WInnComm 2017 Technical Session Paper and Presentation Abstracts

Click to jump to:  Wednesday  ~  Thursday
~~~~~~~

Wednesday, 15 November

10:30 - 12:00

14:00 - 15:30

TS3: Heterogeneous Systems Programming 1
TS4: Cognitive Radio 1

 16:00 - 17:30

TS5: Heterogeneous Systems Programming 2
TS6: Cognitive Radio 2


[TOP]

10:30 - 12:00

TS1:Antenna, RF, IF, and ADC Technologies

RF2D Converter Based Software Defined Radio

David Nienaber (Broadband T Corporation, USA)

A receiver signal processing methodology and architecture is presented, which achieves conversion of an entire RF band directly to the digital signal domain, converting a digitally selectable RF bandwidth of signals centered at a digitally selectable center frequency. Conversion is realized for the raw input signal band, prior to execution of frequency translation, filtering, or mixing. The converted band is thereby made available within DSP blocks for Post Conversion Mixing, further down-conversion, mixing, and/or channel selective or rejective DSP based filtering. This conversion technique is uniquely flexible, format and protocol ambivalent, equally effective for all forms of modulation, and uniformly applicable for any frequency within the wide-band response capability of whichever fabrication process node it is implemented into. Theoretical performance characteristics and limits are mathematically developed. Comparative analysis for this technique relative to other and more traditional receiver techniques is explored, with consideration given to this technique as the missing link for Software Defined Radio.

 

 

Dual Band Antenna for 5G Communication Systems

Daniel S Iancu (Optimum Semiconductor Technologies & Tampere University of Technology, Tampere, Finland, USA)

This presentation describes a vertically stacked dual band antenna composed of a low frequency band at 2.45 GHz and a high frequency band at 20 GHz. The low frequency antenna is a single micro strip patch while the high frequency antenna is a multi-patch micro strip array with beam forming/direction of arrival capability. The beam forming/direction of arrival functionality is achieved by voltage controlled phase shifters. Each micro strip patch is connected to a DC voltage controlled phase shifter. Each phase shifters is digitally controlled through a high speed digital to analog converter. The high frequencies antenna array is on top while the low frequency patch antenna is under the array antenna, acting as ground layer. The multi layer configuration, housing two different frequency band antennas is unique as far as we know. It saves significant real estate in the receiver/transmitter enclosure without compromising significantly the performance of each antenna as an independent structure. The concept is proved both by computer simulations and measurements on fabricated prototypes. The prototype antenna has the following physical dimensions: 3.5 mm x 7 mm for each of the high frequency patch (at 20 GHz) and 27 mm x 40 mm for the low frequency patch.

 

Trade-Off Analysis for Complexity versus Cost in Map-Reduce Based Hybrid Beamforming

Ture Peken, Tamal Bose and Ravi Tandon (University of Arizona, USA)

High data rates ups to 10 Gbps can be achieved for next generation wireless communication by using the millimeter wave (mmWs) bands. Hybrid beamforming, which combines analog beamformers in the RF domain and digital beamformers in the baseband domain, allows the reduction of RF chains while achieving high performance gains in mmWs. Therefore, wireless systems operating at mmWs are expected to use hybrid beamforming. Analog and digital beamformers, which achieve the maximum mutual information over the channel, should be designed in hybrid beamforming. The computational complexity of finding optimal beamformers is significantly high since it grows exponentially with the number of subarrays at the transmitter and the receiver. The MapReduce framework gives promising results by distributing the computation among the multiple cores. The optimum number of cores to use in the MapReduce based hybrid beamforming has not been studied to the best of our knowledge. In this paper, we give a theoretical analysis to find the optimum number of cores in terms of the speed-up and the cost.

[TOP]

TS2: Spectrum Sharing Platforms

A spectrum broker service for CBRS Priority Access Licenses on Illiquid, non-realtime, secondary markets

Heikki Kokkinen (Fairspectrum, Finland); Seppo Yrjölä (Nokia, Finland); Topias Kokkinen (Fairspectrum, Finland)

This presentation discusses a spectrum broker service for micro-operator and Citizens Broadband Radio Service (CBRS) Priority Access Licenses (PAL). The FCC licenses for the PAL layer users will be assigned via competitive bidding and they are allowed to operate up to a total of 70 MHz of the 3550-3650 MHz spectrum segment enjoying interference protection from the GAA operations. A PAL non-renewable authorization is for a 10 MHz channel in a single census tract for three years, with the ability to aggregate up to six years up-front. The PAL layer may cover critical access users like utilities, Internet of Things (IoT) verticals, governmental users, and non-critical users, e.g., MNOs and WBS after a final five-year term on the 3650-3700 MHz band. PALs are auctioned to the licensee within their service area on a census tract basis but the specific channels are assigned, re-assigned and terminated at the end of the term by the Spectrum Access System (SAS). The PAL will be opened for the third GAA tier users when unused and further automatically terminated and may not be renewed at the end of its term. The 'use' status of PALs in the CBRS 'use it or share it' approach is determined using two engineering approaches. First, a PAL licensee should report their PAL protection areas (PPAs) on the basis of actual network deployments, and second, to maximize an objective protection area, the SAS must not authorize other CBSDs on the same channel in geographic areas and at maximum power levels that will cause aggregate interference in excess of -80 dBm/10 MHz channel within a PPA. The FCC revisited rules for CBRS in 2016, and introduced the light-touch leasing process to enable secondary markets for the spectrum use rights held by PAL licensees. Under the framework, no FCC oversight is required for partitioning and disaggregation, and PAL licensees are free to lease any portion of their spectrum or license outside of their PPA. The PPA can be self-reported by the PAL owner or calculated by the SAS. The PAL Radio Frequency channel can be re-allocated beyond the PPA, but within the census tract. Introduced low additional administrative burden with a minimum availability of 80 MHz GAA spectrum in each license area will provide the increased flexibility to serve targeted services to geographic areas or quantities of spectrum. Furthermore, the FCC