Our Research
Alongside the various projects, we are carrying out research in the fields of:
Software Defined Radio, Signal Processing, Cognitive Radio, RF-SoC, Cellular Networks, O-RAN, AI Acceleration, Shared Spectrum and Dynamic Spectrum Access
A list of recent publications is provided below, with links to open access copies of the work.
Resource management in dynamic shared spectrum networks
Ehinomen Atimati, Tawachi Nyasulu, David Crawford, Robert Stewart
IEEE International Symposium on Dynamic Spectrum Access Networks (DySPAN), August 2025
Abstract — With the ever-increasing demand for wireless connectivity for a growing range of applications, there is a compelling need to maximize the available radio spectrum where and when it is needed. In certain situations, the traditional approach of static license allocation leads to spectrum shortage. This paper proposes a tailored approach where base stations/nodes are dynamically assigned spectrum based on their real-time requests/demand needs. It develops a dynamic shared spectrum environment for training its central Deep Reinforcement Learning (DRL) agent, a resource management model (RMM) to assign limited spectrum and transmit power resources to many nodes based on their needs and specifications. The model learned to manage interference among nodes according to IEEE 802.19 coexistence management principles. RMM showed a 20% improvement in request performance compared to the traditional exclusive resource sharing approach. Its assignment performance was better than other methods, such as random, recursive (non-optimization techniques), and a staged Q-learning algorithm (optimization technique), as it consistently provided a sufficient number of requesting nodes with good quality of service (QoS). This improvement diminished with the increased number of bands. Future work will focus on enhancing DRL performance for larger state-action pair sizes.
https://doi.org/10.1109/DySPAN64764.2025.11115965
CP-OFDM PUSCH model-based design for 5G new radio transmitter on ZCU216 RFSoC
James Craig, Louise Crockett, Robert Stewart, Ian Bowyer, Garrey Rice
30th IEEE Symposium on Computers and Communications (ISCC), July 2025
Abstract — 5G New Radio (NR) aims to provide a technological solution to the growing demand for faster data rates and lower latency in mobile communications through its improved features and flexibility; however, this comes at the cost of increased design complexity when targeting hardware devices. To address this, high-level design tools such as Simulink can be used with iterative model-based design to shorten development cycles and reduce human error. This paper demonstrates this design flow through the implementation of a hardware-compatible NR Physical Uplink Shared Channel (PUSCH) model, targeting the Zynq UltraScale+ Radio Frequency System-on-Chip (RFSoC) ZCU216 development board. By interfacing with the hardware through MATLAB, the design supports multiple symbol modulation schemes, and parameters such as the number of layers, antenna ports, and selected precoding matrix can be altered. The design was tested by targeting a 10 MHz bandwidth, 60 kHz subcarrier spacing (SCS) waveform with 132 active subcarriers and met timing with 0.276 ns Worst Negative Slack (WNS) for a 245.76 MHz clock frequency, demonstrating its standard compliancy. The Cyclic Prefix Orthogonal Frequency Division Multiplexing (CP-OFDM) modulated signal generated on the board was looped back to MATLAB and verified against an existing software implementation, and the hardware usage was recorded.
DOI Coming Soon!
Farrow-based true time-delay FPGA wideband digital beamforming architecture
Ryan Provan, Louise Crockett, Robert Stewart, Stephan Weiss
30th IEEE Symposium on Computers and Communications (ISCC), July 2025
Abstract — This paper presents a flexible FPGA digital beamforming architecture to steer an array for wideband radio frequency signals. The architecture combines True Time-Delay (TTD) units and phase-shifting to beamform at digital baseband. The TTDs utilize low order, coefficient-symmetric Farrow structures to rapidly and flexibly adjust fractional sample delays applied to wideband signals with minimal FPGA resources. Bandpass sampling is employed to further reduce resource and power consumption. The designed Farrow structure’s group delay and magnitude characteristics are evaluated. Measured beam patterns are demonstrated through simulation of the proposed FPGA receive array using fixed-point arithmetic. The multiplier utilization of the proposed system is estimated and compared with the literature. Promising results open discussions to hardening Farrow structure cores on FPGAs to serve multiple signal processing techniques required in future communications systems without consuming programmable logic fabric.
