MELODY: A Platform-Agnostic Model for Building and Evaluating Remote Labs of Software-Defined Radio Technology

Studies have emphasized the pivotal role of remote laboratories in delivering high-quality engineering education. Specifically, within the domain of wireless communication, remote laboratories offer students the opportunity to actively engage in practical Radio Frequency (RF) experiments, much like their traditional in-person counterparts. Through online platforms, students can construct RF prototypes, gaining hands-on experience in communication theories. However, the lack of a standardized model for designing and evaluating remote labs in RF education limits their consistency and effectiveness. This paper introduces MELODY, a model and classification framework tailored explicitly for Software Defined Radio (SDR) remote laboratories. This model is characterized by its technology-agnostic and open-source approach. The paper evaluates existing SDR remote lab projects, analyzing their architecture and design choices. MELODY is presented in details, encompassing process, services, platform, and infrastructure layers. Additionally, MELODY provides a classification framework, assigning ratings to SDR remote labs on a 1 to 5-star scale aligned with engineering standards. The classification framework’s rubric assesses isolation challenges, calibration, scalability, and remote SDR lab availability in order to compare with other SDR remote labs. Practical applications developed by the Remote Hub Lab (RHL), such as the remote labs RHL-RELIA and RHL-RADAR are explored, showcasing how MELODY can be effectively applied. MELODY aims to establish standardization, ensuring consistency and quality assurance in remote SDR labs, fostering innovation, skill development, and collaboration within engineering education.


I. INTRODUCTION
Courses in electrical and computer engineering commonly incorporate a laboratory component that aims to provide students with practical experience in system design, programming, and problem-solving [1].Traditionally, students are given physical lab kits and granted access to lab facilities to carry out their assignments [2].However, the shift to remote instruction prompted the need for a more stable, efficient, and reliable solution for hardware access [3].As a result, remote labs have emerged as an appealing approach that continues The associate editor coordinating the review of this manuscript and approving it for publication was Nkaepe Olaniyi .
to be a viable solution for offering engineering labs in the post-pandemic era [4].
The pedagogical effectiveness of remote and virtual laboratories has been consistently validated [5], [6], [7], [8], [9] and their development is closely intertwined with the role of internet technology.With the advancements in cloud computing, it has become increasingly feasible to achieve a high level of realism and establish fast and reliable setups, enabling hardware to be accessed conveniently at any time and from any location.This technological progress facilitates the creation of a more authentic learning environment where students can configure actual hardware and observe its functionality through a simple web browser, without the need for specialized software [3].This immersive experience enhances student engagement and facilitates a seamless transition from physical to remote labs.
Remote laboratories have garnered significant attention in educational research since the 1990s, as evident from the various definitions found in the literature [10], [11], [12].While the terms ''remote lab'' and ''virtual lab'' are often used interchangeably, it is crucial to distinguish between the two.Virtual labs are computer-based applications that simulate non-physical environments, whereas remote labs enable users to access and control physical equipment from a remote location through computer and communication infrastructure [13].Remote labs offer students convenient access to equipment at any time, without geographical restrictions, promoting collaboration among peers and improving accessibility for students with disabilities [14] and promoting equitable access [15].
In addition, the abundance of available technology and internet tools has made it feasible to deliver a positive user experience.Simultaneously, the development of remote labs that actively engage students in hands-on participation presents technical challenges that need to be carefully addressed and overcomed [16].
In the realm of Electrical and Telecommunication Engineering, wireless communication courses that utilize radiofrequency (RF) techniques are integral components of the curriculum.Within a traditional communication laboratory, students are provided with a kit of components to facilitate their lab assignments [6].This kit includes RF devices for transmitting and/or receiving signals, accompanied by other essential components such as antennas and cables.The hands-on nature of the lab requires students to connect and engage in various practical activities.Software Defined Radio (SDR) technology significantly simplifies this process, reducing connection time and offering flexible hardware that can be modified through lines of code.This adaptability and efficiency make SDR a valuable tool for both traditional in-person labs and remote learning environments.SDR allows the same hardware to be used for different types of communication, enhancing its versatility and applicability in various educational settings [17].
In the domain of SDR remote laboratories for educational purposes, there exist multiple initiatives that pursue unique concepts without adhering to a standardized model that can serve as a reference.This lack of a unified approach hinders the ability to offer consistent and standardized products and perform effective evaluations.By adopting a model and incorporating feedback, the quality assurance based on technological advancements can be elevated.This approach will have a positive impact on education, empowering users with improved learning experiences and fostering research and collaboration among developers and instructors [18].Consequently, these projects will align more closely with the high standards of educational quality [19].
The primary objective of this paper is to introduce MELODY (Model for Evaluating Labs Of Software Defined Radio Technology), a model and classification framework designed for SDR remote laboratories.To provide a comprehensive perspective, this study also includes implementations of remote labs based on the MELODY model (RHL-RELIA and RHL-RADAR).These implementations are presented in detail, along with references to specific educational scenarios where RHL-RELIA has been applied.This contribution intentionally excludes discussions related to student engagement or pedagogical performance, as its primary aim is to address conceptual and functional aspects of MELODY.Instead, the core objective involves an examination of existing remote lab SDR projects for educational purposes, evaluating their contributions through an analysis of their architecture and design.Some projects are in the form of prototypes, offering hardware and software developments.
The model and classification framework play a pivotal role in the standardization and consistency of remote labs in SDR technology, offering a structured approach for design, evaluation, and comparison.This ensures fair assessments and promotes effective evaluation of lab capabilities and performance.Additionally, it guarantees quality assurance by establishing guidelines and criteria, leading to improved learning experiences for students.Furthermore, the model fosters the advancement of SDR technology by encouraging innovation and research in remote labs.It also facilitates cost-effective solutions and aligns with industry needs, equipping students with relevant skills.In summary, the model and framework impact education, research, and collaboration while driving advancements in SDR-based remote labs.
The paper is structured as follows: Section II provides an overview of five contemporary remote SDR labs dedicated to engineering education, highlighting their key features.Also, explaining the concept of Ultra-concurrent laboratories.In Section III, the MELODY model is detailed, explaining its features like flexibility, architectural design, and layered approach.Additionally, the evaluation of MELODY composed by Datasets for standardized testing and Classification Framework are detailed.Section IV details the rubrics for the classification framework of the MELODY model, which serves as a valuable tool for comparing it with other projects.This framework adheres to engineering standards and employs a rating system ranging from 1 to 5 stars.Each rating is accompanied by an explanation of the evaluation criteria utilized in the assessment process.
In Section V, the paper elucidates practical applications developed using the MELODY model such as the RHL-RELIA and a protoype of RHL-RADAR implementations.Section VI compares the five SDR remote lab projects analyzed evaluated under MELODY's Classification Framework plus the RHL-RELIA and RHL-RADAR implementations.Finally Section VII, encapsulates the conclusions drawn from this comprehensive study.