will permit stand-alone or an SAS-managed spectrum exchange and let market forces determine the role of the SAS value added services. Traditional spectrum allocation for mobile networks is mainly done in the primary market where the government authorities sell long term licenses by auctions. While these auctions have many benefits, and are accepted as the standard method, they still lead to inefficient situations in particular circumstances. The demand for the spectrum can change rapidly and drastically due to factors such as changes in traffic demand, spectrum applications, and technologies. However, the static, long term licenses do not adapt to these changes. This leads to situations where the licenses are not held by the parties that value them the most. Another problem is that the packaging of licenses to large blocks leads to oligopolies where there are only few large license holders. The winning bidder might not need all bundled licenses so parts of the spectrum remain unused. Additionally, restricted competition and static licenses hinder new innovation. A solution to this problem is to establish a secondary market for the licenses. An open access market for spectrum would increase competition and make the process more efficient, transparent, fair, and simple. There are many proposals of real-time secondary marketplaces where capacity is auctioned according to current demand. These marketplaces are mainly designed for liquid licenses. However, this presentation studies the research question how to facilitate the exchange of spectrum resources that are used for applications such as micro-operator licenses or Priority Access Licenses (PAL) in Citizens Broadband Radio System (CBRS). These licenses are often illiquid micro licenses. Thus, we introduce a non-real-time marketplace for buying and leasing both exclusive and shared access to spectrum. The main function of the marketplace is to allow fast, convenient, and low-cost exchange of local licenses. If there is a high demand for a particular micro license, sellers can use auction instead of a buy now price. Auctions can be used to find the equilibrium price but they do not work as efficiently if there are only few buyers. It is reasonable to assume that the number of buyers is relatively small in micro licensing cases, because the licenses are local and they benefit only few buyers. This presentation examines the pricing of illiquid micro licenses. The price of liquid licenses can be determined for example by auctions or by comparing the sales prices of similar licenses. However, when there are not enough buyers or sellers, market based methods are not reliable. The valuation can also be done by evaluating factors such as the potential economic benefit of the license and the opportunity costs of alternative options. However, this is a labor-intensive method and might not be economically worthwhile when considering the size of the micro licenses. This presentation introduces an automatic valuation method for these small, illiquid licenses. We list a number of features for the proposed system. To increase efficiency, the marketplace automates labour-intensive processes by filing regulatory documents and checking compliance with the law. Additional services, such as analytics tools and consulting could provide additional value for both the buyer and the seller. The presentation shows how the marketplace could be implemented by providing examples of the interfaces for both the buyer and the seller. The revenue section shows that there are different business models that could be used to generate revenue without raising the prices too high for the buyers or the sellers.

 

Radio platforms for spectrum monitoring and mmWave communications

François Lefebvre (Communications Research Centre Canada, Canada)

The Communications Research Centre (CRC) is the Canadian federal government's centre of excellence with the mandate to perform wireless telecommunications R&D that advances the efficient exploitation of the radio spectrum, and to serve as the government's leading source of scientific knowledge and long-term technical advice on spectrum management, regulation and policy. This presentation will provide an overview of two of CRC's main research programs and will describe the radio platforms, frameworks and tools that were developed, integrated and deployed as part of those programs. Under its Spectrum Environmental Awareness program, CRC has developed a cloud-based spectrum monitoring system that relies on a wide range of spectrum sensors consisting of off-the-shelf radios, software defined radios, integrated vehicular radios and high-end radio frequency instruments. These sensors provide a mix of measurement capabilities: wideband spectrum scanning to analyze such parameters as channel occupancy and signal-to-noise ratio, to the analysis of specific wireless communications waveforms such as Wi-Fi and LTE. As these measurements are collected, the sensors upload the data to the cloud for later analysis and visualization by CRC's new Big Data Analytics Centre. Under the Breaking the Frequency Barrier program, CRC is researching materials and surfaces to engineer the radio spectrum environment at mmWave frequencies. CRC is showing how these engineered surfaces can be used to improve mmWave propagation conditions and to optimize network deployments. To demonstrate these materials and to show outdoor mobile communications at mmWave frequencies, two types of radios were deployed to conduct large-scale outdoor experiments at City Hall in downtown Ottawa during the summer 2017. The first type is based on off-the-shelf 28GHz point-to-point radio heads. The other type was integrated at the CRC to allow for the transmission and reception of LTE signals at mmWave frequencies. This platform uses 28GHz up/down radio converters to convert LTE signals generated by USRPs running srsLTE, an open source LTE library. These radios were integrated into CRC's mmWave base stations and user equipment prototypes for testing and demonstrations of CRC's engineered surfaces and pedestrian-speed mobility.

[TOP]

14:00 - 15:30

TS3: Heterogeneous Systems Programming I

Programming Heterogeneous Systems using HSA

John Glossner (Optimum Semiconductor Technologies, USA)

Heterogeneous System Architecture (HSA) Foundation is a not-for-profit industry standards body focused on making it dramatically easier to program heterogeneous computing devices. The consortium comprises various semiconductor companies, tools providers, software vendors, IP providers, and academic institutions that develops royalty-free standards and open-source software. Recently, production systems have been deployed using HSA technology. In this presentation an overview of how the systems are programmed will be presented along with a real-time demonstration of running an HSA program on a laptop computer.