DOI Coming Soon!
A single chip Split-6 PHY implementation for 5G NR, using RFSoC and PYNQ
Lewis J. Brown, Louise H. Crockett, Robert W. Stewart
Franklin Open 13, November 2025
Abstract — Indoor traffic has been estimated to account for over 70% of all mobile data in urban areas. We propose a small cell Split-6 radio unit (RU) to meet the indoor use case, presenting a hardware implementation of the physical layer Physical Downlink Shared Channel, targeting an AMD UltraScale+ Radio Frequency System on Chip (RFSoC) device. Several custom IP cores are developed and used in conjunction with integrated resources
available on the target RFSoC chip. The design uses less than 10% of Block RAM memory, and under 3% of other available Programmable Logic resources; meeting timing at 122.88 MHz, enabling 100 MHz bandwidth channels. This modest resource utilisation leaves capacity for integration of other physical layer channels. Total power of 4.33 W and dynamic power of 2.533 W confirm that the baseband RFSoC power draw is very low. A test harness developed using the PYNQ software/hardware framework is presented, which provides a low-complexity approach to radio testing without external network software. The PYNQ system provides a user interface to update resource allocation in real time, and produces resource grid and constellation plots
based on data captured from the design operating on the target RFSoC4x2 development board. The system is tested through two case studies, validating each sub-step of the physical layer. Latencies of 0.332 ms and 0.719 ms, with a 122.88 MHz clock, meet key HARQ requirements. The success of the study confirms RFSoC devices as candidates for small cell Split-6 deployment.
https://doi.org/10.1016/j.fraope.2025.100408
Intelligent joint resource management of shared spectrum HetNets
Ehinomen Atimati, David Crawford, Robert Stewart
IEEE Future Networks World Forum, June 2025
Abstract — Shared Spectrum networks are a means of providing affordable connectivity to digitally excluded regions. Equal priority radios/AP/BS/nodes with similar and dissimilar MAC protocols in shared spectrum homogeneous and heterogeneous networks compete for scarce spectral resources. Coordination of these resources is challenging, especially with heterogeneous networks (HetNets). The task of achieving and maintaining reduced contention among requesting radios is worsened when many dynamically located radios request few available resources. To address this, we adopt an IEEE 802.19 coexistence framework and a Deep Q-Network (DQN) reinforcement learning (RL) algorithm for joint central coordination, thus improving the Quality of Service (QoS) experienced by equal-priority sharers. A central joint optimization algorithm (JOA) was built, trained, and tested. The results show that the novel model’s resource assignment averaged 98%, similar to the exclusive(fixed) resource assignment. The model’s DQN policy outperformed random, recursive, and a two-staged Q-learning allocation, as it consistently assigned resources that enabled requesting nodes’ receivers to provide a good quality of service. The results indicate significant potential for automated spectrum coordination, and efforts are ongoing to enhance the optimization performance of vast state-space DQN.
https://doi.org/10.1109/FNWF63303.2024.11028798
Low latency wireless broadcast production over 5G
Samuel R. Yoffe, Douglas G. Allan, Kenneth W. Barlee, Dani Anderson, Damien Muir, Malcolm R. Brew, Cameron Speirs, Robert W. Stewart, Mark B. Waddell, Jonas Kroger-Mayes, Giulio Stante
NAB Broadcast Engineering and IT Conference, April 2025
Abstract — Wireless camera feeds are an integral source of content for programme making, and are typically done using licensed point-to-point radio links or “bonded-cellular” devices. Cellular bonding using public 4G and 5G networks has become a mainstay for electronic newsgathering and remote contribution feeds. These contributions can tolerate latencies up to several seconds. The use of private5G to support this contribution workflow has been well documented, such as the coronation of King Charles III. However, such latencies are far too long for broadcast production, where wireless cameras are cut in with cabled systems, or where remote camera control, tally and return video are required. Since 5G is a native IP technology, networks are able to support bi-directional connectivity and facilitate these additional services alongside ultra-low latency video. This paper explores the use of (private) 5G to support full low-latency wireless production workflows, along with discussion on device connectivity, augmenting existing wireless systems, and practical advice for configuring camera control systems over 5G.