II. BACKGROUND
In order to introduce a model and framework, conducting a thorough review of the state of the art in remote lab projects for educational purposes is essential.By leveraging the insights gained from this review, MELODY seeks to make valuable contributions to the advancement of SDR remote labs, ultimately enhancing their educational impact and fostering an environment conducive to effective evaluation practices.

A. SDR REMOTE LABORATORIES
In a conventional SDR remote laboratory, users gain access to both transmitter and receiver components via a central server, typically hosted in some university's infrastructure.This configuration enables students to access the SDR hardware either from their own homes or within their university campus.Notably, the physical hardware may be located within the student's own university or potentially at a different university in the case of federated remote laboratories [20].SDR lies in its capacity to configure multifunctional wireless systems through software, all without the need for any modifications to the SDR hardware.
In the realm of SDR remote labs, contemporary examples can be found showcasing cutting-edge developments and advancements.The diversity of these studies offers a comprehensive perspective on the current state of SDR remote labs and their applications, highlighting their relevance and potential impact in different fields of study.Important aspects in each reviewed project are centered around: • User requirements for necessary software tools.
• Type of network used for secure access.Usually, Virtual Private Network (VPN) is the most commonly utilized technology.
• Access options, including in-person at the lab or campus, as well as access from home.
• Mechanisms of isolation in case of large scalability.Five initiatives are described below: Emona TIMS offers a comprehensive collection of hardware and software solutions, along with modular experiments that are well-suited for the curriculum of both analog and digital communication systems [21].As a result, it has gained significant popularity and is extensively utilized in communication systems laboratories across numerous universities.Its foundational understanding of employing it as an educational tool can be found from [22], [23], and [24].
Hardware is composed by TIMS-301C which is a flexible platform that is able to support other hardware like the TIMS-SDR Plug-in module and other peripherals that permit to complete an experiment.Augmenting the student experience, the system incorporates a built-in virtual instrument that connects to PCs, providing oscilloscope functionalities and spectrum analysis using Fast Fourier Transform (FFT).Its user-friendly design boasts color-coded inputs and outputs for easy identification [25].
EMONA TIMS represent an efficient teaching solution for introducing students to the world of SDR in a simple and expedited manner [26].By utilizing experiment template files, users can develop their own experiments without the need for laborious file downloads, Linux installation, or code compilation.The TIMS-SDR Kit is designed to be a zero-install and plug-and-play solution, comprising a complete Linux system and the latest version of GNU Radio on a USB stick.This approach unlocks the full potential of GNU Radio, enhancing its accessibility and user-friendliness for students.
The primary contribution of this product is its provision of an educational solution that enables the teaching of SDR courses without the requirement of installing specialized software or prior knowledge of programming languages like C, Python, or GitHub.To utilize EMONA SDR, students only need to insert a bootable USB drive into their computers, perform a simple reboot, and the EMONA operating system will seamlessly launch.However, it should be noted that this solution necessitates the use of EMONA hardware, specifically a core hardware module (TIMS-301C) and a plug-in module (TIMS-451 SDR) [26], which means that students must access their university's lab facilities to operate it.
While the EMONA TIMS-SDR Kit offers numerous advantages, such as ease of use and streamlined setup, its dependence on specific EMONA hardware may limit its scalability for providing flexible laboratory access to a large number of students.Table 1 summarizes most important features.

2) HUAZHONG UNIVERSITY
This project from Huazhong University of Science and Technology offers a platform for learners to gain practical knowledge in communications by providing access to a remote lab environment [27].
Employing the Browser/Server (B/S) architecture, this web-based system establishes connections among teacher servers, student clients, hardware devices, and other functional elements through the network.This network framework allows students to access all available resources without the need to physically be present in the laboratory [28].The experimental equipment encompasses a software radio platform and an online test instrument, which is selected based on specific requirements.The client interface facilitates the display of experiment results, measurement data, and instrument images [29].To this end, two oscilloscopes are utilized to measure the modulated signal of the transmission section and the demodulated signal of the reception section.The results of operations and measurements can be sent back to the client interface for visualization.Additionally, students have the capability to observe real-time experimental images through a camera [30].Further details about the features of this project are outlined in Table 2.

3) TECHNICAL UNIVERSITY OF CLUJ NAPOCA
Technical University of Cluj Napoca in Romania is actively engaged in a remote laboratory project utilizing the SDR technology.To access remotely the system a Virtual Private Network (VPN) is used and students can interact with the lab via virtual machines [31].Featuring a setup comprising of 2 ADALM-PLUTO SDRs and one computer per unit.The user requirements include the use of GRC, which is an opensource software [32], [33].
After establishing the VPN connection, computer and ADALM-PLUTO SDR work as though they are part of the same local network.Through this remote control mechanism, students gain the capability to access SDR resources and engage in diverse application development from any place with an internet connection [34].This approach is highly valued by students, particularly since the available platforms are limited and not accessible beyond the laboratory premises.Consequently, students can conveniently access the equipment from their homes or dorm rooms, allowing for continuous work without any disruption.
However, the level of SDR isolation is not specified in the available information (Table 3).