 

MDD software tools for SCA4 developers

Steve Jennis (PrismTech, USA)

Software productivity tools for SCA developers are vital to: lower time-to-value, reduce complexity/bugs, minimize maintenance/porting costs and ease SCA compliance. PrismTech has been a leader in these tools for over a decade and its Spectra CX tooling has provided significant technical and commercial benefits for scores developers of SCA2 radio platforms and/or applications (e.g. waveforms). The emergence and gradual adoption of SCA4 requires even more sophisticated tooling to support development in compliance with the new (more flexible) standard. This paper/presentation will cover the following topics with regard to software tools for SCA development: • The proven benefits of model-driven development to design complex SDR platforms and waveforms • An overview of the generated XML and also the optional C++ code that can be produced with SCA4 development tools • The SCA4 features which are now supported in Spectra CX4 • The key differences between developer tooling for SCA2 and SCA4 environments • The technical challenges encountered and overcome in evolving/adapting development tools for SCA2 to SCA4 compliance

 

 [TOP]

 

TS4: Cognitive Radio 1

The Impact of Delay on the Decisions of a Cognitive Radio Engine

Hamed Asadi (University of Arizona, USA); Haris Volos (DENSO International America, USA); Michael Marefat and Tamal Bose (University of Arizona, USA)

In this paper, we initiate the study of a cognitive radio engine's decision-making in scenarios where the result/feedback of a decision has a non-zero delay. Although, this delay has been mostly absent in the literature involving the link adaptation problem. In certain applications, such as long-distance shortwave (high frequency) radio communication, or deep-space communications which requires communicating with space exploration equipment throughout the solar system, the roundtrip delay can be from seconds to minutes. In these scenarios, the cognitive engine (CE) is faced with the task of making decisions that it will not know of their outcome after a considerable delay. In this paper, we provide the evaluation about the relation of the delay and the CE's performance to find the relation between delays and regret bounds. We also propose a system model, and we evaluate various decision strategies taking into account the expected channel states during the transmission period. Our results show that the expected state and variance of the channel conditions have a significant impact on the CE's decision-making process and performance of the system.

 

Stochastic Models for Optimization of Software-Defined Radio Operation

Marilyn Wolf (Georgia Institute of Technology); Shuvra Bhattacharyya (University of Maryland, USA)

We propose stochastic models for SDR systems based on Markov decision processes (MDPs) which allow us to optimize systems at both design and run time. Compositional models control model complexity and provide structure for solution. Our results show that this approach provides a larger design space that can be used to improve the system.

 

Multi-hop performance assessment of incoherent autonomous cooperative networking in the presence of spatial inhomogeneity

Cenk Köse (Trellisware Technologies, Inc., USA); Andreas Polydoros (University of Athens, Greece)

Cooperative distributed networking is important in mobile wireless applications where significant sources of vulnerability (shadowing/blockage, interference/jamming, dynamic fading) may exist. Multi-hop military networks (MANET) and mobile wireless sensor networks with delay-sensitive applications are two such examples. It is therefore important to assess network performance under a variety of such conditions, in order to arrive at robust solutions which utilize the network resources the best possible way. The assessment of cooperation gains has been addressed extensively in the context of cognitive radio, where various distributed radios (sensors) receive energy, so that collaborative sensing (primary source detection and localization) is the main focus. The current scenario is the dual one: a collection of radios (relays) transmit towards a given receiver (destination or relay) or towards a collection thereof. In both cases, it is important to understand the spatial distribution ("field") of the resulting arriving power in order to translate it to meaningful networking metrics (SNR, outage, capacity/throughput) and accordingly optimize signal design and related protocols (e.g., routing). Recent research work [1] has addressed the problem of determining performance metrics associated with different one-hop cooperative transmission schemes, including incoherent co-transmission, orthogonal transmission and beamforming. This was done under a stochastic-geometry, Poisson Point Process (PPP) model, which facilitates closed-form analysis in the asymptotic SNR regime, emphasizing the role of multipath in providing diversity gains. The present paper differs from previous research on cooperative relayed communication in the following aspects: (i) it contributes to the computation (via simulation) of achievable information rates in a multi-hop scenario while explicitly accounting for the strongly-deleterious impact of correlated shadowing (due to large obstacles) and/or interference sources (such as jammers), as well as small- scale fading for both wideband and narrowband channelization. (ii) it generalizes to more practical node configurations, beyond PPP or grid-based node placement, such as random placement within the "remote-disk" or the "uniform-annulus" topologies. (iii) it compares the efficiency of Incoherent Autonomous Cooperative Networking (I-ACN) versus traditional route-based counterparts, which require rapid assessment of the ambient propagation plus interference fields across the network. (iv) it documents that, in challenging propagation environments, the low-latency data dissemination advantage of incoherent I-ACN does not come at the expense of reliable end-to-end throughput. [1] S. Haykin, "Cognitive radio: Brain-empowered wireless communications," IEEE Journal of Selected Areas Commun., Feb. 2005. [2] A. Polydoros, S. Evangelatos and A. Moustakas, "The Price of Incoherence on Co-Transmission Under a Stochastic-Geometry Model," IEEE Trans. Wireless Communications, April 2017.

[TOP]

16:00 - 17:30

TS5: Heterogeneous Systems Programming II

The Enduring Myths of the Software Communication Architecture (SCA)

Steve Bernier (NordiaSoft, Canada)

As with any new technology that makes bold promises, the adoption of Software Defined Radio (SDR) and the Software Communication Architecture (SCA) went through multiple phases that are best described by the Gartner Hype Cycle chart. The hype behind SDR technology and the early success stories of the SCA triggered significant publicity. The SCA was propelled by hype to a tipping point that Gartner refers to as the "Peaks of Inflated Expectations". The interest in the SCA started to wane as experiments and implementations failed to meet all the expectations. The bad press that followed dragged the SCA in a downward spiral of negative hype that is recognized as the "Trough of Disillusionment" phase. The SCA only started to pull out from the spiral after 2010 as second-generation SCA-based products were brought to market. With third-generation products and the early international adopters, the SCA eventually started its recovery and reached the "Slope of Enlightenment" phase. Nevertheless, the bad press that came with the disillusionment gave birth to many enduring myths of the SCA. This paper identifies the most damageable myths and explains why they can be put to rest.

 

SCA as infrastructure code: A seamless migration from SCAv2.2.2 to SCAv4.1

Juan Pablo Zamora Zapata (NordiaSoft, Canada)

Software Defined systems are the industry's response to the ever increasing complexity and flexibility required in today's military systems. The Software Communications Architecture enables the fulfillment of such requirements for sophisticated heterogeneous embedded distributed systems. In this presentation, NordiaSoft illustrates how to seamlessly convert from SCAv2.2.2 to SCAv4.1. Using Zero Merge code generation technology, the business source code is kept separate from infrastructure source code. The source code for the waveform signal processing or the radar tracking algorithms is kept separate from the source code that deals with deployment, instantiation, configuration, inter-process communications, and the like. The presentation also highlights the track record of NordiaSoft's team from it's implementation of SCAv0.3 to its leadership role in the WInnF's workgroup for SCAv4.1 backwards compatibility. The presentation will also introduce NordiaSoft's signal processing probes that can display signals being processed in real time by a radio.

