufficient frequency agility to allow access to numerous frequency bands at different times and locations. Therefore, future designs sampling directly at radio frequency (RF) have been proposed. The purpose of this paper is to demonstrate an implementation of such a transceiver at a scaled-down frequency implemented on the OMAP-L137 evaluation module, whereby the RF link can be replaced by the device's audio I/O, thus enabling easier observation and algorithm testing for students.

In this paper we consider a robust transceiver design for two hop non-regenerative multiple-input multiple-output (MIMO) relay networks with imperfect channel state information (CSI). The transceiver consists of Tomlinson Harashima Pre-coding (THP) at the source with a linear precoder at the relay and linear equalisation at the destination. Under the assumption that each node in the network can acquire statistical knowledge of the channel in the form of a channel mean and estimation error covariance, we optimise the processors to minimise the expected arithmetic mean square error (MSE) subject to transmission power constraints at the source and relay. Simulation results demonstrate the robustness of the proposed transceiver design to channel estimation errors.

Class D amplifiers are based on a low-bit pulse width modulated (PWM) signal reconstructed through the lowpass characteristic of the loudspeaker. To create a PWM signal, a non-linear conversion from a standard pulse code modulated (PCM) signal has to be performed. The non-linear distortion created in this conversion process can be minimised by non-linear predistortion based on a Wiener-Hammerstein system. In this paper, we compare single and double-sided PWM conversion, whereby the former exhibits a higher resolution but also higher non-linearity compared to the latter. The trade-off between both approaches in creating a highly linear high performance class D audio amplifier is demonstrated by means of simulations and an implementation.

In a Wireless Sensor Network (WSN) the provision of a Guaranteed Time Slot (GTS) in a beacon enabled mode is best suited for applications with low latency and high Quality of Service (QoS) requirements such as those of real time traffic. However the advancements in technology have dramatically increased the volume of applications which require access to this real time service provision in such areas as healthcare, asset tracking, environmental monitoring, and military projects. Each super-frame structure contains only seven GTS's with the remaining time slots provisioned for the Contention Access Period (CAP) where applications with higher tolerance of latency vie for a position to transmit data. The restrictions within the Contention Free Period (CFP) to seven time slots per superframe structure, places enormous responsibilities on the network designers to dimension networks that can guarantee a high QoS. This work plans to apply decay rate analysis to GTS provisioning. Appropriate buffer dimensioning for specific applications can be achieved by predetermining the probability of buffer capacity. This will enable adequate parameter settings to be configured within a WSN to meet all real time traffic needs.

With more and more countries opening up sections of unlicensed spectrum for use by TV White Space (TVWS) devices, the prospect of building a device capable of operating in more than one world region is appealing. The difficulty is that the locations of TVWS bands within the radio spectrum are not globally harmonised. With this problem in mind, the purpose of this paper is to present a TVWS transceiver design which is capable of being reconfigured to operate in both the UK and US spectrum. We present three different configurations: one covering the UK TVWS spectrum and the remaining two covering the various locations of the US TVWS bands.

The use of multiple frequency bands within a wireless network allows the advantages of each band to be exploited. In this paper we discuss how “HopScotch”, a rural wireless broadband access test bed running in the Scottish Highlands and Islands, uses both 5 GHz and ultra high frequency “white space” bands to offer large data rates and expansive coverage whilst reducing the number of base stations or required transmission power. This reduction in energy consumption allows HopScotch to provide a low-cost and green solution for rural broadband delivery.

The provisioning of a Quality of Service (QoS) for applications requiring real-time service demands low latency, low Bit Error Rate (BER) and bounded end to end delay guarantees. Facilitating this within a Wireless Sensor Network (WSN) is Guaranteed Time slots (GTS's) activated using a beacon enabled mode and under the governance of the accepted protocol of the IEEE 802.15.4 standard. However the limitation in design of only seven GTS per superframe duration pre-determines a problem of allocation when requirements by applications in need of good QoS exceed the supply of GTS's. This results in the QoS aware traffic requiring differentiation of service in the buffer of the Personal Area Network Coordinator (PANC). A question arises as to the true ability of service dimensioning of networks without the prior knowledge of buffer allocation. In some WSN the buffer shares memory with the application and radio stack. This memory requirement will vary depending on the application. This research complements a previous paper in which the application of Decay Rate Analysis in relation to the dimensioning of buffer allocation of the various nodes was analysed using simulation and analytical methodology. The results illustrated the probability of queue occupancy for a specific application namely VoIP. This paper proposes that by employing the decay rate analysis a better understanding of buffer requirements at the PANC can be determined for different sensor applications. This analysis will further enhance the pre-deployment dimensioning of a network for greater operational efficiency and overall cost effectiveness.