4) HYDERABAD UNIVERSITY
The FM transmitter project located in Hyderabad, India, utilizes the GNU Radio software and incorporates one USRP device for FM transmission reception [35].
The design of this FM transmitter achieves a high-quality sound signal transmission.This tool is to educate students on fundamental digital signal processing tools and RF concepts, encompassing filtering, sampling rate conversion, and demonstrating the utilization of SDR for real-time application design.Leveraging the SDR package greatly simplifies the utilization of the FM framework.Using LabVIEW as a software tool to program USRP enables the creation of communication system prototypes, enabling real-time performance verification.Substantial research is underway to explore SDR applications in diverse fields, aiming to make it viable for commercial deployment [36].Further insights on the FM transmitter project are presented in Table 4.

5) FORGE
The FORGE (Forging Online Education through FIRE) initiative facilitates access to remarkable FIRE testbed infrastructure for both educators and higher education students [37].Within its various project domains, wireless communication occupies a significant role.In this particular context, the FORGE initiative employs a testbed named IRIS [38], consisting of 16 USRP (Universal Software Radio Peripheral) software-defined radios.
Each individual USRP unit is linked to a virtual machine, which operates either on the IRIS software or GRC.The allocation of resources occurs automatically through the gateway server, which also supports the initiation of experimentation services such as data collection from measurement points.
Within this laboratory environment, students embark on an exploration of the fundamental principles of Orthogonal Frequency-Division Multiplexing (OFDM) for wireless signals.This lab not only connects students with cutting-edge research hardware but also offers the opportunity to delve deeply into the intricate workings of the digital multi-carrier modulation technique as it pertains to wireless communications [39].Over the past year, this course has been delivered nine times in Brazil, Mexico, and Ireland, engaging a total of 148 students and enabling them to conduct at least 1,400 experiments [38].Features of FORGE project are summarized in Table 5.

B. ULTRA-CONCURRENT LABORATORIES
According to Narasimhamurthy et al. [40], an ultraconcurrent laboratory is a type of remote laboratory that enables students to interact with real data, although all of the data and multimedia have been pre-recorded.This type of remote lab is convenient when there are challenges with infrastructure and having a remote laboratory station for each student is not feasible.Soll and Boettcher [41] extends the benefits of ultra-concurrent laboratories pointing out that they have no time limit (as compared to real-time remote laboratories).
Ultra-concurrent lab approach allows to deliver an experience that closely mirrors what students would encounter in both traditional hands-on labs and real-time remote labs [42], [43], [44].Notably, ultra-concurrent labs go beyond mere videos, as they offer interactivity by enabling students to actively engage and make choices.This interactivity is facilitated by a dataset containing numerous experimental combinations.Unlike simulations, these labs rely on authentic data collected from real equipment, avoiding dependency on simulation engines.Within these labs, users can experiment with pre-recorded data, results, images, or videos from actual labs.
This unique solution empowers users to engage with genuine results, while accommodating a multitude of concurrent users without necessitating real-time handling of physical equipment.Instead, requests are directed to a database housing the pre-recorded dataset of the lab.The advantage of these labs lies in their capability to be accessed concurrently by numerous students, eliminating the need to replicate physical equipment.Furthermore, their maintenance primarily involves software, rendering them more cost-effective.Nonetheless, these labs are better suited for experiments with limited variable control.Chemistry practices adhering to set scripts with few variations can fit within these constraints.However, labs involving programming, diverse development boards, or extensive interaction options necessitate a realtime approach.Stanford University1 follows a similar approach by creating digital twins for their remote labs.Similarly, VirtualRemoteLab2 from the Ludwig Maximilian University of Munich offers both real and virtual setups for remote experiments in optical spectrometry due to the equipment's cost and calibration requirements.
Despite these advancements, remote labs have not gained widespread adoption as a substitute or supplementary tool in classrooms, potentially due to persisting practical limitations [45], [46], [47], [48].Addressing and mitigating these limitations is essential to promote greater utilization of remote labs.

C. BACKGROUND SUMMARY
From all these reviewed projects, there are considrable range of contemporary examples in SDR remote labs, showcasing their diverse applications and advancements.Projects like EMONA TIMS, Huazhong University's web-based system, Cluj-Napoca University's VPN-based setup, FORGE and and the Hyderabad University's FM transmitter project all contribute unique approaches to remote SDR labs.
The lack of a technology-agnostic framework model hinders the efficient development of both basic and advanced SDR remote laboratories.Additionally, the absence of such a model makes it challenging to compare these laboratories with existing literature.Thus, it is essential to establish a comprehensive model that addresses these limitations, enabling the improved development and evaluation of SDR remote laboratories.
Ultra-concurrent laboratories offer a distinct solution that allows multiple students to interact with pre-recorded real data, providing an authentic experience without the need for real-time handling of physical equipment.While these labs excel in certain scenarios, such as experiments with limited variable control, challenges persist in fully integrating remote labs into classrooms due to practical limitations.
The MELODY model aims to bridge these gaps by providing a structured framework for designing, evaluating, and comparing SDR remote labs.This model ensures consistent quality and offers a standardized evaluation process.MELODY seeks to enhance the impact of remote SDR labs, fostering innovation, collaboration, and skills development in engineering education.

