Myself, I'm Sir Ramuda. And and. Sorry myself, I'm. And I completed my B.Tech from Rastan Techniques in the college affiliated from Rastan Technical University RU. Kota and then M take from. LPU Jalandhar. And then I worked in some some of the project research project. At one time I worked at the Chennai, another time I worked at the Delhi. So in several project I have worked. And then?

I started my PhD. Uh.

Thank you for sharing that, Professor.

Understood. Before we delve into your research in detail, Professor, could you describe a foundational topic in multimedia or artificial intelligence that you find particularly engaging to teach to undergraduates?

Well, particularly, you know. Computable into teaching core subject. Any one of the core subject. In you know which can be in the.

Artificial intelligence.

You mentioned artificial intelligence as a core subject—can you walk me through a specific concept or module you typically emphasize when introducing AI to new students, and explain why you prioritize it in your curriculum?

Sure.

See, currently I'm working at the Adani city Ahmedabad. So here I'm teaching the data structure.

Thank you. When you introduce that AI concept to students, what sort of real-world example or project do you use in the classroom to help students grasp its relevance?

You mentioned teaching data structures at Adani City, Ahmedabad—can you give a concrete example of how you integrate AI concepts into your data structures course to make the material more relevant for students interested in artificial intelligence?

Now how we can use the AI means the artificial intelligence into the data. Or any of the course? So that. Depends upon the you know also. And it it varies from coast to coast and also. Umm, depends upon the. And the Esplanade, Scaliber and the potential and so on so. Current the the the artificial intelligence can be included in the data structure for suppose different different data structures are there to organize and represent the data. So how efficiently we can, you know, use the? Data structure by utilizing the AIE. To organize and manage the data.

Now it enforces, you know. Time being they consider the data structure like stake or queue or linked list or tree or graph so. Whichever the data structure is there. They have some there, you know, efficiency. I mean like time efficiency and the space efficiency. So how we can? Improve it or how we can. Make it more efficient by utilizing the AI into it.

You mentioned using AI to organize and manage data structures—can you provide a specific classroom assignment or activity where students apply an AI approach to optimize or select data structures, and describe how you assess their understanding?

Thank you, Professor. When you ask students to optimize or select data structures using AI methods, what criteria or evaluation metrics do you have them consider to assess the effectiveness of their solution?

Because we won. To do the use the utilize the AI into it. Reason is that. Like this is the AI era. So we we have to go with the time.

As much as possible or wherever is the possible, we can use AI into. In the course.

You emphasized integrating AI into data structure coursework—can you give a specific example of a project or classroom activity where students apply AI techniques to improve or select among data structures, and explain how you assess their learning from that exercise?

You emphasized integrating AI into data structure coursework—can you give a specific example of a project or classroom activity where students apply AI techniques to improve or select among data structures?

And did suppose a problem is given on the like. And linked list. Support How to find the cycle in a linear linked list? Lincolnshire is already given to them. Suppose problem is that. That you have a linked list and we want to find the a. Circular or a cycle? In the given linked list now.

I have told them suppose one way. To do. To solve this to find the to find the cycle in the singular linked list now. To grab the other ways they started to. Use the, you know, like. ChatGPT or something? To you know how they can find the. Uh, multiple solutions or some different alternative solutions for the given problems? For a suicide problems and also later on by comparing the time complexity that which is more efficient. Or. And we? Compared to, you know, all others. And also. That how if something you know 11 solution is. And lacking in. If you can see let's say for the time efficiency, then.

Right, so with your linked list cycle detection example, how do you guide students to evaluate which AI or algorithmic approach is most efficient for that problem, especially considering practical constraints like time and space complexity?

Is it feasible or is it? And durable to make it more efficient.

Thank you, Professor. When you incorporate tools like ChatGPT into student assignments for data structures, how do you ensure students develop their own critical problem-solving skills rather than relying solely on generated answers?

Or a compared to other solutions.

You mentioned students using ChatGPT or similar tools to find alternative solutions—how do you ensure they critically evaluate those AI-generated answers and understand the underlying algorithmic principles, rather than just relying on the output?

Because the the the fundamentally there can be different different solutions, but. To compare all of them. By using the. And based on the efficiency, specifically time efficiency, we can make some inferior that. This particular solution is efficient compared to others and so on, so it's not. That they are free to use. Any any any any a specific tools? But they have to. Let us know that. AI tool or something they are using? And at what? Level they are using. Why they are using? They means. And how did you they find that which tool is OK or not and so on? See two, you know. And bring it, uh. The notice or in the mind that. OK. This is the way they are using.