https://strathprints.strath.ac.uk/92320/
Off-piste 5G in the Broadcast Auxiliary Service Band
Douglas G. Allan, Samuel R. Yoffe, Kenneth W. Barlee, Dani Anderson, Iain C. Chalmers, Malcolm R. Brew, Cameron A. Speirs, Robert W. Stewart, Nicolas Breant, Jeremy Tastet, Sebastien Roques, Bastien Chague
NAB Broadcast Engineering and IT Conference, April 2025
Abstract — 5G New Radio (NR) can be used to provide flexible, high-capacity and low-latency networks suitable for broadcast content acquisition or delivery, but access to suitable spectrum can be challenging. One of the enablers for private network deployments is shared spectrum licensing, such as the upper n77 band (3.8–4.2 GHz) available in the UK and elsewhere in Europe. The Third Generation Partnership Project (3GPP) was created to develop mobile standards for WCDMA and TD-SCDMA and their respective core networks, and has continued to publish standards as radio access technologies have progressed to 4G and 5G. These standards define frequency bands, numerologies, duplex models and messaging (among many other things). While software-defined radio (SDR) is emerging as a viable and highly flexible solution for core and radio access network (RAN) functions, user equipment (UE) typically remain hardware based with modems that implement the 3GPP standards to ensure device compatibility. The flexibility of SDR RAN allows for wireless radio networks based on 5G NR to be built in non-3GPP defined spectrum bands, but there are no compatible devices to connect. In the USA, broadcasters have access to spectrum in the Broadcast Auxiliary Service (BAS) band (2025– 2110 MHz), which coincides with the programme-making and special events (PMSE) band used in the UK and Europe. This allows for rapid licensing of 10/12 MHz channels for traditional wireless camera systems, such as COFDM, that could instead be used to license low-to-medium power private 5G NR-based networks capable of supporting multiple cameras and other IP-based workflows. This paper discusses the development of a flexible software-defined UE capable of connecting to non-3GPP 5G NR networks in BAS/PMSE spectrum.
https://strathprints.strath.ac.uk/92319/1/Allan-etal-NAB-BEITC-2025-Off-piste-5G-in-the-Broadcast-Auxiliary-Service-band.pdf
RFSoC modulation classification with streaming CNN: Data Set Generation & Quantized-Aware Training
Andrew MacLellan, Louise H. Crockett, Robert W. Stewart
IEEE Open Journal of Circuits and Systems Vol 6, Jan 2025
Abstract — This paper introduces a novel FPGA-based Convolutional Neural Network (CNN) architecture for continuous radio data processing, specifically targeting modulation classification on the Zynq UltraScale+ Radio Frequency System on Chip (RFSoC) operating in real-time. Evaluated on AMD’s RFSoC2x2 development board, the design integrates General Matrix Multiplication (GEMM) optimisations and fixed-point arithmetic. We also present a method for creating Deep Learning (DL) data sets for wireless communications, incorporating the RFSoC into the data generation loop. Furthermore, we explore quantised-aware training, producing three modulation classification models with different fixed-point weight precisions (16-bit, 8-bit, and 4-bit). We interface with the implemented hardware through the open-source PYNQ project, which combines Python with programmable logic interaction, enabling real-time modulation prediction via a PYNQ-enabled Jupyter app. The three models, operating at a 128 MHz sampling rate prior to the decimation stage, were evaluated for accuracy and resource consumption. The 16-bit model achieved the highest accuracy with minimal additional resource usage compared to the 8-bit and 4-bit models, making it the optimal choice for deploying a modulation classifier at the receiver.