In appearance based SLAM (Simultaneous Localisation and Mapping), a robot typically represents its environment through a set of acquired images that are associated with nodes in a topological map. Rather than storing every acquired image, which can be memory intensive, a selection of images (keyframes) representative of the places visited can be stored. Keyframe detection (i.e. choosing when to add a new keyframe) typically requires a means of determining the similarity of images. In this paper we develop three new metrics for computing image similarity. The metrics are based on the degree of feature-flow between features matched in a reference image (e.g. previous keyframe) and a test image (e.g. candidate keyframe), where a low degree of feature-flow indicates a high image similarity value. The new metrics and an existing metric are computed for synthetic and real data and their performance is evaluated with respect to a number of attributes important for keyframe detection. The results suggest similarity metrics based on feature-flow are preferable for use in keyframe detection.

Wireless radio links present one option for rural broadband where population is sparse and many of the existing technologies such as optical fiber or broadband via DSL are not economical due to rugged/mountainous terrain. In this paper, we present a broadband test bed running in the Scottish Highlands and Islands, which is based on a relay network of low-power and cost efficient base stations. In addition to a 5GHz network, the test-bed also provides an overlay UHF white space network sharing the same infrastructure. The low-power base stations forming the relay nodes are supplied by wind and solar energy, establishing a green technology with low requirements on planning permission. Overall, the system offers a robust and independent medium to support broadband communication in rural settings.

We present a white space communications test bed running in the Scottish Highlands and Islands, and discuss its feasibility for smart grid communications. The network aims to serve communities that have great potential for distributed generation of electricity, by means of wind, water, and tidal power. However, smart grid applications such as remote meter reading and load balancing are impaired by the scarcety or lack of communications infrastructure in remote rural areas such as the Scottish Highlands and Islands. We argue that the proposed system is based on a network of energy self-sufficient radio relay nodes that make it a robust and independent medium to support smart grid communications in rural settings.

In this paper we consider the design of minimum mean square error (MMSE) transceivers for non-regenerative multiple input multiple output (MIMO) relay systems. Our design utilises Tomlinson Harashima precoding (THP) at the source along with linear processors in each stage of the network. Assuming full channel state information (CSI) is available at each node in the network the various processors are jointly optimised to minimise the system arithmetic mean square error (MSE) whilst abiding by average power constraints at both the source and relay terminals in the network. Simulations show that the proposed schemes outperform existing methods in terms of bit error ratio (BER).

Empirical results from a performance study of an Integrated Wireless Sensor Network (WSN) and Wireless Local Area Network (WLAN) are presented in this paper. Integration of heterogeneous wireless technologies using different modulation techniques such as direct-sequence spread spectrum (DSSS) and orthogonal frequency division multiplexing (OFDM) require careful deployment planning and configuration if network performance is to be optimized. The impact on network performance of the WLAN and WSN is investigated through a series of experiments on the Athlone Institute of Technology (AIT) campus network under realistic load conditions. Novel software tools from Motorola are used in this research for fast efficient deployment of WLAN's. These software tools utilize satellite maps of the deployment area to predict coverage and range. Combined with management tools for integrated channel assignment, this should provide performance enhancement capabilities. Our findings show that at least more than four channels of WSN were under interference when WSN and WLAN coexisted. The expression for a channel assignment model is adapted from Chowdhury et al [1]. The reference power variables were taken from IEEE802.11g under real traffic loading with eight channels being considered. Achieving effective channel allocation for WSNs offers better performance due to less retransmission and inherently reduces energy consumption. This paper will be of interest to organisations where critical information retrieval over wide area networks in hostile environments (such as disaster recovery situations) is required as well as traditional network deployments.

This paper presents a wireless distributed system to measure transient electromagnetic field in power substations. Taking advantage of flexible short-range wireless technique, the system allows flexible multi-node simultaneous measurement while is of compact size and low cost.

Recently, the Third Generation Partnership standards bodies (3GPP/3GPP2) have defined a channel model for multiple input multiple output (MIMO) systems. This model is a 2D model where the propagating plane waves are assumed to arrive only from the azimuthal direction and does not include the elevation domain. As a result of this assumption the derived angle of arrival (AoA) distribution is characterized only by the azimuth direction of these waves. In this paper the distribution of scatterers is modeled in a novel 3D approach. The Von Mises Fisher (VMF) probability density function (PDF) is used to describe their distribution within the propagation environment in azimuth and coelevation. As a result the AoA distribution of the incoming multipaths is characterized in the same way. More specifically the model proposed uses a mixture of VMF distributions. A mixture can be composed of any number of clusters and this is environment/clutter specific. The parameters of the individual clusters of scatterers within the mixture are derived and an estimation of those parameters is achieved using the soft-expectation maximization (EM) algorithm. The results indicate that the proposed model fits well with MIMO experimental data obtained from a measurement campaign in Germany. The measurements obtained from the RUSK channel sounder were post-processed using the RiMAX algorithm. The data was subsequently clusterized using the soft-EM for mixtures of VMF distributions.

The Research Consortium in Speckled Computing is a multidisciplinary grouping of computer scientists, electronic engineers, physicists and electrochemists with the aim of realising wireless networks of miniature programmable speck devices. Each speck is designed to be autonomous with its own rechargeable battery and combines the capabilities to sense, process data and communicate wirelessly. This review paper outlines the unique multidisciplinary approach to the design of networks of specks and how the design of energy-efficient specks should take into account the intimate interactions between the different sub-systems which make up the specks, such as the energy source, wireless communication, physical layer design and the Medium Access Control (MAC) and routing protocols.

