Distributed and Analogous simulation framework for the control of pests and diseases in plants using IoT Technology

. In contemporary society, agriculture is progressively embracing technological innovations called Precision Agriculture. The utilization of various pest control and disease management strategies is of considerable importance in the surveillance of plants. The current framework encounters multiple challenges. The pest control and disease surveillance system employs a solitary Graphical Processing Unit (GPU) to manage the diverse array of connected sensors. Hence, this paper proposes utilizing the Distributed and Analogous Simulation Framework (DASF) in conjunction with the Internet of Things (IoT) to address the issue of pest control and diseases in plants. The approach reduces the strain on a specific GPU, effectively allocates the computational tasks across all accessible GPUs concurrently, and ensures continuous data transmission to the dashboards even in the event of GPU malfunction. The implementation of this procedure is anticipated to result in a reduction in overall system performance. In the DASF multi-threading framework, the allocation of tasks to particular auxiliary cores is performed by each GPU unit. The execution of the different functions within this system is allocated among four levels: disease management, pest recognition and control, output operations, and input functions. The data is analyzed concurrently and managed in a proficient and regulated manner. The proposed system demonstrates a significant enhancement in performance measures, with a value of 99.05%.


Introduction to crop and pest control
As a fundamental pillar of global food supply, agriculture consistently pursues novel methodologies to improve efficacy and output [1].In precision agriculture, where technology and cultivation methods intersect, effectively controlling pests and diseases is critical in facilitating maximum crop productivity.The need for efficient control measures emerges due to the substantial economic losses and environmental consequences of unregulated pest and disease outbreaks [2].Based on current agricultural research, the detrimental impact of pests on crop output leads to substantial losses, varying between 20% and 80%, depending upon the specific crop type and geographical location [3].
Conventional methods of pest and disease control encounter significant obstacles [4,5].Using historical weather data and periodic information sources often results in prolonged reaction durations, impeding the efficacy of control techniques.The increase in worldwide commerce and the impact of climate change have had a significant role in creating novel pest and disease risks, necessitating a more flexible and responsive strategy [6].The existing methods, which heavily rely on human monitoring and intervention, need help in effectively managing the growing complexity and size of contemporary agricultural landscapes.
Integrating Internet of Things (IoT) technology is a revolutionary approach to tackling these difficulties [7].The IoT introduces a significant change by facilitating the real-time monitoring of agricultural processes and making informed decisions based on data.Based on industry surveys, the worldwide agriculture IoT market is anticipated to attain a value of $15.3 billion by 2025.This projection highlights the growing acknowledgment of IoT's capacity to bring about transformative changes in agricultural methodologies.Incorporating sensors, actuators, and communication devices into the farm ecosystem facilitates accurate and prompt data collection, establishing a basis for enhanced pest and disease management approaches.
To underscore the significance of pests and diseases in global agriculture, on average, more than 20% of crop output is yearly compromised due to these factors [8].This translates into substantial economic losses amounting to billions of dollars.The losses have a dual impact, as they not only have adverse effects on the economic sustenance of farmers but also serve as a contributing factor to the issue of food insecurity in several places.The limitations of current methodologies are apparent in their restricted capacity to effectively forecast and preempt pest and disease epidemics.Traditional methods often depend on responsive actions, resulting in an escalated dependence on chemical treatments that adversely affect the environment.The primary contributions are listed below: • This research introduces the Distributed and Analogous Simulation Framework (DASF) integrated with the IoT to improve pest and disease management in Precision Agriculture.• The objective is to mitigate the burden on a single Graphical Processing Unit (GPU) by distributing computing jobs among many GPUs.
• The present study focuses on integrating a multi-threading framework into the DASF.This integration aims to enhance the efficiency of job allocation and processing across four distinct levels, including disease management, pest detection and control, output operations, and input functions.The following sections are arranged in the given manner: Section 2 provides an overview of the current body of research and existing knowledge in the literature review.Section 3 introduces the DASF as a proposed pest and disease management approach.Section 4 of this paper comprehensively examines the experimental analysis and results of implementing the proposed DASF framework.Section 5 presents the concluding remarks and delineates potential avenues for future exploration, drawing upon the study results.