https://doi.org/10.1109/OJCAS.2024.3509627
Using a private 5G network to support the international broadcast of the coronation of HM King Charles III
Samuel R. Yoffe, Malcolm R. Brew, Douglas G. Allan, Kenneth W. Barlee, Dani Anderson, Odianosen Ighagbon, Damien Muir, Joshua Goldsmith, Cameron Speirs, Robert W. Stewart, Ian Wagdin, Mark B. Waddell, Purminder Gandhu, Andy Reed, Simon Ashton
BEIT Conference Proceedings, April 2024
Abstract — Wireless cameras for news contribution feeds regularly use “bonded-cellular” devices, which connect to and split the encoded video across multiple public mobile network SIMs. However, in high demand density environments with large crowds, the public networks can quickly become saturated and unable to sustain the necessary bitrates to support high-definition video. To overcome this and provide uncontested wireless connectivity, the largest pop-up 5G standalone non-public (private) network of its type was deployed outside Buckingham Palace and along The Mall to Admiralty Arch to support news contributions for domestic and foreign broadcasters at the Coronation of HM King Charles III, without changing the contribution workflow.
https://strathprints.strath.ac.uk/88699/
Intelligent shared spectrum coordination in heterogeneous networks
Ehinomen Atimati, David Crawford, Robert Stewart
IEEE Virtual Conference on Communications, March 2024
Abstract — Global connectivity requires reliable and affordable access to the internet for digital inclusiveness. Shared spectrum technologies are one of many technologies that can help provide affordable connectivity. They require available spectrum (channels) from Primary Users (PUs) and share these among future dynamic heterogeneous secondary user (SU) networks. Coordination of these transient available resources is exacerbated by dynamic SU network scenarios, thus raising the risk of poor SU experience and inefficient spectrum and power usage. Therefore, adopting reinforcement learning (RL), terrain-based propagation models, and IEEE 802.19 coexistence principles, a central intelligent real-time shared spectrum coordination algorithm is proposed to coordinate resource allocation among dynamic SUs operating at low bands. The proposed two-stage algorithm was compared to existing shared spectrum allocation techniques deployed in dynamic spectrum access (DSA) networks to quantify RL’s impact on low-band wireless networks. 66% to 100% of SU nodes/access points (APs) that used the proposed algorithm experienced good quality of service (QoS) in most scenarios examined. A good QoS meant that 75% of APs receivers experienced signal-to-noise plus interference ratio (SINR) greater than 5. This was achieved using minimal AP transmission power.
https://doi.org/10.1109/VCC60689.2023.10474686
Ultra-wideband SDR architecture for AMD RFSoCs and PYNQ based GNU Radio blocks
Marius Siauciulis, Louise H. Crockett, Robert W. Stewart
Proceedings of the GNU Radio Conference Vol 8, September 2023
Abstract — The AMD RFSoC (Radio Frequency System on Chip) architecture has gained significant attention within the Software Defined Radio (SDR) community for its integration of Radio Frequency (RF) frontend, FPGA fabric and Linux-capable Arm-based processing system in a single package. Despite its accessibility to researchers via the RFSoC 2×2 and RFSoC 4×2 development board platforms, its adoption within the GNU Radio community has been limited. This work demonstrates the potential of combining RFSoC with GNU Radio by utilizing a bidirectional QSFP network link to transmit and receive a wideband Orthogonal Frequency-Division Multiplexing (OFDM) signal. Using the remote procedure calls we are able to control the Tx/Rx center frequency and RFSoCs Digital Up/Down Converter (DUC/DDC) rates from the host PC to achieve runtime configurable bandwidth. Additional signal inspection and visualisation is implemented using existing GNU Radio GUI widgets and analysis blocks.
https://pureportal.strath.ac.uk/files/183404896/Siauciulis-etal-GNURC-2023-Ultra-wideband-SDR-architecture.pdf
Streaming Convolutional Neural Network FPGA Architecture for RFSoC Data Converters
Andrew Maclellan, Louise H. Crockett, Robert W. Stewart
21st IEEE Interregional NEWCAS Conference, August 2023
Abstract — This paper presents a novel Convolutional Neural Network (CNN) FPGA architecture designed to perform processing of radio data in a streaming manner without interruption. The proposed architecture is evaluated for radio modulation classification tasks implemented on an AMD RFSoC 2×2 development board and operating in real-time. The proposed architecture leverages optimisation such as the General Matrix-to-Matrix (GEMM) transform, on-chip weights, fixed-point arithmetic, and efficient utilisation of FPGA resources to achieve constant processing of a stream of samples. The performance of the proposed architecture is demonstrated through accuracy results obtained during live modulation classification, while operating at a sampling frequency of 128 MHz before decimation. The proposed architecture demonstrates promising results for real-time, time-critical CNN applications.
