Description:Dynamic Resource Allocation System for Mitigating Malicious Activities and Optimizing Energy Consumption in Cloud Operations

2. Problem Statement:
Turning to the definition of computer clouds, the very term refers to a style of computing that relies on shared computer resources, especially storage capacity, processing power or networking capability. However, with time, the adoption of cloud services has been on the rise, which has brought about an increase in the attack on cloud systems. These are cyber criminality or hacking, which may result in to security threats, violation of secure information or disruption of services. These activities have a detrimental effect on cloud environments as they cause losses to the organizations using the cloud platforms both in terms of finance and reputation.
One key issue deemed vital to cloud service providers is that of proper management of the cloud resources especially in as much as resource provisioning in response to dynamic-cloud user requirement/profiles and in an effort to counter acts of susceptibly to malicious events. Dynamic resource allocation centers on the capability of providing the proportional and precise distribution of the system resources of clouds (CPU, storage, band width etc.) along the real requirements or threat factor. Lack of efficiency to use the resources results in performance degradations, wastage of energy or system compromise since cloud services may be proactively attacked for over-provisioned or under-provisioned resources.
In addition, energy related issues are also apparent in the case of cloud systems. Data centres are certainly more power consuming and the consumption rate is even higher with cloud data centres and the chances are also more with the growth of number of users and usage of applications. This means that the resources will either not be well utilized, or they are overloaded, they will consume much power and will also be much destructive to the environment.
The issue is as follows: On the one hand, cloud environments contain ways which are prone to malicious activity by inefficient resource handling. However, inadequate resource management leads to over-utilization of resources and this results in high operational cost with consequences on the natural resources. That is why there is a call for a workable dynamic resource allocation resource that coordinates the movements of the cloud operations; conserve as much energy as possible; and should be able to protect the system from any sorts of threats/attacks. Establishing such a system would effectively increase the security, effectiveness, and stability of cloud computing systems. Thus, it is possible to list the following main concerns:
• Cyber crimes that affect cloud environment due to poor utilization of resources.
• This problem can be attributed to wasteful use of energy in cloud data centers due to unproper energy resource management.
There are a lot of challenges that have risen and acted as a hindering factor to implement the cloud computing technology, some of these challenges includes; This upcoming calamity has necessitated the need for a new generation dynamic resource allocation mechanism that can mediate on the current issues such as cloud security and efficient energy usage.

3. Existing Solutions
At the moment, there are different approaches used by CSPs to assign and partition resources in their clouds. However, most of them do not adequately capture all the issues related to malicious activities and energy consumption optimisation in an elastic cloud system.

Static Resource Allocation: Most of the conventional cloud models are the static resource allocation mechanisms, where the resource usage and demand of the consumers such as computation capacity, storage, and network bandwidth are predetermined and cannot be decided dynamically by the consumers. While this is advantageous in some cases, it is not very efficient most of the time specifically in the period that experiences varying demands or attack rates. It may be over-provisioned that always leads to the wastage of energy, as well as increase the operational expenses unnecessarily, or under-provisioned that could lead to slow down of the system or even to service interruptions, or even to attacks from malicious users.

Measures That Will Be Taken Against The Above Unwanted Activities: Cloud service provides traditional security measures comprised of firewalls and intrusion detection systems and other access controls. Although they can identify some threats/ intrusions or unauthorized attempts, they cannot optimize the usage of the resources in order to counteract the possible threats. These solutions act at the ‘catch’ side of the attack where resources are already lost or even damaged rather than preventing attacks by maintaining the elasticity needed to manage the ever-changing threat landscape of any cloud resources.

Resource scheduling/ Resource Scaling: to increase efficiency, most of the Clouds utilize the method of scheduling of resources and auto-scaling. These solutions adjust resources up or down regarding the workload requirements to ensure equal distribution of load at any given time of the day, especially during the time when the demand is higher. However, these techniques are based on specific threshold and rule-based which do not consider the effect of malicious actions or an efficient use of energy. As compared to scaling to address such issues as fluctuations in traffic or increases in load, it does not deal with the problem of resource flexibility in response to the threat or with concerns related to the power consumed during scaling.

Some of the leading cloud service providers have started coming up with energy efficiency algorithms In this case. These algorithms balance resources to cut down the general power consumptions by making efforts to evenly distribute the cloud resources as per the energy price that come with different workloads. Thus, although those solutions improve the efficiency of utilizing energy in data centers, they do not apply dynamic security mechanisms. That does not deem with regards to the changes in security threats and the flexibility to allocate necessary resources for responding to threats without consuming more energy than is needed.