 

Implementation of Improved omniORB for SCA 4.1 with VxWorks

Shan Wang (National University of Defense Technology); Fanglin Gu, Li Zhou and Wei Ji-bo (National University of Defense Technology, P.R. China)

Although there is no clear restrict for the middleware used by SCA 4.1, it still is a high priority fact because its performance influences the scalability of the SCA framework. OmniORB is an efficient and concise middleware, but such a pity that its support for VxWorks is almost vacant. Firstly in our work, omniORB is improved and implemented in an SCA 4.1 framework with VxWorks based on PowerPC processors. Also, we evaluate the implementation and make a complete comparison between different middlewares in the same operational environment. The results testify the correctness of our proposal. The work provides a more efficient implementation for SCA based platforms.

[TOP]

TS6: Cognitive Radio II

VITA 49 Radio Transport: The New Software Radio Protocol

Rodger Hosking (Pentek, USA)

Officially designated by the VITA Standards Organization as VITA 49 Radio Transport Protocol, this evolving specification defines consistent methods of delivering digitized radio signals, control, status, and metadata between elements of software radio systems. The presentation shows how legacy "stovepipe" architectures consisting of specialized receiver hardware and cabling can be replaced with flexible, network-based VITA 49 links. Nicknamed VRT, this new standard offers vendors and users many benefits including reduced costs, improved inter-operability, maintainability, and easier insertion of new technology.

[TOP]

 


Thursday, 16 November

10:30 - 12:00
 
14:00 - 15:30
 
 16:00 - 17:30

10:30 - 12:00
TS7: SDR and Baseband Technologies

GWN: a Framework for Packet Radio and Medium Access Control in GNU Radio

Victor Gonzalez Barbone (Universidad de la República, Uruguay); Pablo Belzarena (Universidad de la Republica, Uruguay); Federico Larroca and Mariana Gelós (Universidad de la República, Uruguay); Martín Randall (Instituto de Ingeniería Eléctrica -- Facultad de Ingeniería -- Universidad de la República, Uruguay); Paola Romero (Universidad de la República, Uruguay)

Software Defined Radio, and GNU Radio in particular, were conceived for communication systems such as radio and TV, where information is conveyed in a continuous flow, called a stream. Data networks by contrast use small portions of information, called messages, frames or packets according to the context. Another important difference is that in data networks many actors share the same communications channel, called a shared medium. Some discipline must be imposed to avoid "all speaking at the same time". This discipline is called the medium access control or channel access method. GNU Radio was originally stream oriented, but more recently added support for message communications. A block may thus comprise two different types of inputs and outputs: stream ports for continuous flows of data, and message ports for discrete portions of bytes. Some projects exist to implement data network standards in GNU Radio, but they are oriented towards partial implementations of specific communication protocols. In this paper we present GWN (GNU Radio Wireless Network), an open and free extension to GNU Radio specifically oriented to data networks, but not tied to any specific protocol (please visit https://github.com/vagonbar/gr-gwn/wiki). Its aim is to provide a framework for experimentation and development, working either on existing protocols or devising entirely new ones. To this purpose, GWN provides a new generic block (gwnblock) which adds the tools necessary for data network designs, and at the same time decouples all GWN data network blocks from the GNU Radio generic basic_block. This means a new GWN block only needs to inherit from gwnblock and follow GWN design rules, shielding from users most of the complexity of GNU Radio. The GWN generic block adds the following facilities to GNU Radio: - Message orientation. GNU Radio is stream oriented, GWN is message oriented; items interchanged among GWN blocks are discrete groups of bytes. GWN makes use of the message mechanism of GNU Radio, but provides some blocks to interact with stream GNU Radio blocks when necessary, thus relieving users of stream oriented worries. - Events. GWN elaborates on the message interchange mechanism of GNU Radio into a more structured item of interchange called an event. GWN blocks interchange events. The event inner structure reflects the needs of network data protocols and is closer to their design conception. - Handling of time. This is a feature absent in GNU Radio, and essential in data networking. Answers are waited for a certain time, keep-alive signals are emitted at regular intervals; timing pervades data communications. GWN provides two forms of handling time: timeouts and timers. A timeout just waits for some time and emits a timeout event; it is a one-shot gun. A timer emits timing events regularly. - Finite State Machines. Most data communication protocols involve a complex logic usually described in a mathematical model of computation called a Finite State Machine (FSM). An FSM comprises "states" and "transitions", and reacts to "events": when the machine is in a certain state and receives an event, a transition to another state is performed, optionally with some parallel task. FSMs are a very powerful tool, and the complexity of some protocols makes it almost impossible to implement them otherwise. GWN includes a simplified version of an eXtended Finite State Machine (XFSM) which has been used to implement complex packet processing tasks inside network switches, and is considered a powerful enough tool to implement any protocol for data networks. As a proof of concept of GWN, and to illustrate its usage, we briefly present two examples: an ARQ (Automatic Repeat Query) protocol with its different flavors and a CSMA/CA protocol.

 

Rapid Software Defined Radio Waveform Development of DVB-S Transceiver

Ashwin Amanna (ANDRO Computational Solutions, LLC, USA)

Most waveforms operating on Software Defined Radios (SDR) are actually 'firmware' defined implementations with significant elements operating in the FPGA on closed platforms. This diminishes the full potential of SDR in terms of rapid development time cycles, accessibility to waveform code, and adaptable operations for cognitive radio. We address the challenges of rapid waveform development on SDR using the DVB-S as a case study with a true software defined implementation where all I/Q processing is performed on an Intel processors in conjunction with low-cost Ettus B205mini and bladeRF SDRs. We distill the waveform to core functional components and sequence the implementation into sprints while maintaining end-to-end functionality at each iteration. Innovations include strategic use of machine code for computational intensive operations and efficient multi-thread management. Our approach yielded a full transceiver implementation within 2 man-weeks using only the published specification for reference. The transmitter was tested for interoperability with a consumer satellite receiver. CPU usage on an I7 processor was approximately 22%, with a memory usage of 16MB out of the 8GB available. The mean latency of the system is around 50 milliseconds. A similar test showed that the system latency is effected by symbol rate and decreases as the rate is increased. With the symbol rate set to 15 M the latency dropped to 17 milliseconds.