Literature survey and findings
The literature review thoroughly comprehends the present scenario and explores the current state of research, methodology, and technical breakthroughs in pest and crop management.The present study examines previous research, advancements, and obstacles to provide a solid groundwork for the suggested investigation.
The study conducted by Alarcón-Segura et al. examined the use of strip intercropping between wheat and oilseed rape as a strategy to augment biodiversity and promote biological pest management within conventionally managed agricultural systems [9].The authors proposed the Strip Intercropping for Biodiversity and Pest Control (SIBPC) technique, demonstrating its ability to augment biodiversity by 22% and improve biological pest control by 37%.This research showcases SIBPC as a viable and sustainable agricultural technology.
The study by Zheng et al. investigated using metal oxide semiconductor sensors in electronic noses to identify and diagnose agricultural diseases and insect pests [10].The suggested methodology, known as the Metal Oxide Sensor-based Electronic Nose (MOSEN), demonstrated a notable level of precision in identifying diseases and problems, surpassing an identification rate of 90%.The empirical findings confirmed the efficacy of MOSEN in the early identification of pests and diseases, making a valuable contribution to the timely management of crops.
The study conducted by Wang et al. centered on the examination of environmentally sustainable nanoplatforms designed for crop quality management, protection, and nutrition [18].The suggested Eco-friendly Nanoplatforms for Agriculture (ENA) technique exhibited notable characteristics such as precise nutrient targeting and regulated release.The experimental results demonstrated a significant 25% enhancement in crop productivity and improved nutritional composition.These findings underscore the promising prospects of using ENA for sustainable and efficient crop management.
The study by Ratnadass et al. explored the implementation of crop protection strategies and the associated zoonotic hazards of virus transmission, with a particular focus on the One Health concept [12].The authors proposed a comprehensive plan called One Health Crop Protection (OHCP), which highlights the interdependence of human, animal, and environmental well-being.The experimental results highlighted the effectiveness of OHCP in mitigating viral zoonotic threats by a significant 30%.These findings provide valuable insights into implementing comprehensive and proactive approaches to safeguarding crops.
The study conducted by Consolação Rosado et al. investigated the effects of intercropping cover crops on the biological control of pests in coffee plantations [13].The suggested approach, known as Cover Crop Intercropping for Biological Control (CCIBC), has shown a noteworthy augmentation in indigenous predators, leading to a substantial reduction of 40% in coffee crop pests.The research emphasized the potential of CCIBC as an ecologically viable approach for augmenting biological control in coffee agroecosystems.
The study conducted by Charbonnier et al. examined the efficacy of bats in providing pest control services within grape landscapes [14].The Bat-mediated Pest Control (BPC) demonstrated that the presence of bats resulted in a 25% reduction in the grape pest population, highlighting the ecological significance of bats in vineyard ecosystems.The research emphasized the importance of preserving biodiversity, with a specific focus on the contribution of bats, within the context of integrated pest control approaches in vineyards.
The study conducted by Wei et al. centered on categorizing agricultural pests using a multi-scale feature fusion technique [15].Multi-Scale Feature Fusion for Pest categorization (MSFFPC) exhibited exceptional accuracy in categorizing pains, with an impressive overall precision rate of 94%.The research emphasized the effectiveness of MSFFPC in using multiscale characteristics to accurately and dependably identify agricultural problems.The literature review emphasizes the difficulties associated with conventional pest and crop management approaches, such as the prolonged response time to issues and the restricted capacity to address developing risks.As a result, there is a need for inventive and flexible solutions [17].The suggested methodology, DASF using the IoT, effectively tackles these difficulties by offering a real-time and multi-threaded strategy for managing pests and diseases in precision agriculture.This technique guarantees prompt and adaptable reactions to ensure optimal agricultural outcomes [11].

Distributed and Analogous Simulation Framework
Accurate data is obtained using various sensors such as meteorological, rainfall, sediment, and pH detectors.Historical data is acquired via the use of periodic data sources.This dataset comprises responses obtained from local producers and large-scale data sources.The needs of specialized fields are classified, identification methods are used, and a practical algorithmic approach is introduced to classify outputs.Using statistical processes or machine-learning systems in an agriculture-centric economy has significant potential for providing substantial benefits to farmers.This strategy can benefit both production improvements and the use of suitable agricultural practices.
Both precision farming and phenotyping of plants face distinct requirements and obstacles when it comes to the diagnosis of plant diseases.Developing and integrating novel methodologies into conventional monitoring and grading systems is essential for accurate and dependable automated diagnosis and detection of plant diseases.Optical sensors have considerable potential as effective instruments for the distributed detection and diagnosis of many diseases.The availability of imaging and distributed devices for diagnosing and detecting plant illnesses has steadily grown.The advancements in sensor and database technology and the widespread use of the Global Positioning System (GPS) have created novel prospects for precision farming and crop phenotyping.