• Security and Resource Management Models: There is a list of security and resource management models like CloudSim and Openstack which mainly deals with the use of security for cloud computing and control of available cloud resources. Though these tools are good for utilizing resources for cloud environment, they do not self-adapt to resource usage based on security threats. Also, these frameworks do not also provide sufficient solution for dealing with spikes and attacks in the demand point and secondly they do not consider possibility to integrate the efficient resource allocation with security mean monitoring during the peak.

• Recent Trends: In regard with the recent decades, the AI and ML approaches help in optimizing the resources in cloud computing as well as improving the security aspect. For instance, smart systems are currently being used to make machine learning-based prediction of traffic density and consequently distribute the resources according to the expected traffic demand. Besides, anomaly detection systems are also employed for screening out activities that may be different from the norm thus pointing to malicious activities. Although these techniques are helpful in this regard, they remain partially integrated as they address security as a distinct problem from resource management. Moreover, they are usually designed to operate only for resource allocation optimization, or for the detection of an attack, and not for both concurrently as two interconnected and continuously alternating processes.

• Advanced Scheduling: There has been attempts to develop advanced techniques of the energy conscious scheduling that seeks to allocate the available cloud resources using energy signatures. They adapt resources depending on the workload that is expected to be solved and the required energy for such a purpose. Yet, these methods are not aligned with security-steered approaches and, apart from some basic means, are unable to adapt dynamically to offset a threat in the environment at the cost of partially energy savings.
. Preamble
The current invention is concerning to a dynamic resource allocation system to solve some of the main problems of cloud computing dealing with the security of cloud and energy consumption. Today cloud computing acts as the basis for many industries as an easily accessible infrastructure that provides computing power needed. However, as cloud services grow in size they unfortunately remain open to various activities of an illegitimate nature including cyber-crime, hacking and denial of service (DoS). At the same time, there are some concerns related to the energy consumption of cloud data centers, usually large-scale cloud operations imply a considerable power consumption that may result in both high operational cost and impact the environment.

Current systems in the cloud resource management are mainly developed to increase the efficiency and utilization of resources with respect to the amount of work that has to be processed, but they do not offer the security and energy aspect in an integral dynamics model. Most of the traditional models are based on static capacity planning, in which resources are either are provisioned beforehand or based on pre-set deviation points. Though such strategies do help in making the optimal use of the resources to some extent, they cannot adaptively respond to the occurrence of security threats or change the energy efficiency rather reducing it while enhancing the performance.

The above mentioned issues have been effectively eliminated by the dynamic resource allocation system by invoking a global method of security monitoring and energy conservation in real time. The system requires the overall workloads of cloud and is built with the support of superior algorithms as well as machine learning methods to identify any security threats and risks, including unauthorized access, botnet, or resource consumption attack. In case of such incidents, the system adapts the cloud resources to withstand the impact of attack and at the same time keeps the overall cloud functioning while preserving energy utilization to the maximum.

The novelty here is in the coupling or security responses with the ability to manage energy consuming resources in real time. Self-learning algorithms are employed to analyze historical and current data, thus the workload changes beforehand, before it is affected by security threats or energy concerns. For instance, during periods of high attack risk, the system can identify extra security permissions necessary such as the firewalls and the intrusion detecting systems without having to load power hungry services. Also, at the same time, with the decrease in traffic, it can decrease resource-demanding factors, for example, computing or storing powers, while maintaining the quality of services.

Also, it incorporates a mechanism to develop on its own by recalibrating through machine learning. The system adapts its predictive and decisional model as new data comes in, thus it serves as a better solution in the enhanced detection of new forms of threat to security and better utilization of energy. This makes the system more flexible in order to adapt to changing needs of cloud workloads as well as the ever-evolving threats in the cybersecurity context making the operation of clouds smarter and lasting.

The aim of this solution is to make an adaptive, efficient and real time system that allows better security, better resource management and efficient energy consumption on cloud systems that also proposes and implement security solutions against security threats. This invention serves as a keynote for cloud service provider since it avails a perfect solution in terms of scalability and cost efficient solution to the need of cloud which is steadily on demand due to its ability to improve performance while at the same time looking at the energy efficiency of the cloud computing.