 

Improved Physical Layer Implementation of a MIL-STD-188 CPM Modem

Frederic j harris (San Diego State Univ, USA); Richard Bell (SPAWAR, USA)

CPM signals are phase modulated sinewaves that maintain a constant amplitude. The constant envelope permits use of very efficient power amplifiers operating near their saturation boundaries. One early model of the CPM modulation process is Offset-Quadrature Phase Shift Keying (O-QPSK) with half sinewave shaping filters. Modern implementations entail direct phase modulation of Direct Digital Synthesizers. The direct phase modulation is understood to be a non-linear modulation process [1]. In spite of this fact most receiver structures demodulate the CPM signal by treating it as an O-QPSK signal. For small phase modulation indices, the errors are small but are not zero. We examine and present CPM receiver structures that are not related to O-QPSK but rather to time reversed and conjugate phase profiles of the multiple phase options of the modulation processes. We also examine innovative processing techniques that extract Doppler offsets, and perform carrier phase recovery and symbol timing recovery from traditional structured preambles.

 

TS8: Quality, Reliability and Security

LSA evolution enables local high-quality wireless networks

Seppo Yrjölä (Nokia, Finland)

The Radio Spectrum Policy Group (RSPG) of the European Commission (EC) have identified 700 MHz, 3.4-3.8 GHz and 26 GHz spectrum bands as pioneer bands for 5G in Europe, recommends the band 3.4-3.8 GHz as the primary band for introduction of services in its strategic roadmap [1], and calls for industrial user experiments for the digitization of industry in its 5G Action Plan [2]. Furthermore, Ofcom's statement defines the same spectrum bands for the first wave of 5G in the UK [3]. Groupe Speciale Mobile Association (GSMA) recommends at least one frequency band allocated to 5G from each of the following frequency ranges: sub GHz, 1-6 GHz, and above 6 GHz [4]. The Global mobile Supplier Association (GSA) recommends 3.3-4.2 GHz frequency range, the 3GPP is working on in 5G New Radio (5G-NR) channel arrangement [5], as the primary band in the spectrum below 6 GHz, for the global introduction of 5G [6]. There is, on the other hand, a great variation of the current 3.4-4.2 GHz spectrum use and authorization in the EU member states as well as globally. The incumbents include, e.g., Fixed Wireless Access (FWA), satellite communications, and fixed links, with highly varying expire dates of their radio licenses. Moreover, some of the member states plan to clear and auction at least parts of the band with nationwide licenses, while others have already prepared to have regional licenses on the band. There are member states, which think of having primary and secondary 5G allocations with over 10 year and less than 2 year license periods, respectively. From the above examples, it is obvious that Europe has to prepare for a diverse 5G spectrum use on the primary 3.4-3.8 GHz band. The key objective of the European Telecoms framework [7] is to provide a pro-investment framework to support 5G development through new bands, new users and usages, and increased more flexible use of spectrum. Proposed European Electronic Communications Code (EECC) framework promotes shared use of the spectrum. The prominent spectrum sharing concepts, the LSA in Europe and the CBRS in the US enable a flexible adaptation to both temporal and regional changes in spectrum use while guaranteeing interference free radio communication for incumbent users. This paper discusses the regulatory and standardization status of the LSA, reviews results from the ongoing feasibility study in the European Telecommunications Standards Institute Reconfigurable Radio Systems technical Committee (ETSI RRS) on temporary spectrum access for local high-quality wireless networks [8], and presents the early results of the world first LSA evolution field trial. In this trial, we responded to the discussed early 5G deployment requests by demonstrating how LSA evolution can move towards more dynamic and flexible spectrum management concept. LSAevo concept and system architecture can be applied to 3.4-3.8 GHz band so that current individual fragmentation challenges to take the band into 5G use in the respective member states can be solved, while ensuring that the communication of the incumbent users, FWA, fixed links, and satellite earth stations do not experience any harmful interference. In the e2e field trial, an industrial automation micro operator use case on 3.4-3.8 GHz band was validated. In this use case, the objects, special needs (low latency), and other solutions are local, e.g., industrial machinery never leaves the site or they need special connectivity only when in the local area, like an airplane when landed. This leads to a different type of network, opportunities, and requirements. Private micro-operator networks offer relatively speaking unlimited capacity and speed by tapping into large pools of local spectrum. Network architecture is built around distributed and edge clouds offering low latency and local content management to boost use case development with the domain specific ecosystem. Validation platform utilizes open APIs, native could architecture and leverages the sharing economy principles to create a sustainable business models across stakeholders and interfaces. The demonstrated Network as a Service (NaaS) deployment consists of commercial Long Term Evolution (LTE) User Equipment (UEs), 3.5 GHz eNodeBs under LSAevo control, and virtualized hosted Evolved Packet Core (EPC). The implemented spectrum manager demo system runs on commercially available virtualized Network Management System (NMS) and Self Organizing Network (SON) platforms, and is built on synergies between the existing LSA and CBRS standards. The LTE test network is installed in the Nokia factory in Oulu, Finland. The incumbents were created for the demonstration purpose, but their types and protection criteria are taken from the real life in Finland and the EU member states [9]. Proposed LSAevo builds on proven LSA benefits of leveraging scale and harmonization in regulation & standardization, and utilization of existing commercial assets and capabilities. Introduced new features includes, new frequency bands towards 5G, localization of spectrum with novel 5G use cases, e.g., for verticals, horizontal sharing & sub-licensing for efficient use of the spectrum assets, and as a recapitulation lowers entry barrier for new service providers through unbundling investments in spectrum, infrastructure and services. References [1] RSPG16-032. Final Radio Spectrum Policy Group strategic roadmap towards 5G for Europe. Opinion on spectrum related aspects for next-generation wireless systems. Brussels, Nov. 2016. [2] EC COM(2016) 588 final. Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions. 5G for Europe: An Action Plan. Brussels, Sept. 2016. [3] Ofcom Statement. Update on 5G spectrum in the UK. Feb. 2017. [4] GSMA Spectrum. 5G Spectrum Public Policy Position. Nov. 2016. [5] 3GPP RP-170855, "Work Item on New Radio (NR) Access Technology", 2017. [6] GSA, "The Future of IMT in the 3300-4200 MHz frequency range," June 2017. [7] European Commission, Proposal for a Directive of the European Parliament and of the Council establishing the European Electronic Communications Code COM(2016) 590, Articles 2 (26), 45 (2) (e), 46 and 47, 2016.