https://doi.org/10.1109/NEWCAS57931.2023.10198198
100GBit/s RF sample offload for RFSoC using GNU Radio and PYNQ
Marius Siauciulis, David Northcote, Josh Goldsmith, Louise H. Crockett, Sarunas Kalade
21st IEEE Interregional NEWCAS Conference, August 2023
Abstract — Modern software defined radio systems are capable of multi-gigabit-per-second sampling rates producing unprecedented amounts of digitized RF data. In applications such as wideband spectrum sensing and machine learning algorithms for cognitive radio, prototyping, and instrumentation, it is often impractical to process the acquired data locally in real-time. This motivates the need for a high speed connection to offload data to an accelerator application running on a secondary processing resource. In this paper, we present a novel hardware and software co-design using the AMD RFSoC 4×2 platform, PYNQ and GNU Radio projects. The demonstrated system is capable of continuous 80GBit/s offload in a 100GBit/s channel, utilising a GPU acceleration to rapidly process the Fast Fourier Transforms of a full 2GHz bandwidth RF signal at 60 frames per second.
https://doi.org/10.1109/NEWCAS57931.2023.10198070
An interference management system for a shared spectrum access network
Ehinomen Atimati, David Crawford, Robert Stewart, Ifeyinwa Achumba, Longinus Ezema, Uchenna Diala
IEEE Nigeria 4th International Conference on Disruptive Technologies for Sustainable Development, June 2022
Abstract — Internet access, in developing and underdeveloped countries, remains a huge challenge despite advancements in technology. Shared resources, amongst telecommunication systems, offer an affordability context to this problem. A shared spectrum interference management system is implemented by designing a geolocation database, for a television white space network, for a location in Nigeria. This is achieved using the Dynamic Spectrum Alliance framework (a rarely used methodology) and robust terrain-based propagation models. The designed spectrum coexistence manager (geolocation database) was created, presented, and evaluated, based on its channel availability, predictions, and ability to protect very weak TV signals. The result showed a 15% channel utilization of Analogue and Digital Terrestrial Television channels within the study area. Finally, key components of the framework, that can be adopted for further studies, were identified.
https://doi.org/10.1109/NIGERCON54645.2022.9803165
Comparison of graph-based and hypergraph-based models for wireless network coexistence
Tawachi Nyasulu, David H. Crawford
IEEE International Mediterranean Conference on Communications and Networking, September 2021
Abstract — Dynamic Spectrum Access (DSA) is regarded as a promising solution for efficient spectrum management. Regulators have also approved licence-exemption or general authorisation access (GAA) to further improve spectrum accessibility for DSA systems in the Television (TV) and 3.5 GHz bands. However, heterogeneous DSA radio standards have been developed and the gains in spectrum efficiency could be undermined by coexistence challenges. Hence, the IEEE 802.19.1 standard for wireless network coexistence methods was published, but it leaves algorithmic implementation of the methods to the industry. When the spectrum is not sufficient for exclusive channel allocation, the standard includes a method for co-channel sharing among coexistent neighbour networks. In previous work, channel sharing was introduced on top of the exclusive channel allocation. However, channel sharing options could be significantly limited by the outcome of the exclusive channel allocation. Alternatively, this paper proposes use of hypergraph theory to model the co-sharing strategy for coexistence management of heterogeneous radio systems. Results demonstrate that the hypergraph method achieves higher average spectrum utilisation by up to 8% and requires up to 5 fewer channels to achieve, on average, 100% operational networks than the previous method.