III. MELODY -MODEL
Building upon the preceding explanation concerning remote laboratories founded on SDR technology, this paper introduces a model characterized by its technology-agnostic framework and utilization of open-source software.Both of these elements constitute fundamental components that facilitate the streamlined development of both basic and advanced SDR remote laboratories.
The benefit of proposing a framework-agnostic model lies in its capacity to construct applications and solutions without being tied to a specific framework or platform technology which empowers developers to harness various frameworks and platform technologies while creating a portable and adaptable program [49].Consequently, developers can leverage the most suitable framework and platform technologies for their projects while retaining control over the overall design and architecture of the application.
Agnostic models also empower developers to contribute feedback and drive the model's evolution.This dynamic process addresses the evolving requirements of education and the progressing technology landscape within the SDR field.This aspect holds substantial significance, especially considering that the SDR community consistently advances their technology due to its modular design.This adaptability extends to a wide range of devices, such as computers, tablets, and cell phones, underlining the model's versatility and accessibility across platforms.
In particular, given the array of options available in SDR technology, which depend on the lab's objectives and educational level, MELODY is structured with a layered approach, fostering increased interoperability, and emphasizing modularity and flexibility which is composed by: • Standardization: The model proposes a gold standard that establishes the technical challenges to be considered in advance.This standardization ensures a consistent and systematic approach to evaluating SDR projects, promoting quality and reliability.
• Layered Approach: The model facilitates the organization of the communication process into distinct layers, delineating interactions between software and hardware stages.This approach provides clarity and structure, enabling efficient development and evaluation of SDR systems.
• Interoperability: The model emphasizes the importance of promoting interoperability between different SDR capabilities.By defining common protocols and interfaces, it fosters seamless communication and compatibility between diverse SDR devices and platforms.
• Modularity and Flexibility: Understanding the unique characteristics of SDR devices and their scalability is crucial.The model encourages modularity and flexibility in system design, allowing adaptability and expansion as SDR technology evolves.
The creation of the MELODY model can be classified into four primary segments: Architecture and shared terminology, Creation guidelines, Datasets for standardized testing, and Classification Framework.The initial two elements constitute the model's design, while the latter two are essential for evaluating its effectiveness.A visual representation of these key components is presented in Figure 1 through a block diagram.

A. ARCHITECTURE AND SHARED TERMINOLOGY
The architecture and shared terminology aspect pertains to establishing a comprehensive structure, components, and interactions within the system among various elements through a layered design.When considering an RF remote laboratory built upon SDR technology, the hardware, firmware, and software components are interconnected in a modular manner.This design approach ensures that the hardware and firmware elements are treated as software entities, resulting in independent layers.This configuration facilitates a streamlined redesign process and offers flexibility to accommodate emerging hardware and software technologies.
Within MELODY, this particular element comprises of four distinct components: Process, Services, Platform, and Infrastructure.Each of these components fulfills a specific role within the system.A block diagram with all components are illustrated in Figure 2.

1) PROCESS
Within this element, the focus lies on the precise actions undertaken by users within the system.This layer encompasses two distinct user categories: students and instructors.Notably, instructors possess enhanced privileges, enabling access to private files and the ability to monitor the activities of registered students in a laboratory setting.Moreover, this layer is responsible for validating user credentials and managing ongoing user sessions.

2) SERVICES
This software component incorporates modules that furnish essential functions facilitating interaction between the Process layer and the Platform layer.MELODY integrates 12 high-level software blocks, each characterized by distinct functionalities, as detailed below: • access(): This block verifies the availability of lab resources for a designated user.
• interactLab(): Enabling dynamic access, this module allows users to engage with the lab in real time.
• loadFile(): This functionality permits the inclusion of scripts within the design, such as sequences for modulation or coefficients for specific filters.
• saveFile(): Of particular significance for RF applications, this block enables the preservation of received data, whether real or simulated, in either the time or real domain.
• dataExchanger(): This module enables the visualization of received data within a web browser, enhancing user interaction.
• authentication(): This module verifies user permissions and privileges, assigning the appropriate resources as required.
• scheduledLab(): Responsible for time allocation and queue management, this module coordinates student access to specific lab modules.
• accessLab(): Enabling lab entry, this component facilitates the allocation of a designated SDR module.
• checkAvailability(): This module ensures the availability and access of both Transmitter and Receiver modules.
• getResults(): Acknowledging correct outcomes and data transfers from the Platform layer, this component confirms successful operations.
• getStudentInfo(): Recognizing and retrieving user information, this module retrieves pertinent user data.
• setStudentInfo(): This block associates user data with a specific name, allowing students to access and retain their files for future sessions, enhancing usability and continuity.

3) PLATFORM
Within the Platform component, essential system resources are encompassed, potentially subject to alterations during runtime.Operations within this layer constitute a combination of software and firmware interactions, engaging with the computer or embedded system responsible for managing SDR hardware.This layer comprises 14 distinct blocks.
• updateVariables(): This block receives user inputs to modify SDR parameters or view parameters, facilitating updates to acquisition or display settings.• preProcessing(): Responsible for applying preliminary processing before data streaming is displayed, this block often involves domain transformation, such as time or frequency domain integration, to effectively mitigate noise.
• queueWaiting(): Utilizing computer resources, this block buffers data streaming to prevent potential data loss and maintain smooth data flow.
• accessAssignment(): Enabling data transfer within an assignment rather than a test scenario, this block ensures proper data display or storage.
• dataTransfer(): Charged with the complete and uninterrupted transfer of data, this block safeguards against glitches or losses by transmitting a sequence of zeros in such cases.
• melodyEngine(): A dynamic component housing signal processing blocks applicable to transmission and reception stages.Tailored to each experiment and application, this block showcases distinct configurations.
• melodyGRCLibraries(): Housing the Gnu Radio Companion Libraries, this block establishes standardization for operations.
• melodyReset(): Offering the ability to reset acquisitions, this block is triggered when an experiment starts or when data loss or corruption is detected.
• setStudentInfo(): Similar to the description in the Services layer, this block shares user information.
• getStudentInfo(): As in the Services layer, this block retrieves user information.
• addWidget(): Incorporating widgets to visualize varying data or the same data in different domains (Time, I/Q, Frequency, etc.), this block enhances visualization flexibility.
• dataUploader(): Collecting data from the Platform layer and facilitating its transfer to the Services layer, this block supports efficient data movement.
• reserveLab(): Ensuring exclusive SDR component access for specific users, this resource maintains operational integrity.