Understood. When you assign this type of problem—such as detecting cycles in linked lists—how do you evaluate whether a student’s approach demonstrates genuine understanding of both the data structure and the AI tools they apply?

Given your approach, how do you assess whether students are genuinely learning to select and justify appropriate data structures when AI tools provide them with ready-made solutions?

So that's why I told you no that. To just.

If they are using fine but. At which parameter? Hey at at what? You know a specific sub, you know activity or a particular task in the given problem or a subjective or subtask? They are using.

Understood. Shifting slightly, Professor, could you share an example of a student project you have supervised where your industry experience directly influenced the project's direction or outcome?

And why they are using? And at what? You know, level or at what I would say instance they got stuck. So they. Required to use the AI tool or something? Because that's why. So if they. Use the AI tool then they they can that OK, we use the AI tool for at this instance or at this particular time. In this problem or for this problem? And then we. Find the you know. This is the. Uh, you know, once we're stuck, then we in in the dialogue, then we find that OK, we come out from the deadlock or is stuck from by using this. So listen also sometime in the, in the, in the, in the lab or some. To to ask them. Do a hands in experience by using AI. Tools. Uh. For their problems so they can. You know, they can learn how efficient sometimes sometime if we have the. You know, adequate time or something. By using, you know, AI tools also. By letting. Them know how prompt engineering can be used. To make the.

Let's shift focus to your research. Professor, can you outline your three-year research roadmap—including the main themes you intend to pursue, anticipated outputs, and strategies for securing external funding to sustain your scholarly activities?

Efficient utilization of the AI into the. Any over the course for for in my case it.

You described guiding students on prompt engineering and using AI tools when they're stuck—how do you evaluate the quality of their problem-solving process versus their final solution when AI assistance is involved?

Understood. Looking ahead, could you outline your three-year research roadmap—specifically, what main themes you plan to pursue, what outputs you anticipate, and how you intend to secure external funding or industry partnerships to support your work?

See, in my PhD I work the.

You mentioned your PhD work—can you describe how you plan to leverage your doctoral research as a foundation for obtaining grants or building collaborative industry projects in your next three years?

Trying to get or find that. How we can use some advanced technology like? AI even though. At say till some extent to be used the. And some ML mechanism in my work.

But also how we can use the you know? Other, you know, like AI make advanced technology like any any any AI or a computer reasons and you know. And and ML, whatever the case will be there so but TLX. So, you know, some extent or till some extent or certain extent, we use the computer vision. And the AI both in my work.

You mentioned using some machine learning mechanisms in your current work—can you specify which ML models or techniques you've applied so far, and how you measured their impact on energy efficiency in your wireless sensor network experiments?

So I would like to pursue. If possible, wherever and by getting to get the external funding. For for my research or my. In project, whatever I would like to do, I will try to collaborate. I would like to. Pursue to collaborate with the my.

You noted using computer vision and AI in your work—can you give a concrete example of how you integrated computer vision techniques into energy-efficient communication for sensor networks, and what measurable improvement resulted from that integration?

Cliques in the department or outside the department? Within the Institute. Or also outside the institute by trying to get. So we can get the collaboratively we can. Pursue the. For to seek to acquire this grant from the external agency like DST or Sir or CSR, whatever.

Professor, for a large undergraduate course, how would you implement an active learning model—such as a flipped classroom—without relying on traditional lectures or slides? Please describe the specific activities and structure you would use to maximize student engagement.

Let's shift focus to departmental metrics. Suppose your department needs to improve its “Research Output per Faculty” to meet an upcoming accreditation review. What systematic steps would you take to help the department achieve this goal?

OK, the first thing the the the engagement, the faculty's engagement in the research. So the the for that there should be some balanced day. You know, like approach. Especially for the workload, you know it, it cannot be like that, that we are doing the. A more teaching load. And so if they have, the more teaching load, if any faculty has the more teaching load. Then he or she might not be able to contribute. Or able to provide they'll give. More time. To the I would say the research. So if we approach. If we. Prefer or if we adopt the balanced approach, the adequate teaching load will be there. So he or she can pursue his research.