 

Improving PHY-Layer Security with Traffic Formation

Garrett Vanhoy and Tamal Bose (University of Arizona, USA)

The exploitation of side-channel information (SCI) poses a threat to the security of even the most sophisticated systems. SCI generally refers to any information that is exposed from a system employing encryption other than the original or encrypted data. In wireless systems, this information can include signal attributes such as received signal strength, bandwidth, burst duration, modulation, and others. Although encryption prevents an eavesdropper from being able to completely understand traffic being generated between devices, SCI can be exploited to potentially circumvent encryption. Traffic patterns are an especially revealing form of SCI. For example, it has been shown that particular traffic patterns can be used to identify a web page that a user is currently browsing. Many existing techniques used to exploit or extract SCI require knowledge of the protocol being used between devices or being able to extract commonly unencrypted information from protocol headers. In this paper, we discuss how physical layer parameters, which are impossible to hide, can be exploited to do traffic classification on an 802.11 system. We also propose potential mitigation techniques.

 

Bringing new levels of reliability to deployed Software Defined Radios

Tim Fountain (National Instruments & National Instruments, USA)

In this paper we will discuss new tools that improve the reliability and maintainability of remotely deployed Software Defined Radios. As SDR's are increasingly being used in remotely deployed applications, a critical gap has emerged in how a remote operator of the SDR can ensure the SDR is operational and has not exceeded critical operational environmental parameters. We will discuss the role of remote management tools to update firmware, recover from errant s/w or firmware malfunctions and push critical patches to multiple deployed radios and how these tools improve the maintainability of deployed SDR's. In this session we will outline how a new generation of SDR's are being implemented with these Reliability, Availability Serviceability and Maintainability (RASM) capabilities and how these features were implemented as part of the design of the radio

14:00 - 15:30
TS9:  Advanced Signal Processing

An Evaluation of Adaptive Equalizers on Non-Linear HF/Ionospheric Channel Models

Noel Teku, Garrett Vanhoy and Tamal Bose (University of Arizona, USA)

HF communications (3-30 MHz) have been a popular medium for emergency, military, and hobbyist applications because they generally do not require significant architecture or equipment. For long range HF communications, which are common, this involves having signals reflect off the Earth's ionosphere. However, the ionosphere is highly unstable, varying with respect to space, time, and frequency, which has prompted research into compensating the effects of the ionosphere in the receiver to maintain robust communications even when faced with harsh channel conditions. This instability in the ionosphere causes the channel to change between non-linear and linear models at different times. Thus, the objective of this paper is to utilize and evaluate the performance of three non-linear adaptive equalizers in recovering distorted signals transmitted through a non-linear HF channel. The performance of the equalizers will be characterized using bit error rate (BER) and convergence rate.

 

3.5 GHz Waveform Generation for Testing and Development of ESC Detectors

Raied Caromi (National Institute of Standard and Technology, USA); John Mink and Michael R. Souryal (National Institute of Standards and Technology, USA)

Environmental Sensing Capability (ESC) sensors will be used in the 3.5 GHz Citizens Broadband Radio Service (CBRS) to detect and report the presence of federal incumbent radar signals in 100 MHz of spectrum. Unlike traditional radar detection schemes, ESC sensors will not have full knowledge of radar waveform parameters such as pulse repetition, pulse duration and center frequency of the incumbent radar. Furthermore, ESC sensors are expected to detect incumbent radar and identify its operational channel in the presence of interference from CBRS devices and adjacent-band emissions. This paper presents signal processing procedures and a software tool for generating ESC test waveforms. These waveforms cover multiple testing scenarios in which one or more radars operate in the presence of interference signals such as LTE TDD signals and adjacent-band radar emissions. We utilize field-measured radar waveforms acquired by the National Advanced Spectrum and Communications Test Network (NASCTN) in the 3.5 GHz band with a 225 MHz sampling rate. Field-measured waveforms include channel propagation effects such as time-varying multipath fading and pulse dispersion, similar to what an actual ESC sensor will observe. We present the signal processing blocks for decimating the measured waveforms and mixing them with interference signals at specified frequency offsets. Gains are adjusted to achieve a desired signal-to-interference ratio (SIR), defined as the ratio of the peak power of the measured radar waveform to the peak or average power of the interference. In addition, we provide an open-source software tool with a graphical user interface (GUI) to visualize the resulting waveforms and to automate the process of generating the waveforms. The tool can randomize signal parameters such as start time, frequency, and SIR. The generated waveforms are saved as 90 second, 25 MHz sampled in-phase/quadrature (IQ) data files, and their parameters are saved in JavaScript Object Notation (JSON) format. The waveforms and their parameters can be used by ESC applicants and developers for training and testing incumbent radar detection algorithms.