4) INFRASTRUCTURE
This component refers to the physical elements that support the system's operation.It includes SDR devices, antennas, and mechanisms for isolation.The selection of SDR hardware is made within the SDR Lab Station, allowing developers to choose the appropriate SDR hardware for their needs.This layer is composed by four integral elements.Authorized licensed use limited to the terms of the applicable license agreement with IEEE.Restrictions apply.In scenarios involving wireless communication, it typically holds less relevance, as both transmitter and receiver units cooperate in the communication process.Nonetheless, in use cases like radar systems, the target subject to analysis remains autonomous from the system and is consequently categorized as a non-cooperating unit.
• Ultra-Concurrent SDR Lab Station: This element comes into play when the remote lab is intended to operate in an ultra-concurrent manner.This configuration is typically chosen in SDR applications where resource consolidation is paramount, physical space is constrained, or when the components themselves are too large to be feasibly scaled.This scenario often arises in applications such as radar systems.
It's worth noting that initially, MELODY does not incorporate a camera web in the Infrastructure layer.This omission is rooted in the understanding that RF signals cannot be visually perceived by the human eye.Moreover, among the SDR remote projects analyzed in Section II, three of them do not utilize a camera.Similarly, the VISIR remote lab, renowned as the most popular remote laboratory for basic analog electronics, also operates without a camera for its functionality [50].However, it's essential to highlight that a camera can be optionally added if a developer deems it advantageous for their specific application.

B. CREATION GUIDELINES
In essence, the Creation guidelines serve as a comprehensive set of rules and recommendations for developing successful SDR lab projects.These guidelines encompass various aspects, including design principles, coding standards, best practices, and quality assurance procedures.These principles guide developers in creating SDR designs that meet the necessary timing requirements for an effective real-time SDR lab.
For instance, careful selection of appropriate SDR devices, antennas, and supporting hardware components is crucial to ensure optimal functionality and performance while avoiding data quality issues.Additionally, the development of software tools and platforms plays a vital role in facilitating data processing, visualization, and reliable data transfer to prevent any loss of information.
MELODY suggests to follow four guidelines when a SDR remote lab is going to be planed either to develop or to acquire.While these guidelines are technically optional, it is imperative to underscore that considerations such as the selection of Appropriate SDR Hardware and Scalability Concerns wield substantial influence over the overall cost associated with establishing a remote laboratory.

1) CHOOSING APPROPRIATE SDR HARDWARE
Selecting the suitable SDR device empowers users to optimize their usage according to specific requirements.The selection of SDR devices for educational purposes hinges on variables such as the academic level (undergraduate or graduate) and the scope of the course.The SDR community provides an extensive array of devices, each differing in attributes such as bandwidth, transfer rate, channel count, external synchronization capabilities, dynamic range, and more [51].VOLUME 11, 2023 127557 Authorized licensed use limited to the terms of the applicable license agreement with IEEE.Restrictions apply.

2) SOFTWARE DEPENDENCIES
Within an SDR application, users have the flexibility to set up their experiments through coding or by employing a graphical modeling and simulation environment.They can then analyze the outcomes by examining the streaming data generated.Such inputs and outputs might necessitate dedicated software, which could involve licensing prerequisites.At this juncture, developers need to deliberate on whether to utilize proprietary software or opt for open-source alternatives.For instance, SDR provides software like GRC (GNU Radio Companion), an open-source toolkit that grants users access to a wide array of SDR devices [33].

3) PLATFORM COMPATIBILITY
When embarking on the development of an SDR remote lab, it becomes crucial to assess the technological tools available to users.Consequently, the choice of a system that ensures compatibility across diverse operating systems should be carefully deliberated.

4) SCALABILITY CONSIDERATIONS
When making a decision about an SDR remote lab, it's essential to evaluate the projected user count.If there is a considerable user base, scalability options should be carefully considered.In this scenario, the concept of federated remote laboratories comes into play.This approach involves the interconnection and integration of multiple remote lab facilities, often located in diverse institutions or places, to form a cohesive system [20].

C. DATASETS FOR STANDARDIZED TESTING
Developers possess the capability to conduct unbiased performance assessments of SDR remote labs and make meaningful comparisons with other implementations.
SDR systems offer distinct features that provide flexibility in terms of hardware specifications, performance capabilities, data quality, and the number of inputs and outputs.Some systems even support external inputs for synchronization, enabling seamless integration with other devices or systems.Consequently, the datasets required for testing and evaluating SDR applications can vary depending on the specific application or use case.
By employing standardized evaluation methods and utilizing appropriate datasets, developers can accurately assess the performance of their SDR remote lab in relation to others.This facilitates the identification of strengths and weaknesses, thereby fostering advancements in SDR technology and promoting its continuous improvement.

D. CLASSIFICATION FRAMEWORK
The evaluation process enables the systematic categorization and organization of various aspects and characteristics relevant to SDR remote applications.
This classification system employs a rubric that encompasses multiple dimensions, with each aspect being evaluated and assigned a rating ranging from 1 to 5 stars.These dimensions include: • Ways of accessing the laboratory • Mechanism for isolation • Lab calibration • Dynamic parameter adjustments • Cost optimization By utilizing this rubric, the evaluation process effectively assesses and rates each aspect, providing a comprehensive overview of the strengths and weaknesses of the SDR remote applications under evaluation.This systematic approach enhances the understanding and comparison of different applications, enabling stakeholders to make informed decisions and advancements in the field of SDR technology.Next section will explain the rubric from a technical point of view.