This paper presents preliminary investigations of the short range (~10 cm) narrowband wireless channel appropriate to specks operating in the 2.4 GHz ISM band. Path loss measurements have been made in the laboratory using rectaxial antennas at 2.45 GHz. The transmitted signal is an unmodulated carrier and the receiver is a spectrum analyzer. Characterization of signal power variations are important for system design and so path loss and fading models have been derived. The paper presents a comprehensive narrowband channel model (including interference due to neighboring nodes) for application to asynchronous short-range wireless networks. Medium access is assumed to be CSMA based and SIR will therefore depend on inhibition distance and antenna characteristics. Aggregated interference due to neighboring nodes has been calculated. Finally, cumulative SIR values have been used to construct link budgets and bit error rates.

This paper evaluates the introduction of pulse shaping to RF communications between nodes in an ultra-small, wireless sensor network technology. With low power consumption crucial, and in the context of a digital transceiver which utilizes Manchester encoding to reduce the synchronization overhead, we consider the addition of pulse shaping filters to an existing on-off keyed transmission scheme. Pulse shaping is shown to reduce spectral leakage and thereby to allow the transmit power to be reduced. These evaluations are made in the context of 125 mm 3 nodes transmitting over channels of up to 30 cm, with potentially poor clock oscillators and limited energy storage capacity

Speckled computing is a novel vision of a wireless sensor network consisting of small nodes which can sense, compute and network wirelessly. The nodes individually have limited power and processing resources, but together forms a powerful processing system. Electrical power resources at such a volume are severely restricted, and as such design decisions are made with low-power as the first priority. This work examines the use of Manchester encoding in the digital transceiver to reduce the complexity of symbol synchronisation. A Manchester decoder has been implemented which has the useful property of being tolerant to oscillator inaccuracies, allowing a cheap and low-power clock source to be employed. A realistic implementation of the decoder using rectangular pulse-shaping and an oversampling ratio of 8 allows an on-chip oscillator tolerance of more than 11%

In this paper, a distributed mechanism used by a robotic swarm to facilitate the building of a loose wall structure is discussed and results presented. The ability of a robotic swarm to utilise a spatio-temporal varying template in construction tasks is verified. The work is seen as a step towards the development of a generalised methodology for collective construction building tasks and establishes a means for the building of more complex structures.

The cascaded integrator comb filter is commonly used as a first stage decimation filter in communications systems. This paper reviews sharpening techniques for correcting the passband 'droop' present in the frequency response of these filters. A variation on an existing technique that would allow a partially non-recursive (and, therefore, potentially higher speed) implementation is presented for CIC filters with power of two decimation ratios.

IEEE 802.15.4 compliant design forms the basis of ZigBee technology. Intended for low-power, short range sensor networks, these devices are likely to operate in offices and factories: environments with strong multipath channel components. This research introduces a novel adaptive MMSE receiver design for the ISM 2.4 GHz band, comprising a bank of adaptive MMSE correlators. Design concentrates on methods for the efficient training of each correlator and on de-multiplexing the data after the correlation process. Results show a BER performance comparison of the MMSE receiver design with a traditional correlation receiver. Although the adaptive receiver adds complexity to the transceiver, the improved performance could aid coexistence with other technologies or allow a longer range of operation.

A new algorithm is presented that synthesises multiplier blocks with the goal of minimising FPGA hardware cost. Comparisons with existing algorithms are made via implementing synthesised blocks as the multiplication hardware of fully-pipelined, full-parallel transposed form FIR filters. Results establish that the classic optimisation goal of minimising adders does not minimise FPGA hardware. Instead, minimising multiplier block logic depth is shown to be the primary factor for low area FPGA implementation. Filters generated using the new algorithm are also shown to consume less FPGA area than equivalents implemented using the distributed arithmetic technique.

This paper introduces an oversampled subband approach to linearly constrained minimum variance adaptive broadband beamforming. This method is motivated by the considerable reduction in computation over fullband implementation and resulting large computational complexity when fullband beamformers with high spatial and spectral resolution are required. We present the proposed subband adaptive beamformer structure, discuss the advantages and limitations of it, and comment on the correct projection of the constraints in the subband domain. In a simulation, the proposed subband structure is compared to a fullband adaptive beamformer, highlighting the benefit of our method.

For fast and numerically stable algorithms, array processors with floating point multiplication, division, and square rooting are necessary. The authors consider the use of the arithmetic operation of square rooting in the QR algorithm as used in many linear algebraic signal processing algorithms. Rather than reformulating the algorithms to be square root free with the inherent problems of numerical instability, loss of orthogonality, and overflow/underflow, the square root is reconsidered from first principles and arrays are designed that are as fast and have a smaller chip area than the analogous division arrays. This implies that implementations such as square foot free Givens rotations should not be considered in an application-specific integrated circuit or similar design due to their potential instability and susceptibility to overflow. >