In the in his or her, you know, I would say the domain or area wherever he or she wish to pursue. And so if he or she. Pursue, then he's can. Provide the good research papers. Also, there should be some incentives mechanisms or some. Some some something should be there at the department level. For if the faculty has the good. Published the paper then. APC will be provided to the if the. If the journal is Open Access, then they. APC will be bound up. You know it will be will be taken care by the department itself. If the paper is published in the conference, then all the registration fee or travel all the etc and they. The expenses will be bound by the department cell. Itself also. By because it's, you know, like a. Should not be like a one man army. It's a collaborative, so students should also be getting participated into the with the faculty into their so they can also get some hands free experience or so in in the team or in a combined manner they can work and they can pursue.

You emphasized balancing teaching and research loads—what specific mechanisms or policies would you propose to track and ensure faculty are meeting agreed-upon research targets without compromising instructional quality?

So it ultimately improved their. Performance as well. Even though they are getting the degree but. With some project or research project or. In in what? In whichever the project they would like to continue, and also for the faculty also. Because he has some. We take a question on term, take PhD, accept the PhD. Also some B take questions should be there.

You emphasized balancing teaching and research loads—what specific mechanisms or policies would you propose to track and ensure faculty are meeting agreed-upon research targets without compromising instructional quality?

So he or she can. Continue with them and. Uh, the things can be improved further. Depart.

Now, Professor, imagine a situation where you are accused of grading bias in a course, but adjusting grades would conflict with the university’s minimum pass-percentage targets. How would you handle this conflict between fairness in assessment and institutional pressure to increase pass rates?

You mentioned that departmental support for conference costs and student involvement can boost research output—how would you ensure fairness in distributing these resources among faculty with different levels of research productivity?

And chances for each and every faculty. It should not be that some of are getting the. High number of the resources others are getting the. Least number of the resources based on their seniority or junior to your experience, whatever. Yeah, if somebody is getting the more external funding that. Completely depends upon him or her how his profile is good or how he's trying. Putting the effort, more effort to get the external funding.

But from the within the. Department or within the Institute if. Uh, he or she should get the equal number of the, you know, share from the available resources. Which can be the journal access. You know, you know Open Access because now the world is. Approaching towards a open exchange channel, so. It should be some.

Thank you for clarifying your approach. Now, can you describe a concrete strategy for leveraging your professional network to create internship or employment pipelines for students, specifically within the media or AI in media sectors?

Freely available, open exchange and limited. In A, whether it's I, triple E or ACM. Or the OR springs. Or whatever the publishing house will. Will be there. It's freely and also like, you know, some kind of the. That library access or book access also be available that OK.

Now, let's discuss your expertise in multimedia or AI in media. Can you describe a specific project or research application where you directly applied AI techniques to solve a media-related problem?

And like before the start of the next semester, they. The requirement for the books can be asked to the faculty that in in which courses you are teaching or he or she can let us. And let the library know the concerned person know that I am teaching this and this course. Then I'm requesting or I'm I might require this and these books for this and this course so. All the pre requirement can be acquired from the consent department.

Okay. You mentioned open access to journals and resources—how would you handle situations where limited departmental funds can't cover all the faculty requests for publication or conference support?

By the sod or some of the concerned person and that can be forwarded to the library or wherever they. So in this similar manner we can apply for the TSF. You know, we can know the how much? Effort. I would say the working teaching loaded faculty has so how many number of Tasers is required based on the course? Or the number of the student enrolling the course, all these things. Gonna be so in if in advance all these can be. And you know, I would say. Can be acquired or can you know can information can be get? From all these and. Then the things can be formalized or formulate to know they. Things in advance. Before things getting started.

Right. Earlier you mentioned involving students in research projects to enhance both their and faculty outcomes—can you give a specific example of how you’ve structured undergraduate research involvement in the past, including how you balance their training needs with project deliverables?

Okay. Earlier you mentioned integrating AI and computer vision into your research—can you describe a specific instance where you applied machine learning to improve energy efficiency in WSN-based IoT environments?

See in the, in in the we use the clustering mechanism. To improve the energy efficient communication for the devolution. So the point is that in the literature study we found there are lots of the. I would say. And the parameter which can affect the. Energy efficient. Communications. While we are forming the cholesterol, making the cholesterol in clustings for the network. Now the point is that how we can make. So there can be N number of the parameters.

Now if we use the old, there can be some trade off between that how we use the, you know the. You know there should be some decisions are required. Which parameters should be used, which old? Would not be used, so in my we we we we decided several kilometres. And then be some importance has been assigned to all these parameters by using the MCDM methods.

So when selecting parameters for your clustering mechanism, how did you determine which features to prioritize for the machine learning model, and how did this choice impact overall energy consumption in your network?

And for then.