IV. RUBRICS FOR CLASSIFICATION FRAMEWORK
Engineering metrics serve as quantifiable indicators that measure the level of quality and capabilities of remote lab designed.This Classification Framework gathers, categorizes, and organizes the diverse aspects and characteristics of SDR remote applications.MELODY employs a five-level classification system for quality assessment, encompassing isolation, calibration, scalability, availability, and usability with each one being rated on a scale of 1 to 5 stars according to the features it incorporates.

A. ISOLATION
When considering isolation, it is important to address the issue of interference that arises when multiple SDRs units works simultaneously within the same or nearby bandwidth, a situation often encountered in scalable remote labs [52] or when labs are confined to limited spaces.As shown in Figure 4, this interference can significantly interfere other stations reducing the probability of a success transmission for other receivers.
While there is not a standardized metric to quantitatively measure interference in RF labs, lessons from other engineering contexts provide valuable insights for ensuring satisfactory data quality in similar situations.In standards for wireless communication, such as WiFi and Bluetooth coexistence, various effective techniques have been developed.Adaptive Frequency Hopping (AFH) enables Bluetooth devices to dynamically switch between frequency channels, preventing interference with WiFi transmissions.Conversely, Dynamic Frequency Selection (DFS) is utilized by WiFi devices, enabling them to detect and avoid radar signals and other sources of interference, thus minimizing disruptions to Bluetooth devices [53], [54].
In the realm of Analog Digital Converters (ADCs), manufacturers offer detailed technical specifications to customers, which include information about the interference present in their products.While the Signal to Noise Ratio (SNR) is a well-known parameter that provides insights into this aspect, there is another important metric called Spurious Free Dynamic Range (SFDR).SFDR evaluates the converter's ability to differentiate between the desired signal and any undesired spurious signals in the output [55].
Drawing from these insights, MELODY establishes its own criteria for qualifying standards.Employing a methodology similar to SFDR analysis, as shown in Figure 5, MELODY compares the transmitted signal power of a unit to the power of the maximum interference originating from neighboring units.For data telecommunication applications, achieving a Bit Error Rate (BER) of 10 −13 is essential to have a coexistence that ensure a proper communication [56].
In the context of Quadrature Phase Shift Keying (QPSK) modulation, this value, when translated into Signal-to-Noise Ratio (SNR) levels, approximately equates to 19 dB, as illustrated in Figure 6.This outcome serves as the fundamental benchmark for the rubric, earning it an initial rating of one-star.Subsequently, each additional 3dB exceeding this minimal threshold contributes an extra star to the evaluation.As a result, the rubric isolation criteria is expressed in Table 6, with the highest five-star rating extending from 28 dB to the point where the receiver reaches full scale (FS).

B. CALIBRATION
Calibration of RF devices ensures that the receiver provides accurate measurements and readings.This is crucial in  various wireless applications, such as communication systems and radar applications, where precise data is essential for proper signal processing and analysis.Within the MELODY framework, five essential engineering methods (Linearity, Calibration, Input and Output Noise Power, Minimum Detectable Signal, and Frequency Stability) are introduced for proper calibration.

1) LINEARITY OF THE RECEIVER
A receiver is considered linear when there exists a constant proportion between the input and output, inclusive of any associated noise.This test determines whether the receiver operates in a linear manner.Linearity is expressed in equation 1 as presented in [57]. where: G RX = Power Gain of the receiver.

2) CALIBRATION OF RECEIVER CONSTANT
Once it's established that the receiver operates linearly, the Receiver Constant (RC) can be calculated.RC is a parameter that quantifies the relationship between the input and output power of the receiver.This input power of the receiver is sourced from a Signal Generator or Function Generator device, which produces a Test Signal Power in watts.It's crucial that this generator can be precisely calibrated and adjusted in milliwatt increments.
On the other hand, the measured output power of the receiver depends on the receiver bits resolution and it will be tested in Arbitrary Digital Units (ADU).Consequently, RC is mathematically represented as: To ensure reliable measurements, the Test Signal Power must surpass the receiver noise by at least twice its power or 3dB more in order to disregard the noise term.The constancy of RC is pivotal for maintaining precise calibration, as emphasized in [58].Ideally, this value should be set to one.Nonetheless, if it deviates, the RC value is utilized to transform the detected output power of an actual signal back to its corresponding input power.

3) INPUT AND OUTPUT NOISE POWER
The input and output noise power of a digital receiver are critical factors that impact its performance in various communication systems.When a receiver's input is properly terminated with a matched load, the input noise can be expressed by the equation: where: = System Bandwidth in Hz Regarding the calculation of output power noise, it can be estimated as the product of the measured noise power and the previously calculated RC: where: N OUT = Output Noise Power in Watts.ADU = Arbitrary Digital Units.RC = Receiver constant.For comprehensive insights into both N IN and N OUT , along with the derivation of these equations, refer to [59].

4) MINIMUM DETECTABLE SIGNAL MEASUREMENT
Minimum Detectable Signal (MDS) defines the sensitivity of a receiver.It denotes the starting point of the linear region, and it can be considered as equal to N OUT .In radar applications, MDS is typically defined as twice the value of N OUT or with an additional 3dB margin.Measurement of this parameter is detailed in [60].

5) FREQUENCY STABILITY
Another significant examination conducted on SDR devices is the frequency stability test, which evaluates the device's capacity to maintain a specific frequency consistently without undergoing drift over time.This assessment involves measuring and analyzing the extent of frequency deviation within a designated time frame.
For example, Analog Devices, the manufacturer of the ADALM-PLUTO SDR, conducted a test on the device and found that its oscillator exhibited a drift of 100 Hz within a 10-minute time frame.It's important to note that the level of frequency stability can vary among different SDR devices, as each device has its own unique characteristics and performance specifications.The test procedure is detailed in [61].
To conclude, the five calibration procedures outlined above play a pivotal role in enhancing the data quality and ensuring a robust learning experience.These calibration processes can be executed either automatically or manually by the user each time the system initiates.The complexity of the calibration tasks and the specific characteristics of the SDR hardware unit determine the level of detail in MELODY's calibration rubric, which awards more stars as more tests are completed, as shown in Table 7.