https://doi.org/10.1109/MeditCom49071.2021.9647587
Hypergraph-based model for coexistence management of heterogeneous wireless networks
Tawachi Nyasulu, David H. Crawford
DNAC Wireless Days Conference, June 2021
Abstract — Traditional graph theory is typically used to model interference relations among networks to realise channel assignment that enables improvement in spectrum utilisation through spatial re-use of the channels. Studies have shown that spectrum utilisation could be further improved through co-sharing among networks that are capable of spectral coexistence as long as the channel load is not excessive. The co-sharing networks use their inherent media access control (MAC) techniques to coordinate access to a shared channel. However, the concept of an edge in a traditional graph, which is a two-element subset, is not sufficient to model subsets of potential co-sharing networks because such subsets may have cardinality of greater than 2. Instead, this paper proposes use of hypergraph theory to model the co-sharing strategy in an environment that comprises heterogeneous radio systems. The model could be applied in centralised coexistence management frameworks such as IEEE 802.19.1-based systems. Results demonstrate that spectrum sharing using the hypergraph model achieves higher average spectrum utilisation by up to 17.5% when there are 3 available channels, and requires up to 7 fewer channels to achieve, on average, 100% operational networks than spatial re-use alone.
https://doi.org/10.1109/WD52248.2021.9508293
Rapid Prototyping and Validation of FS-FBMC Dynamic Spectrum Radio with Simulink and ZynqSDR
Kenneth W. Barlee, Robert W. Stewart, Louise H. Crockett, Neil C. MacEwen
IEEE Open Journal of the Communications Society, Nov 2020
Abstract — This paper 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 paper 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.
https://doi.org/10.1109/OJCOMS.2020.3039928
Control and Visualisation of a Software Defined Radio System on the Xilinx RFSoC Platform Using the PYNQ Framework
Josh Goldsmith, Craig Ramsay, David Northcote, Kenneth W. Barlee, Louise H. Crockett, Robert W. Stewart
IEEE Access, July 2020
Abstract — The availability of commercial Radio Frequency System on Chip (RFSoC) devices brings new possibilities for implementing Software Defined Radio (SDR) systems. Such systems are of increasing interest given the pace of innovation in wireless technology, and the pressure on RF spectrum resources, leading to a growing need to access the spectrum in more dynamic and innovative ways. In this paper, we present an SDR demonstration system based on the Xilinx RFSoC platform, which leverages the Python-based `PYNQ’ (Python Productivity for Zynq) software framework. In doing so, we highlight features that can be extremely useful for prototyping radio system design. Notably, our developed system features Python-based control of hardware processing blocks and Radio Frequency (RF) data converters, as well as direct visualisation of communications signals captured within the chip. The system architecture is reviewed, hardware and software components are discussed, functionality is demonstrated, and aspects of the system’s performance are evaluated. Finally, it is noted that this combined RFSoC + PYNQ approach is readily extensible for other SDR systems; we highlight our online shared resources, and invite other engineers to investigate and build upon our work.
https://doi.org/10.1109/ACCESS.2020.3008954
Generic Compression of Off-The-Air Radio Frequency Signals with Grouped-Bin FFT Quantisation
Damien Muir, Louise H. Crockett, Robert W. Stewart
EUSIPCO 2020
Abstract — 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, Groupedbin 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.
https://www.eurasip.org/Proceedings/Eusipco/Eusipco2020/pdfs/0001767.pdf
FPGA Accelerated Deep Learning Radio Modulation Classification Using MATLAB System Objects & PYNQ
Andrew Maclellan, Lewis McLaughlin, Louise Crockett, Robert Stewart
29th International Conference on Field Programmable Logic and Applications (FPL), September 2019
Abstract — 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.