You mentioned assigning importance to parameters using MCDM methods—can you give a specific example of how you validated that your parameter selection led to measurable improvements in network energy efficiency?

Interesting. When you applied MCDM to weigh those parameters, how did you ensure your machine learning model generalized well to new WSN deployments, not just your training data?

If the VBB simulated the network. And in networks we placed some nodes. And on nodes we run the DBSCAN algorithm to form the cluster.

In that we have the nodes. Now we want to form the clusters. We want to form the group of the nodes. Based on their.

So network we have.

You mentioned using DBSCAN for clustering—what challenges did you encounter when tuning DBSCAN’s parameters for real-world WSN data, and how did you address them?

You mentioned running DBSCAN on your simulated WSN—how did you evaluate whether your clustering improved the network’s energy efficiency compared to baseline approaches?

Like the number of notes? Should be in the minimum number of nodes. Should be in the in the cluster or the cluster radius. Based on these two parameters. Because the DB scan requires that these two parameters only at the initial stage, so we use the Debiscan algorithm. To form the clusters once the cluster formed. Then by using the MCDM. Method we selected the cluster heads.

You said you assigned importance to parameters with MCDM methods—can you describe exactly how you measured the resulting improvement in network energy efficiency after clustering, and what baseline you compared against?

So normally what happen while we are using the AI or any ML techniques, they require lots of the you know. Labeled unlabeled data. To train their model, all these things also. If they might require some. And trial and error approach. So ultimately in our in in my case the doubleson had already the resource constraint means. They are the way trip operated devices. So if we. We might not have the sufficient data. And even though if we have. Then the uh. Battery depleted frequently. We do not want to eat so that's why we use the DB scan, so it requires the less. Initial knowledge about the network. And these? These two things are decided by the the initial stage that. What should be the degree? What should be the number of nodes in the cluster? And what should be the? And I would say radius. That maximum distance of the clusters so we can. Based on these parameter. The cluster formation. Can happen.

You mentioned using DBSCAN for clustering and MCDM for cluster head selection—how did you validate, in your simulation, that these choices actually reduced the overall energy consumption compared to traditional clustering approaches?

Okay. How did you ensure that your simulation using DBSCAN and MCDM actually reflected real-world energy consumption patterns in deployed WSN-based IoT systems?

See MCDM is used for the multiple decision making. Because we have the multiple parameters, so one way is that we can assign the weights. In a fixed way, Certain fixed weight. You know, like to. Multiple parameters will have. Suppose there are some 5 parameters, Suppose P1P 2P3. 5:00 So we can extend some fixed weight.

To a parameter P1 then 2A parameter P2 then. Parameter theory and so until P5. Now if we use this approach. Then there can be, you know, maybe that. Might work for the in one environment or one domain or for one applications. So that we do not want, we want to make it. You know the applicable for.

Understood. When you assigned fixed weights to parameters in your simulation, how did you determine which parameters were most critical for energy efficiency and which could be deprioritized?

Based on the requirements means the based on the domain or applications.

Because in some domain or application. Parameter P4 might have the more importance than P1. In another P2 might have the more importance. Than other. So these are the. Based on the case to case. So for the 2 resemble. These things. We use the MCDM. So because based on the network or based on the requirement. We use the in the. The the importance can be importance can be assigned to the parameters.

Okay. When you switched from fixed weights to more adaptive weighting in your MCDM approach, how did you validate that this generalization actually led to better or more consistent energy efficiency across different WSN scenarios?

Right, so when you adapted your weighting approach based on domain or application requirements, can you give a concrete example of how you tuned those weights for a specific scenario and what impact it had on the simulation results?

Now the DB scan as I said. That was we. We might not have. In the multiple data labeled unlabeled unlabeled data for the. To train the models. First thing and even though if we have, we do not want to make the. We do not want more energy consumption for the training purpose only so. We use the DB scan only, so it requires the two basic. Parameters initially. The first is the number of nodes in the.

Thanks for clarifying your approach to adaptive weighting. Shifting focus a bit, can you walk me through your typical process for evaluating students during exams or practicals—what steps do you take to ensure fairness and consistency across different batches?

You explained how energy and data constraints led you to choose DBSCAN over more data-hungry AI models—can you describe a specific difficulty you faced when tuning those two key DBSCAN parameters for your network, and how you resolved it?

It's going in for the training purposes, all these things, and they are targeting only the energy efficiency.

Here also so we use the MCT, the DB scan along with the MCDM. To make it more energy efficient. Compared to the. In the already existing sum.