C. LAB SCALABILITY
Frequently, laboratories are established within constrained physical spaces, capitalizing on a notable advantage they provide: cost savings in contrast to conventional as highlighted by Heradio et al. [7].Nonetheless, a cooperative strategy that interlinks and integrates numerous remote laboratory installations into a cohesive system, known as Federated remote labs [62], grants users access to resources from diverse remote lab configurations.This approach not only increases the number of available units for users but also allows developers to enhance scalability without requiring additional resources.
The central components within an SDR remote lab are the SDR unit and its controller, as elaborated upon in the MELODY's architecture.The cost associated with both components are contingent upon specific needs and data quality, and these expenses also influence the scalability potential.A compilation of available SDR options can be referenced in [63].Other components of remote labs are servers, camera and software license.
Depending on the increasing number of users, the MELODY model offers the ability to evaluate whether a remote laboratory has the capacity or flexibility to integrate extra modules while maintaining a cost-effective approach.The criteria for this evaluation are outlined in Table 8.

D. LAB AVAILABILITY
In the domain of RF experiments, the time taken for can frequently be substantial, primarily due to the necessities of initializing transmissions and procuring data of superior quality.To mitigate prolonged waiting queues, remote laboratories have progressed to accommodate a substantial influx of users while safeguarding them from adverse effects.Within this context, there exist both uncomplicated and intricate strategies aimed at optimizing utilization efficiency, thereby ensuring fair access, a fundamental element during the formulation of a remote laboratory.The level of accessibility, appraised using a 5-star rating scale, is succinctly outlined in Table 9.

E. LAB USAGE SESSIONS
The availability of scheduling and time distribution options in the remote lab is of utmost importance for various reasons.First and foremost, it ensures equitable access to lab resources, allowing all users to utilize them fairly and efficiently [64].
Two methods for accessing remote labs are queuing and calendar-based booking.The queuing approach is optimized for short-time sessions, while the calendar-based booking system is tailored to facilitate longer laboratory sessions.In the queuing mode, access is granted on a first-come, first-served basis, necessitating users to wait until ongoing users complete their sessions.On the other hand, the calendar-based mode ensures that users have exclusive access to the equipment during their designated session and helps prevent overutilization of the lab, ensuring that the system functions smoothly and performs optimally at all times [20].
In summary, having scheduling and time distribution options in the remote lab not only promotes fair access for all users but also helps maintain the lab's efficiency and facilitates valuable insights into user behavior and resource utilization.The rubric to evaluate this metric is summarized in Table 10.

V. IMPLEMENTATIONS BASED ON MELODY MODEL
A. RHL-RELIA RHL-RELIA (Remote Engineering Lab for Inclusive Access) is the first implementation based on the MELODY model.It consists on a distributed network of SDR stations for the purpose of creating a wireless communication laboratory.The primary objective of this application is to provide students with a remote lab that mimics the experience of hands-on work with physical SDR kits.This project is a collaboration between Remote Hub Lab (RHL) [65], and LabsLand. 3igure 7 encapsulates the MELODY components integrated into the design of RHL-RELIA.An aspect within the Infrastructure layer is the inclusion of 2 ADALM-PLUTO SDR devices.This feature was implemented to provide separate SDR devices for transmission and reception, streamlining the user's programming process.For more detailed technical information, refer to [65], [66], and [67].
The affordability of RHL-RELIA facilitates extensive scalability, enabling multiple stations to operate in a single room.This is why an Isolation block is incorporated into the Infrastructure layer.To prevent interference, a Faraday Cage encloses an SDR transmitter/receiver station, as illustrated in Figure 8.A thorough approach was employed to minimize interference and unwanted elements [68].
To ensure proper RF transmission within the Faraday Cage, it's crucial to maintain a suitable separation between the transmitter and receiver, placing them in the far-field region.This separation depends on factors like antenna dimensions and the frequency range being utilized, as explained in [69].The specific antennas provided with the ADALM-PLUTO device, the Jinchang Electron JCG401, are designed for operation in the frequency range of 824-894 MHz and 1710-2170 MHz [70].According to this information, a minimum distance of 3.60 cm is required to ensure far-field communication.
Furthermore, the implementation of RHL-RELIA empowers students to perform RF experiments from their personal computers via a standard web browser, as exemplified in Figure 9.This figure illustrates a QPSK constellation experiment captured using the RHL-RELIA system.For detailed examples of complete experiments, please refer to [66].

B. RHL-RADAR
The RHL-RADAR project aims to create a new SDR remote laboratory with a specific peripheral which is a fan to test  Radar applications [71].While RHL-RADAR is still under development, its features fit to better explain the MELODY model in this contribution, so we can evaluate MELODY with a future remote laboratory.RHL-RADAR strives to offer students the ease of performing experiments from a distance, negating the requirement for direct manipulation of physical hardware or manual radar parameter configuration.
Within a radar context, a far entity termed the ''target'' comes into play.As such, a potent antenna becomes indispensable to detect for echoes, called backscatter.Moreover, the of a stationary target has been introduced to facilitate the estimation of certain physical characteristics.Visual representation of components derived from the MELODY model can be observed in Figure 10.
For this application, MELODY's architecture is composed by only on SDR unit.This design is driven by the critical role synchronization plays in radar applications, as it enables the Regarding the choice of an antenna, the preference was for an antenna that primarily emits radio waves in a particular direction, enabling efficient energy transmission in that specific path.For this experiment, the selected antenna is the Log Periodic Antenna.This antenna type is characterized by its directional radiation pattern, offering a considerable gain ranging from 9 to 10 dBi.Moreover, it exhibits exceptional adaptability across a wide range of frequencies.Additional details regarding its specifications can be found in [72].
Using this model as a foundation, a prototype was constructed to assess the rotation velocity of a rotating structure, as illustrated in Figure 11.This prototype serves as a foundation for future development of a new remote radar laboratory.