https://doi.org/10.1109/FPL.2019.00045
Secondary user access for IoT applications in the FM radio band using FS-FBMC
Kenneth W. Barlee, Louise H. Crockett, Robert W. Stewart
IEEE First World Form on 5G, 2019
Abstract — In this paper a Dynamic Spectrum Access (DSA) Physical layer (PHY) technique is proposed that allows Secondary User (SU) access to the traditional FM Radio spectrum (88-108 MHz) for alternative data communication applications. FM radio waves have excellent propagation characteristics for long distance transmission, and have high levels of penetration through buildings. Using tools such as a structured geolocation database of licensed Primary User (PU) FM Radio transmitters, unlicensed SUs can access portions of the 20 MHz-wide band and transmit signals that place spectral ‘holes’ with suitable guard bands around all known PUs. Based on the PU protection ratios published by Ofcom and the FCC, the operation of a FBMC (Filter Bank Multi-Carrier) transmitter is demonstrated for an urban environment, and through ‘field test’ simulation it is shown that the Out Of Band (OOB) leakage of the proposed PHY (energy in the ‘holes’ that can interfere with the PU) is 47 dB lower than that of using an equivalent OFDM PHY. The results show that the proposed PHY is a suitable candidate for DSA-SU communication (e.g. in smart city IoT applications), whilst ensuring the integrity of incumbent PU signals.
https://pureportal.strath.ac.uk/files/80843172/Barlee_etal_NEWSDR_2018_Secondary_user_access_for_IoT_applications_in_the_FM_radio_band.pdf
Malawi’s TV white space regulations: a review and comparison with FCC and Ofcom regulations
Tawachi Nyasulu, David Crawford, Chomora Mikeka
IEEE Wireless Communications and Networking Conference, April 2018
Abstract — Regulators are in the process of framing regulations to allow secondary use of vacant TV channels while protecting TV broadcast services from harmful interference. While the US and UK regulators have already passed such regulations in 2008 and 2015 respectively, other countries are still in drafting stages and the underlying circumstances in these countries could be different from those of the US and UK. Malawi released its final draft regulations in 2016. While the US and UK legislate for dynamic spectrum access and licence-exemption for secondary users, Malawi’s draft regulations require such users to apply for a licence for assigned TV white space spectrum. This paper provides an analytical review of Malawi’s regulations and a comparison with FCC and Ofcom regulations, which new regulations can build on. This analysis will also inform future work on network management tools that can enable practical deployment and coexistence of large-scale TV white space networks in a dynamic spectrum access environment in Africa.
https://doi.org/10.1109/WCNC.2018.8377175
A low-cost desktop software defined radio design environment using MATLAB, simulink, and the RTL-SDR
Robert W. Stewart, Louise H. Crockett, Dale Atkinson, Kenneth W. Barlee, David Crawford, Iain Chalmers, Mike McLernon, Ethem Sozer
IEEE Communications Magazine, 2015
Abstract — In the last 5 years, the availability of powerful DSP and Communications design software, and the emergence of relatively affordable devices that receive and digitize RF signals, has brought Software Defined Radio (SDR) to the desktops of many communications engineers. However, the more recent availability of very low cost SDR devices such as the RTL-SDR, costing less than $20, brings SDR to the home desktop of undergraduate and graduate students, as well as both professional engineers and the maker communities. Since the release of the various open source drivers for the RTL-SDR, many in the digital communications community have used this device to scan the RF spectrum and digitise I/Q signals that are being transmitted in the range 25MHz to 1.75GHz. This wide bandwidth enables the sampling of frequency bands containing signals such as FM radio, ISM signals, GSM, 3G and LTE mobile radio, GPS and so on. In this paper we will describe the opportunity and operation of the RTL-SDR, and the development of a hands-on, open-course for SDR. These educational materials can be integrated into core curriculum undergraduate and graduate courses, and will greatly enhance the teaching of DSP and communications theory, principles and applications. The lab and teaching materials have recently been used in Senior (4th year Undergraduate) courses and are available as open course materials for all to access, use and evolve.