You mentioned that many approaches in the literature focus only on energy efficiency—did you explore or measure any trade-offs, such as network latency or data delivery reliability, when prioritizing energy optimization in your clustering approach?

So we can because MCDM is used for the multiple parameter selection. We have the multiple parameters so based on multiple parameter which. Parameters would be more. You know, say.

Right, so you emphasized minimizing energy used for model training in WSNs—how did you measure or estimate the battery savings achieved by using DBSCAN and MCDM compared to a traditional supervised learning approach?

Import important for the particular environment or the applications. Now these parameters can be the distance. These can be raised to energy of a node. Distance from the base states. To the node or average distance from the among the cluster members or within a cluster and the cluster had. Count how many times you know to become the cluster head and all these things. There are several parameters which can be.

Understood. Let's shift focus—how do you typically structure your lecture when teaching a foundational multimedia theory course to undergraduates, especially for students without a strong technical background?

Consider into. It so it resembles. Is there real life environment? Because all these parameters affect the. I would say that. Cluster head selection. In the cholesterol algorithm. For the West and basic source constraint, I.

You mentioned considering parameters like residual energy and distance from the base station when selecting cluster heads—can you describe how you would explain the reasoning behind multi-parameter selection to students who are new to wireless sensor networks?

You mentioned that parameters like residual energy, distance to base station, and cluster head rotation frequency make your simulation more realistic—can you give an example of how you validated that your chosen parameter set actually matched or predicted behavior in a real WSN deployment?

OK. And did the. You know, sensitivity analysis. I mean for two, you know, by changing some. Uh values of the parameters or because in MCDM also we that if we use this method. Then. And what should be the? In the parameter which has the. The I would say the importance so by doing some sensitive analysis, we.

Got it. When you performed sensitivity analysis by varying the importance of parameters in MCDM, what was the most surprising impact you observed on cluster head selection or network performance?

To, to, to, to for the validation of our. In a study.

Understood. When you adjusted the weights of parameters during sensitivity analysis, how did you document or communicate those findings to ensure reproducibility for someone else repeating your study?

We we enable you know, documented the sensitive analysis. Or in the in the in the paper itself. So.

Our paper is currently in in the in the in the review stage. So we have already submitted. Hey sorry, already documented into the paper itself.

Thanks. For the parameters you documented in your sensitivity analysis, did you also provide access to your simulation scripts or datasets for peer validation, or was the supporting material limited to what was included in the paper?

Thanks for clarifying. Since your paper is under review, did you also include a discussion of any limitations or potential biases in your sensitivity analysis methodology?

It's not, you know, see every research or the work. Has its limitation or a future directions.

So it's not related to the sensitive analysis. It it it work because we we we we fix some I think 6 parameters into it so the things can be. Improved by using more than six parameters. Or some. You know, like. How the parameters how they? Because ultimately it becomes some. You know, optimization of the. Saying how we can optimize the.

Right, so for your documented limitations and future directions, can you give a concrete example of a scenario where your parameter weighting approach might not generalize well to a different wireless sensor network application?

And the automatic and the notes of the energy. Into the. Network. So the point is that it's a limitation. We not only fixed or targeted about the sensitive analysis, but we discussed about the work. The paper. Overall.

That makes sense. Since you mentioned using six parameters for optimization, can you describe how you prioritized which parameters to include and if there were any that you considered but ultimately left out?

Understood. Earlier, you mentioned using DBSCAN and MCDM together for energy efficiency—can you walk me through how you integrated these two methods in practice within your clustering algorithm?

Oh sure, as I already said that DBS can be used for the cluster formation and MCD used for the clustered selection. Umm yeah, within the cluster so DBS can use it. We have the, we have the network. The nodes are. Distributed in the network.

Physically or geographically? Now the deviation is applied to form the clusters. Suppose they're. Some supports in network we have the 10 nodes, sorry 100 nodes and a zoom. 10 clusters formed. So the device can slow is to form the clusters. Now we applied the MCDM.

Okay, focusing on the DBSCAN step, did you face any challenges with nodes that were classified as noise, and if so, how did you handle those nodes in your clustering process?

Alright. When you use MCDM for selecting the cluster head from within each DBSCAN-formed cluster, how do you balance between energy level and node centrality if they suggest different optimal nodes?

To make one of the node. Within the cluster as a rope. Hope that cluster. Or that group. So they call it as a cluster head. So to select the cluster head. We use the MCDM mechanism. So, uh. Then both. Together. You know. Integrated for the energy efficient communications. In in this case so because. The first is use the. You know, I would say.