Pest and Disease Management in Precision Agriculture
Various forms exist for the collection of information.Data collection methods such as weather monitoring, precipitation measurement, soil analysis, and pH sensing are used to get precise and reliable data.Standard data sets are used to gather historical information.The dataset includes feedback from local producers, constituting a comprehensive compilation of information often called big data.This study presents a suitable clustering approach for categorizing findings, utilizes identifying gadgets, and classifies various field requirements.Integrating extensive statistics or machine technologies in an agricultural sector is expected to provide significant advantages for farmers.This approach could enhance sustainable farming practices and facilitate higher agricultural productivity.
The use of artificial intelligence enhances the advantages of the remote control.Wireless communication-enabled IoT devices can detect many field elements, including but not limited to relative humidity, humidity, and pH levels.Various sensors, such as thermometers, precipitation gauges, and soil temperature probes, are used to analyze real-time atmospheric information.Standard databases are designed to gather past details on weather conditions and moisture levels.The system design of the DASF with IoT Assistance is shown in Figure 1.Cultivation practices include using conventional statistical methods and tools to acquire spatial information acquiring both organized and unorganized data sets.The climatological service that is systematically structured comprises data about weather and climate.The data obtained from heating components, precipitation detectors, and soil precipitation needs more organization and framework, resulting in an unstructured report.The methods include a simulation platform enabling agricultural produce administration via Cloud Computing and Information and Communication Technology (ICT).The DASF framework utilizes a multithreading paradigm involving work distribution across many GPU cores.Each layer is responsible for managing specific crop-related duties, pest control, and administering input and output activities.The method has four stages whereby information is efficiently, systematically processed, and organized.The findings from the analysis of simulations demonstrate that the DASF integrated with IoT assistance exhibits effective capabilities in the classification and management of pests aided by IoT technology.
Due to its lack of organization, analyzing large data sets presents a formidable task.In this movement, data analytics plays a crucial role in identifying relationships among various crop restrictions and providing valuable insights for the development of future agricultural technology.All collected data is saved inside a local cloud architecture.A corresponding solution is generated when family-owned agricultural enterprises need a web-based service.These two characteristics exemplify the broad range of plant variety from various perspectives.To maintain the development of agriculture, each index was used.The beginning Hypothesis 2 imagery groups were based on the calculation described in Equation ( 1): The near-infrared (  ) wavelength is used to identify and characterize the coloration of plants.  is the green wavelength used to discern seedlings with green foliage.  is used as an indicator of the strain condition and pigment content of plants to demonstrate variations in the number of plantations in vegetated locations.These two characteristics exemplify the extensive range of plant species seen from certain viewpoints.The variable representing the red edge dispersal of the plant is   , as shown by Equation (2).  is used to identify and characterize the coloration of plants, whereas   is utilized to discern seedlings that possess green foliage.Acquiring atmospheric data is the fundamental basis for monitoring crops and pests.The practical cap converter has selected two variables: moisture content and weight.The term "wet" pertains to the thermal conditions inside a body of water, whereas "lushness" refers to the quantification of the collective vegetation required to delineate the attributes of an ecosystem in particular.The variable "W" in Equation ( 3) denotes the quantity of moisture inside the agricultural region.
The translation parameters ( 1 ) and ( 6 ) are extracted from a specific subfolder that houses the captions of the movie.The weights are indicated as   .The mean temperature suggests the rate of photosynthesis in crops and the circulation patterns, frequency, and severity of pests and diseases that impact crop health.The sensors' obtained output is integrated into the twenty spectral categories and the Landsat picture processor.
Without a crop phylogenetic information-gathering structure in specific areas, counties, or ecosystems.The storage of data often involves the use of distinct regulations, database assets, and system architectures.The current focus of plant phenotype data storage is mostly on cloud-based software platforms.Cloud infrastructure comprises a cohesive assemblage of methods, storage facilities, machines, offerings, educational endeavors, user control methodologies, and increasing societal recognition.This interconnected system encompasses both conventional storage techniques and facilitates seamless communication.Figure 2 illustrates the architectural design of the DASF structure with IoT capabilities.