VI. EVALUATION OF SDR REMOTE LABS
Within this section, an evaluation is conducted for each project, namely RHL-RELIA and RHL-RADAR.Both projects are assessed against the MELODY standard, with each being subjected to scrutiny in five distinct categories.

1) EMONA TIMS SDR
• Isolation: The hardware design of Emona TIMS is notably sturdy.While it may not be inherently scalable, the isolation levels are above 40 dB [73].5 stars.
• Lab scalability: One unit per user, not scalable.1 star.
• Lab Availability: Designed for on-campus usage, remote access is not supported.1 star.
• Lab Usage: Device can't be operated by a home computer but campus one. 1 star.
• Lab scalability: One unit per user, not scalable.1 star.
• Lab calibration: Information no available.
• Lab Availability: System can be accessed by a web browser remotely without software installation.5 stars.
• Lab Usage: Students may need to wait for others to complete their usage before accessing the lab. 2 stars.
3) CLUJ-NAPOCA UNIVERSITY • Isolation: Not designed to be scalable and there is no isolation.0 star.
• Lab scalability: One unit per user, not scalable.1 star.
• Lab calibration: Uses ADALM-PLUTO SDR which has been calibrated by its manufacturer Analog Devices [32].5 stars.
• Lab Availability: System can be accessed through GRC. 4 stars.
• Lab Usage: Students may need to wait for others to complete their usage before accessing the lab. 2 stars.

4) HYDERABAD UNIVERSITY
• Isolation: Not designed to be scalable and there is no isolation.Not applicable.
• Lab scalability: One unit per user, not scalable.1 star.
• Lab Availability: System can be accessed through Lab View. 4 stars.
• Lab Usage: Students may need to wait for others to complete their usage before accessing the lab. 2 stars.
• Lab Availability: System can be accessed through GRC. 4 stars.
• Lab Usage: It requires a calendar-based system for the allocation of resources.3 stars.
• Lab calibration: Uses ADALM-PLUTO which has been calibrated by its manufacturer Analog Devices. 5 stars.
• Lab Availability: System can be accessed through a web browser.5 stars.
• Lab Usage: The system allocates a 30-second window for each user to utilize the lab, ensuring efficient rotation.5 stars.

7) RHL-RADAR
Despite being in the developmental phase, the preliminary tests conducted suggest that the results could be included in the evaluation of the MELODY framework.
• Isolation: Not applicable.Radar signal holds significant power that makes effective attenuation unfeasible.
• Lab scalability: The system is limited to supporting a single module.To accommodate additional modules, a transition to the ultra-concurrent mode is necessary.1 star.
• Lab calibration: Uses ADALM-PLUTO which has been calibrated by its manufacturer Analog Devices. 5 stars.
• Lab Availability: System can be accessed through a web browser.5 stars.
• Lab Usage: The system allocates a 30-second window for each user to utilize the lab, ensuring efficient rotation.5 stars.A summary of the rubric assigned to every project in each of the 5 areas MELODY's classification framework is in Table 11.

VII. CONCLUSION AND FUTURE WORK
In this paper we presented the MELODY model, a framework for creating and evaluating SDR remote laboratories.The framework-agnostic nature of MELODY provides developers with the freedom to create applications and solutions that are not bound to specific frameworks or platform technologies.This approach promotes flexibility and portability, enabling the integration of various technologies while maintaining control over the application's design and architecture.The model's classification framework enables structured evaluation, enhancing decision-making by highlighting strengths and weaknesses across various dimensions.Through the use of rubrics covering various engineering aspects, a comparison was conducted among five other SDR remote labs and two implementations using MELODY model.This comparative analysis underscores the technical characteristics of each individual remote lab.The practical implementations, such as RHL-RELIA and RHL-RADAR prototype, exemplify MELODY's potential to adapt to dynamic SDR landscapes.These implementations aim to streamline experimentation by eliminating the need for physical hardware handling or complex parameter configuration.However, it's important to note that the effectiveness of RHL-RELIA and RHL-RADAR falls outside the scope of this paper.
Our future work will be directed towards realizing substantial scalability for RHL-RADAR by advancing its infrastructure to achieve ultra-concurrency as a remote lab.Additionally, a crucial objective will involve seamlessly integrating the RHL-RADAR prototype into the RHL-RELIA network, thereby empowering users to effectively engage with hands-on radar technique topics.Furthermore, the RHL-RELIA application presents opportunities for expansion in terms of interoperability.This entails enabling the inclusion of various SDR hardware units within its framework, offering a more streamlined approach for selecting the appropriate SDR device for specific applications within the same remote lab.Such enhancements will contribute to the broadening of the MELODY model and its classification framework, enriching the metrics available for comparing with other projects.Furthermore, future endeavors related to the MELODY model encompass the creation of Lite implementations, allowing students to access SDR experiments from more affordable devices such as smartphones or tablets.Currently, work is underway to adapt the RHL-RELIA implementation to accommodate this approach [19].

FIGURE 1 .
FIGURE 1. Block diagram MELODY model and implementations.

FIGURE 2 .
FIGURE 2. MELODY architecture and its components.

FIGURE 6 .
FIGURE 6. Measurement of SNR of a QPSK modulation.

FIGURE 9 .
FIGURE 9. Constellation plot of a QPSK communication in a RHL-RELIA transmission.

FIGURE 10 .
FIGURE 10.RADAR architecture and its components.

TABLE 9 .
Rubric -availability of the laboratory.

TABLE 11 .
Classification framework rubric -comparison of 7 SDR lab projects.