Some optimal or I would. You know whether that some efficient. For the cluster formation and the second is for the. To lower the computational overhead. So both are togetherly for the cluster formation and the cluster selection.

Alright. When applying DBSCAN for cluster formation, did you have to tune the epsilon or minimum samples parameters for different network densities, and if so, how did you determine the best values?

Thanks for clarifying. Focusing on reducing computational overhead, did you benchmark or compare the run-time performance of your integrated DBSCAN-MCDM approach against any other clustering algorithms?

Yes, we did.

Understood. When you compared your DBSCAN-MCDM method to other algorithms, what specific metrics did you use to evaluate computational overhead, and what were the results?

We find the time complexity. And our approach time complexity. What's the? Order of N square.

Given your approach has O(N squared) time complexity, did you consider any potential optimizations to reduce this, or do you think a lower complexity is realistically achievable for this type of clustering in WSNs?

You see? Yeah, it can be achievable. It's not like that. Where there's uh. Achievable or not if the. But. You know it it it it's doable. Anyhow, we can reduce it.

Understood. Since you mentioned it's doable—what specific strategies or algorithmic changes would you consider first to try reducing the time complexity of your clustering method?

Because every. Problem or a solution is open for the optimization or a further enhancement? So yeah.

Things can be reduced.

Right. Since you mentioned that further optimization is possible, can you give a concrete example of one algorithmic change or alternative approach you would try first to lower the time complexity in your clustering method?

Alright. Since you emphasized optimization potential, what would be your criteria for deciding whether an optimization to your clustering approach is worth implementing in a real-world wireless sensor network deployment?

Optimizing optimization in my I, I said that in in my way that. We somehow we are optimizing the energy open node. So energy of a node means we have their parameters like. The distance of a node from the cluster head. Distance of a node from the base station. The the the the the energy level of a node. And how many number of the?

In the neighboring nodes surrounding that node. Means what will the node degree and so on.

Understood. When considering parameters like node distance and energy level, how do you prioritize which parameter has the most impact on energy optimization in your clustering approach?

So based on these. We, we, we, you know, we can optimize the things. By using the MCD methods.

Thanks for outlining those node parameters. When you use MCDM to select a cluster head, how do you assign weights to these different factors—do you rely on empirical data, simulations, or a specific weighting method?

Understood. When applying MCDM to choose the cluster head, how do you ensure that the weighting of parameters like node degree and energy level remains robust if the network topology changes dynamically over time?

MCD method is used. To form the clusters and is. Sorry. To select the cluster, head for a for a for a cluster. So MC and DB scan is used to form the. And the clusters and MCD is to select the cluster within the.

So I think you got the answer, if not kindly.

Got it. Since your approach relies on parameters like node distance, node degree, and energy level, how do you validate that your chosen weights and optimization strategy actually improve energy efficiency in practice—do you use simulation results, real deployments, or something else?

Understood. In your experience, how do you ensure that the weights assigned to parameters like node degree and energy level in your MCDM-based cluster head selection remain relevant as the wireless sensor network evolves over time?

That's why I said no, we we didn't sign the fixed weight.

As I said, we didn't design the fixed the fixed rate only. You know, maybe if efficient or maybe some kind of the robust thing for the one environment or one network or what application, but may not be feasible or even. It's relevant for the another network, so. Since we want to make it.

Right, you kept the weights dynamic. Can you give an example of how you adjusted those weights in response to changes in network density or energy distribution during your simulations?

And the more relevant for the real life. Based on the application and domain on the network, we use the MCDM. So by using the MCD method, we assign the weights to the each and importance of weights to the weights. For the you know in the in the in the sorry importance in the form of weight to. Each and every parameters. So based on that. If we choose choose the different network then kindly possibility is that that. Importance can be changed. And if you use the another network then again importance can be changed. For the among the parameters. So that's why we use the MCDM.

Right, so you kept the weights adaptable. Thinking back to your simulations, can you describe a scenario where dynamic weighting significantly improved overall energy efficiency compared to fixed weights?

That approach makes sense. Given your emphasis on adaptability, can you describe a time in your simulations when changing the weight for a specific parameter—like node degree—led to a measurable improvement in network lifetime or communication efficiency?

Oh, OK, See, it's not in this way that we. And I can let you know the we, we. Change the node degree or something. We we, as I said.

Thanks for clarifying. When you're running these simulations with dynamic weights, how do you monitor or log which parameter changes have the greatest impact on overall network performance?