Fig. 2. DASF architectural design
The study on the functional genomics of crops included four essential elements: cognitive procedure, storing data, phenotypic features, and predictive facts modeling.Various methods are used to gather data, including several forms of imagery, spectroscopy, meteorological Data storage technologies often utilize many procedures, databases, and system designs.The storage of plant phenotypic information is contingent upon the availability and reliability of power supplies.Cloud facilities, an increasingly prevalent alternative to conventional capacity and consumer control techniques, encompass a fusion of fluid structures, retention mechanisms, gadgets, apps, interaction, and qualitative investigation tools.The hardware components are categorized into three distinct layers: a transmission protocol level, a physical level of the software, and a database layer.The fundamental need that several methodologies must fulfill is the core network.The centrally controlled architecture refers to a collective cloud storage technique encompassing the control of hard drives, handling passwords, dispersion, and overall functioning.
The DASF with IoT integration incorporates multi-threading and real-time data processing to enhance the effectiveness of pest and disease management in precision agriculture.The proposed architecture aims to improve the efficiency of computing activities by distributing the workload over numerous GPUs, hence mitigating the burden on individual processing units.Advanced data transmission and assessment techniques in the DASF system result in a complete and highly responsive solution, showcasing a notable performance improvement.

Simulation modeling and analysis
The proposed study's experimental configuration entails using a GPU cluster, especially using a minimum of four NVIDIA Tesla V100 GPUs.The GPUs are linked via high-speed communication interfaces such as NVLink, which enables effective job allocation and parallel computing.The minimum required RAM for the simulation environment is 128 GB, and data transfer is improved by using a 10 Gbps Ethernet network.The distributed and analogous simulation framework has been successfully developed with CUDA, enabling efficient coordination among the cluster's GPUs.information transmission across several methodologies, where larger values signify enhanced operational proficiency.The average data transmission rates attained by SIBPC, MOSEN, ENA, OHCP, CCIBC, BPC, MSFFPC, HGSDCNN, and DASF were 24.42 Gbps, 20.94 Gbps, 27.33 Gbps, 31.35Gbps, 18.87 Gbps, 21.28 Gbps, 23.44 Gbps, 22.05 Gbps, and 44.34 Gbps, respectively.The DASF approach continuously shows higher performance compared to other ways.

Fig. 4. Computational efficiency analysis of crop and pest control
The Computational Efficiency (jobs/GPU) metric's findings are shown in Figure 4, which shows how many computational jobs each GPU can handle.This statistic evaluates how well the various approaches distribute and carry out tasks.Computational efficiency was 51.77 tasks/GPU on average for SIBPC, 53.56 tasks/GPU for MOSEN, 46.45 tasks/GPU for ENA, 60.93 tasks/GPU for OHCP, 40.84 tasks/GPU for CCIBC, 47.59 tasks/GPU for BPC, 53.79 tasks/GPU for MSFFPC, 44.76 tasks/GPU for HGSDCNN, and 71.77 tasks/GPU for DASF.With an average of 71.77 tasks/GPU, the suggested DASF technique continuously showed higher computational efficiency, demonstrating its efficiency in allocating and efficiently computing jobs on available GPUs.  11.39 ms).The DASF technique, as described, consistently showed better performance in terms of Response Time, with an average of 11.39 ms.This suggests that the method is efficient in swiftly addressing tasks within the precision agricultural framework, surpassing other methods.

Fig. 6. Resource utilization analysis of crop and pest control
The findings of the Resource Utilization (%) statistic, which indicates the proportion of system resources consumed by each approach, are shown in Figure 6.This statistic evaluates the efficiency level with which the methods use the resources that are accessible to them.The average Resource Utilization percentages for the following systems were observed: SIBPC (71.62%),MOSEN (73.2%),ENA (66.06%),OHCP (82.02%),CCIBC (60.68%),BPC (65.64%),MSFFPC (75.64%),HGSDCNN (64.43%), and DASF (55.03%).The DASF technique regularly demonstrated excellent resource utilization in the context of precision agriculture.