 
 
TS10: Dynamic Spectrum Testing and Validation

Test and Evaluation of a Spectrum Access System using a Software-Defined Radio / Cognitive Radio Testbed and Experiment Management Framework

Shem Kikamaze (Virginia Polytechnic Institute and State University, USA); Carl B. Dietrich (Virginia Tech & Wireless @ Virginia Tech, USA); Jeffrey Reed (Virginia Tech, USA)

Demand for wireless data capacity is projected to grow rapidly over the next few years due to Internet of Things / machine-to-machine communications, intelligent transportation systems, unmanned aeronautical systems, and other emerging applications in addition to current application such as video. The U.S. government is pursuing sharing of radio frequency bands among multiple users to help meet this demand, an approach that the U.S. President's Council of Advisors on Science and Technology projected to provide a trillion dollars' worth of societal benefits and millions of U.S. jobs. The Federal Communications Commission is following the recommendation to share spectrum, and so far is using an approach in which spectrum access systems (SAS) manage the radio spectrum. SAS will enable spectrum users to request and reserve frequencies for use in specific times and locations and thus provide centrally managed coordination of spectrum use by multiple users that can have varying levels of spectrum access priority. However, many research issues remain in design and implementation of effective SAS, and experimentation and testing are required to assess, compare, and refine the many possible approaches to SAS design and implementation. We describe comprehensive plans and initial test results for evaluating effectiveness of an experimental open-source SAS for use in managing shared radio frequency bands. This SAS is being developed based on specifications developed by the Wireless Innovation Forum. The open-source SAS employs a radio environment map (REM), a database of observed spectrum use, as well as a database of frequency users and reservations. Potentially, many parameters of the SAS could be varied, including the granularity of spectrum reservations in time, frequency, and space, the number of sensor nodes used to gather information for the REM, and whether and how additional data are crowd-sourced from spectrum users. In the case of crowd-sourced data, user reports of spectral conditions would need to be authenticated and a variety of approaches could be used to assess the accuracy and trustworthiness of these reports. To enable evaluation of the SAS' performance for various combinations of parameters, we allow the SAS to manage multiple radio links or networks in a shared radio frequency band within a confined volume. This is accomplished through use of a cognitive radio testbed (CORNET) that consists of software-defined radio (SDR) front ends attached to servers that perform mush of the physical layer signal processing. While commercial off-the-shelf (COTS) radios will be used in future experiments, the initial experiments use an experimental SDR waveform that uses orthogonal frequency division multiplexing (OFDM, the basis of LTE and WiFi physical layers) within distinct networks of two or more radios each that share a frequency band, with each radio consisting of an SDR front end and a server running the OFDM waveform software. These networks may or may not cooperate directly, but in either case they coordinate their spectrum use through communication with the SAS. An experiment management framework is used to conduct the experiments or tests on the testbed and to measure performance metrics such as throughput and latency of each radio link or network while under management of the SAS. This in turn enables measurement of statistics such as aggregate throughput and minimum link throughput under various combinations of SAS parameters, numbers of spectrum users, and data traffic for each user, link, and network. In addition, the experimental OFDM radios described above are designed for use with the experiment management framework, and report their adaptations, including those based on reservations issued by the SAS, to a scenario manager object that is part of the experiment management framework. Experiments and tests that use the experiment management framework, experimental OFDM waveform, and SDR-based testbed will enable evaluation and refinement of algorithms, parameter settings, and physical configuration of the SAS and associated resources. The end goal of the research is to address questions regarding how to design, implement, and configure a SAS to manage shared bands in the radio frequency spectrum effectively in terms of both performance and cost, through empirical results that can validate analytical and simulation studies

.

Validation of a CRV Model Using TVWS Measurements

Khalil Anderson (University of Maryland, Baltimore County, USA); Lauren Lusk (University of Oklahoma, USA); Marti Hands (Texas Tech University, USA); Garrett Vanhoy (University of Arizona, USA)

Autonomous cars are growing in commercial popularity. This growth will cause the need from wireless transmission of data to these cars to not only aid the software to drive more efficiently but to also entertain the driver. Currently, autonomous vehicles are allowed to transmit using the band specified by the IEEE protocol 802.11p. While autonomous vehicles can transmit data using the 5.9 GHz band(5.850-5.925 GHz), the band may not support wireless transmission of media to vehicles' infotainment systems. This poses a problem as autonomous vehicle users will demand this type of entertainment to be built into vehicles to engage them since they will eventually be free from the task of driving. This requires an alternative to be found. With the switch from analog to digital television, the government has vacated the analog TV bands. This transition provides a possible solution to the limitations of 802.11p transmissions. The vacated space is called TV white space. One proposed use of this white space is to provide Wi-Fi. This idea has been called White-Fi. According to our research, researchers have measured whether the specific frequencies are occupied but do not provide the unprocessed data. With this in mind, we measure the occupancy of the TV white space and we simulate how a network using this band would perform under the multiple scenarios of everyday driving.

 

Building a 128x128 Coherent Massive MIMO Testbed: Architecture, Challenges and Real-time Implementation

Jun Chen and John Ye (National Instruments, USA)

Large-scale antenna arrays or massive multiple-input-multiple-output (MIMO) systems can provide substantial improvements in energy and/or spectral efficiency due to the greatly improved spatial resolution and array gain. It is paramount to develop a real-time prototyping system for testing and validating capabilities, efficacy and limitations of methods, algorithms and signal processing involved in the massive MIMO technology. In this paper, we propose architecture, design and implementation of a 128x128 coherent massive MIMO testbed. This flexible and expandable prototyping system includes total 256 coherent radio-frequency (RF) transceivers at both transmit and receive sides. Real-time baseband signal processing with 10 Tbps throughput capability can be distributed over 64 Xilinx Virtex-7 FPGAs with copper and optical interconnects. We also present effective methods to mitigate the impact of hardware impairments in distributed RF radios and large-scale antenna elements such as time/frequency synchronization, in-phase/quadrature imbalances and antenna coupling effects. Initial measured results demonstrate the desired performance of transmit and receive processing algorithms in massive MIMO systems with non-ideal hardware. Solutions to key design problems and challenges are discussed in massive MIMO prototyping and future improvements are also recommended.