https://pureportal.strath.ac.uk/files/45267371/Stewart_etal_IEEE_CM_2015_A_low_cost_desktop_software_defined_radio_design_environment_using_MATLAB.pdf
Software Defined Radio with Zynq Ultrascale+ RFSoC
Louise H. Crockett, David Northcote, Robert Stewart, Douglas Allan, Ehinomen Atimati, Kenny W. Barlee, Lewis J. Brown, James Craig, Graeme Fitzpatrick, Joshua Goldsmith, Andrew Maclellan, Lewis D. McLaughlin, Blair McTaggart, Tawachi Nyasulu, Marius Siauciulis, David Crawford
Published 2023
Abstract — This book introduces Zynq Ultrascale+ RFSoC, a technology that brings real, single-chip Software Defined Radio (SDR) to the marketplace. RFSoC devices are the first adaptive SoCs (Systems-on-Chip) to monolithically integrate multiple RF signal chains along with Arm application and real-time, multicore processors and programmable logic. RFSoC is not so much a radio on a chip, but almost an entire base station on a chip! We anticipate that the book will be of interest and use across a number of technical areas. It serves as an introduction to the family of RFSoC devices and its key features and programmability. The book explores SDR concepts and architecture and key DSP algorithms.
A selection of hands-on exercises via Jupyter Lab notebooks accompany the book and are available from the companion GitHub repository.
https://www.rfsocbook.com/
Exploring Zynq MPSoC
Louise H. Crockett, David Northcote, Craig Ramsay, Fraser Robinson, Robert W. Stewart
Published 2019
Abstract — This book introduces the Zynq® MPSoC (Multi-Processor System-on-Chip), an embedded device from Xilinx® that combines a processing system that includes Arm® Cortex®-A53 application and Arm Cortex-R5 real-time processors, alongside FPGA programmable logic. The book covers the architecture of the device, the design tools and methods, conventional hardware/software co-design approach, and the newer software-defined methodology, as well as hardware and software development, multiprocessing, safety, security and platform management, system booting, and special features on PYNQ (a Python-based framework) and machine learning applications. This book should serve as a useful guide for those getting starting with, and the working with Zynq MPSoC, and equally as a reference for technical managers wishing to gain familiarity with the device and its associated design methodologies.
https://www.zynq-mpsoc-book.com
Software Defined Radio using MATLAB & Simulink and the RTL-SDR (+ Tutorials)
Robert W. Stewart, Kenneth W. Barlee, Dale S. W. Atkinson, Louise H. Crockett
Published 2015
Abstract — The availability of the RTL-SDR for less than $20 brings SDR to the home and work desktops of EE students, professional engineers and the maker community. The RTL-SDR device can be used to acquire and sample RF (radio frequency) signals transmitted in the frequency range 25MHz to 1.75GHz, and using some official software add-ons, these samples can be brought into the MATLAB and Simulink environment for users to develop receivers using first principles DSP algorithms. Signals that the RTL-SDR hardware can receive include: FM radio, UHF band signals, ISM signals, GSM, 3G and LTE mobile radio, GPS and satellite signals, and any that the reader can (legally) transmit of course! In this free book we introduce readers to SDR methods by viewing and analysing downconverted RF signals in the time and frequency domains, and then provide extensive DSP enabled SDR design exercises which the reader can learn from. The hands-on examples begin with simple AM and FM receivers, and move on to the more challenging aspects of PHY layer DSP, where receive filter chains, real-time channelisers, and advanced concepts such as carrier synchronisers, digital PLL designs and QPSK timing and phase synchronisers are implemented. Towards the end of the book, we demonstrate how the RTL-SDR can be used with SDR transmitters to develop a more complete communications system, capable of transmitting text strings and images across the desktop.
https://www.desktopSDR.com/download
The Zynq Book (+ Tutorials)
Louise H. Crockett, Ross Elliot, Martin Enderwitz, Robert W. Stewart (+David Northcote)
Book published 2013, Tutorials published 2015
Abstract — The Zynq Book is all about the Xilinx Zynq®-7000 All Programmable System on Chip (SoC) from Xilinx. This is the online home of The Zynq Book, designed to raise awareness of the book and host the accompanying tutorials. The Zynq Book is the first book about Zynq to be written in the English language. It has been produced by a team of authors from the University of Strathclyde, Glasgow, UK, with the support of Xilinx. We wanted to create an accessible, readable book that would benefit people just starting out with Zynq, and engineers already working with Zynq.
http://www.zynqbook.com/download-book.php
http://www.zynqbook.com/download-tuts.html