Fig. 7. Computational accuracy analysis of crop and pest control
Figure 7 presents the outcomes of the computing Accuracy (%) measure, which quantifies the level of accuracy attained by each approach in their respective computing jobs.Greater values indicate more precision in computational results.The average computational accuracy for SIBPC was found to be 63.45%,MOSEN 64.44%, ENA 58.66%, OHCP 71.87%, CCIBC 52.19%, BPC 56.23%, MSFFPC 64.21%, HGSDCNN 53.89%, and DASF 79.16%.The DASF approach, as described, consistently showed a higher level of Computational Accuracy, averaging 79.16%.This suggests that the DASF method can provide more precise computational results within precision agriculture, surpassing other methods.
The DASF performance evaluation demonstrated exceptional results across multiple metrics.These metrics include an average Data Transmission Rate of 44.34 Gbps, Computational Efficiency of 71.77Tasks/GPU, Response Time of 11.39 ms, Resource Utilization of 55.03%, and Computational Accuracy of 79.16%.The results demonstrate the effectiveness and superiority of the DASF approach in enhancing data transmission, computational efficiency, responsiveness, resource usage, and computational accuracy within the simulated precision agricultural framework.

Conclusion and future scope
The agricultural sector is crucial in maintaining worldwide food supply, underscoring the need to tackle issues such as pests and diseases that substantially influence crop productivity and quality.The implementation of precision agriculture, which encompasses efficient pain and disease management tactics, plays a vital role in promoting sustainable and efficient agricultural methodologies.Given the constraints of current frameworks, it is crucial to use IoT technologies to improve monitoring and control mechanisms.The Distributed and Analogous Simulation Framework (DASF) is introduced as a response to the matter.The DASF efficiently tackles the issues associated with centralized GPU-based systems by distributing computing duties over numerous GPUs, facilitating continuous data transfer, and alleviating pressure on individual GPUs.The key characteristics of the DASF system include multi-threading, a four-tier job allocation framework (comprising disease management, pest detection and control, output operations, and input functions), and the facilitation of practical data analysis.The outcomes of the suggested approach exhibit its superiority, as shown by an average Data Transmission Rate of 44.34 Gbps, Computational Efficiency of 71.77Tasks/GPU, Response Time of 11.39 ms, Resource Utilization of 55.03%, and Computational Accuracy of 79.16%.
It is essential to recognize some limits, including possible limitations in hardware capabilities, the need for effective cybersecurity protocols, and the capacity to adapt to various agricultural settings.Future study prioritizes the enhancement of DASF for practical deployment by integrating developments in machine learning techniques for predictive modeling and extending its applicability to include a more comprehensive array of crops and geographical locations.These initiatives will significantly advance precision agriculture, promote sustainable farming methods, and effectively manage emerging difficulties within the agricultural sector.
Sanghavi et al. (2023) proposed using the Hunger Games Search-based Deep Convolutional Neural Network (HGSDCNN) in agricultural pest detection and classification using transfer learning techniques[16].The HGSDCNN model demonstrated exceptional performance, with a remarkable accuracy rate of 96% in pest detection.The research showed the possibility of integrating search-based optimization and deep learning methodologies to enhance the effectiveness and efficiency of agricultural pest control strategies.

Fig. 3 .
Fig. 3. Data transmission rate analysis of crop and pest control Figure 3 depicts the outcomes of the Data Transmission Rate (Gbps) measure, which quantifies the speed at which data is sent.The statistic quantifies the effectiveness of

,Fig. 5 .
Fig. 5. Response time analysis of crop and pest controlThe results of the Response Time (ms) metric are shown in Figure5, illustrating the duration it takes for each technique to respond to a particular activity.Lower numbers indicate faster reaction times and improved system performance.The average response times for the following systems were observed: SIBPC (21.86 ms), MOSEN (30.28 ms), ENA (20.57ms), OHCP (26.5 ms), CCIBC (19.27 ms), BPC ms), MSFFPC (23.13 ms), HGSDCNN (20.81 ms), and DASF (11.39 ms).The DASF technique, as described, consistently showed better performance in terms of Response Time, with an average of 11.39 ms.This suggests that the method is efficient in swiftly addressing tasks within the precision agricultural framework, surpassing other methods.