16:00 - 17:30
TS11: SDR and Cognitive Systems

Software Defined Radio integrated with flight safety and mission management into IoT subsystems - A case study at amazonian scenario

Nina Figueira (Brazilian Army & NIPCAD Company, Brazil); Victor Bramigk (Brazlian Army, Brazil); Andre P Mattei (Senai Innovation Institute for Embedded Systems, Brazil); David Moura(Centro Tecnologico do Exercito & Brazilian Army, Brazil)

Due to the variety of Embedded Systems available on the market, the sensors that can be mounted on these platforms, the inherent risks of autonomous navigation and the various waveforms available for communication, it is necessary to develop an architecture in the context of the Internet of Things (IoT) To facilitate the portability of applications. It is also desirable that this architecture enables the adaptability of the waveform according to the application and the operating context. Thus, considering an embedded system with a Mission Oriented Sensors Array (MOSA) that does the mission management associated with a navigation safety system, the In-Flight Awareness Augmentation System (IFA2S), both transmitting data through an Software Defined Radio, it is desired to develop a IoT architecture that standardizes the data processing and communication of the applications that run on the network, with the waveform control system. One of the advantages of this standardization is the construction and portability of adaptive systems, modifying the SDR waveform and processing the data according to the operating conditions. One way to achieve this standardization is to use the standard middleware architecture employed for waveforms, in conjunction with the structure of a MOSA supported by an IFA2S. Thus, an architecture that integrates a MOSA and an IFA2S to an SDR, all in an IoT context, will be presented, which will promote greater interoperability and portability of hardware and software. The introduction of this article presents the SDR Project of the Ministry of Defense of Brazil. Section 2 presents MOSA and IFA2S. Section 3 describes the interaction of MOSA and IFA2S with RDS within an IoT architecture. Section 4 presents a case study of the IoT architecture proposed in the Amazonian environment. Section 5 presents the final considerations.

 

Low-power, Low-cost Software Defined Radio and Its Applications

XiaoDong Zhang (General Processor Technologies, P.R. China)

Currently, most Software Defined Radio (SDR) related products in China focus on research and prototyping of communications systems. However, when mass production is required, the practical design and codes have to be optimized to fit into new platforms, e.g. ASICs. In reality, such a development flow requires too much investment and time to enter into the market. In this paper, we introduce China-based production environments of the Sandbridge Sandblaster SB3500 chip and system. It is a heterogeneous SDR platform for wireless communications. The platform consists of heterogeneous radio hardware programmed using the mainstream high-level C language. The main features include: 1) Unified source programming composed of host CPU and kernel DSP codes. 2) Flexible and powerful instruction set architecture optimed for digital-communications 3) A unified address space between multiple DSP and CPU cores. 4) Unified profiling across heterogeneous cores for functional and timing verification using a system simulator. 5) A microarchitecture that supports instruction-level, data-level and thread-level parallelism. and 6) An extremely low-power interleaved hardware multithreaded implementation. The SB3500 SDR platform has been used in research, prototyping, and most recently in China-based production systems. The software programming model for product development has reduced the typical time to market by more than six months.

TS12: Dynamic Spectrum Interference Mitigation and Management

The Effects of Interference and Mitigation Techniques Upon Wireless Systems

Gregory Buchwald (Motorola Solutions, Inc., USA)

As spectrum becomes more densely populated, the potential for interference and denial of service escalates rapidly. Interference may be categorized in three forms: Unintentional, engineered, and malicious . Unintentional interference has been highlighted in the press on many occasions recently including disruption to 800MHz police narrow-band communications on the west coast, disruption of communications in a North Chicago suburb, and several other noted instances. Engineered interference refers to the purposeful limitation of service area allowing desired communications to occur while also allowing multiple users to occupy the same spectrum within a smaller than normal protection zone. Nearly all LTE systems operate in this manner. In the case of engineered interference, the noise floor is effectively raised, reducing radius of service while allowing substantially larger number of users to share spectral resources. Finally, malicious interference refers to various forms of intentional disruption of communications including the jamming of consumer cellular and LTE services, and WiFi, as well as mission critical public safety, critical infrastructure, and other professional and commercial communications systems. This presentation will discuss each of the three generalized forms of interference, detection of interference, methods to locate and alleviate interference, and mitigation techniques that may be applied towards resiliency from interference sources. Specific examples of interference will be presented, along with methods utilized to locate the interference and nullify it. Additionally, the recent Department of Homeland Security (DHS) exercises that took place in 2016 and 2017 will also be discussed in overview.

 

Spectrum Sharing between the Fixed-Satellite Service and Fixed and Mobile Broadband Services in the 3.7-4.2 GHz Band: A Case Study

Andrew Clegg (Google, USA)

There is growing interest in the use of the 3.7-4.2 GHz band in the U.S. for some combination of mobile and fixed (point-to-point and point-to-multipoint) applications for broadband delivery. However, the band is heavily encumbered by fixed-satellite service (FSS) earth stations that operate in the space-to-Earth (receive-only) direction. The common perception is that sharing with broadband systems would be a significant challenge for FSS, but to date, technical studies have not been conducted to confirm the validity of this hypothesis given the current state of the band, the nature of modern broadband systems, and the emergence of dynamic band management systems, such as Spectrum Access Systems. This talk will present the initial results of a compatibility study between actual FSS deployments and potential fixed and mobile systems throughout a 30x30 km region in the the Kansas City, Kansas, metropolitan area. The study is based upon a manual verification of the FSS licensing data for the Kansas City area, and uses a conservative propagation model (ITM) adopted by WInnForum for protection of a small number of FSS sites operating below 3.7 GHz from interference caused by the Citizens Broadband Radio Service in the same band. Further refinements of the simulation include the impact of more realistic propagation models that take into account the impact of clutter (buildings and trees) along the propagation path. The results of the study include a quantitative understanding of the ability for fixed and mobile broadband systems to operate in the 3.7-4.2 GHz band without causing harmful interference to FSS earth stations, given an actual FSS deployment scenario.

 

Spectrum Data Mining: Measurement-Driven Insights for Sustainable Spectrum Management

Amir Ghasemi (Communications Research Centre Canada, Canada)

Communications Research Centre Canada (CRC) has deployed a network of spectrum sensors with high time and frequency resolution for continuous monitoring and analysis of spectrum activity in different bands. In present work, the RF measurement data collected over a long period of time is used to characterize the spectrum usage of incumbent users in Land Mobile Radio (LMR) bands. Statistical analysis of measured data indicates daily and weekly spectrum usage patterns, especially in the more heavily-used public safety channels. The identified incumbent activity patterns are then leveraged to predict future spectrum usage, thereby enabling a data-driven proactive approach to spectrum assignment and management where spectrum shortage and oversupply across different networks can be predicted and managed. We further investigate correlations between spectrum activity and external factors such as severe weather and special events. The findings confirm existence of such correlation and provide further insights for dynamic spectrum assignment in Land Mobile Radio bands.

 
powered by MemberClicks