6. Methodology
The dissemination of resource allocation models for cloud security and power utilization method targets at solving the two main issues relating to the effectiveness and protection of CSP cloud services as well as energy efficiency in the cloud environment. The system can also holistically adjust the resources in the cloud depending on the required load, security events and energy consumption. The procedures that can be followed in this method are described in the following text.

Step 1: Real-time resource monitoring and data collection for effective capacity building strategy for any organization or government body in developing countries.

The first process of the methodology involves analysing data that is gathered from the cloud environment continuously. This includes:
• Process information: The current work load of the cloud system is defined as the information related to the CPU usage, memory usage, storage usage and network usage.

• Security-related records: The records of the activities of the security systems in the organisation including firewalls, Intrusion detection/prevention systems, and access control to identify any opportunity for security breaches or unlawful activities.

• Energy logs: Information from energy meters and power management systems that are used for measuring the amount of energy use in the various resources such as computing nodes, the storage facility, and the networking equipment. This information is collected systematically and directly into the central processing unit of the system against where it is summed up and analyzed for decision-making purposes.

Step 3: Security threat detection and analysis can be also divided in several stages:
As soon as the data is gathered, the system then moves to the next step which is to assess the security situation of the cloud:

• Security: The system adapts machine learning algorithm and anomaly detection mechanisms to detect threats, for instance Sherman can detect unusually high usage of resource utilization or unauthorized attempts to access it.
• Threat Identification: Threats identified are categorized in accordance to their types, for instance: DoS attacks, unauthorized access, executing of malicious codes and programs. The entire program proceeds on the basis of the severity of the threats that are likely to hinder cloud operations.

It facilitates in establishing level of security resources needed to contain current or future attacks.

Step 3: Dynamic Resource Allocation Decision
Therefore, if a security threat is identified or in accordance with data on the current load and energy consumption, the decision on dynamic resource allocation is made.

• Resource acquisition: The following factors determine the amount of resource that will be attached to a particular project; Thus for instance during such times; when the demand is low or not sufficient enough to warrant the use of various services offered by the business, the system may decide to use less computing capacity or storage capacity. On the other hand during load or attack, it may increase either resources or reconfigure in order to conquer the load or defence.

• Security Pre-Emptive Mechanism: The system automatically adjusts security priorities to tackle different threats, for instance, enhancing firewalls for more capacity, increasing intrusion detection, etc., without significant interferences with other procedures. Some supporting or non-vital activities might have limited or rationed resources in order to compensate for security necessities.

• Energy Optimization: The system also has an energy optimization that would make sure that the available resources are not wasted. When there are no intrusions, or workloads are stable, it assigns resources founded on the original purpose of saving energy without compromising on the service quality.

This step also helps to ensure that the resources of the system are flexible, sustainable, resilient and meets both the security and energy requirements of the system.

Step 4: Personalization is predictive modeling and adaptive learning
This is an effective way to personalize the information presented to the user. To enhance the use of resources in future, the system adapts forecast models in order to anticipate workloads, security threats and energy consumption in the future.
• The forecast: This predicts the amount of work that is expected to be done in the future as well as the security risks that are likely to occur in the process. These are because the algorithms can acquire pattern from previous occurrences and make a better right decision in facilitating resource and demands.

• Online Model Update: The model is updated with further data on regular basis so that the system learns the new trends in usage patterns, emerging security threats or changes in energy efficiency. This kind of learning makes the system thus become more effective in making decision as it dynamically allocates resources.

Step 5: Feedback and Optimization
The latter stands as it involves feedback or assessment of how effectively the decision of resource allocation has been implemented:
• The system also checks on the performance of the decisions made regarding the resources, such as system performance, the state of security [new breaches], energy consumption [usage reduction].

• Self-organization: If existing outcomes fail to meet given objectives (e.g., high energy consumption, unmitig8ated threats), then the upcoming strategies will be modified. Optimizations might be making changes to scaling rules, increasing security resources or optimization the ways of saving energy without losing the performance of the services that are being provided.

Figure 1. Methodology Proposed

7. Result
In this section, the outcomes pertinent to the functionality and efficiency of the dynamic resource allocation system as regards to mitigate malicious activities and energy consumption are discussed. The effectiveness of the proposed system in terms of the self-organizing property in response to security threats and changes in the load can be confirmed by the measures such as system’s performance, energy efficiency, and security features. Besides, this paper compares the new approach with common cloud management systems and identifies the beneficial effect this approach has on cloud processes.

7.1 Model Performance Evaluation
This assessment of the dynamic resource allocation system was done according to its robustness and efficiency depending on the fluctuating loads of work in the cloud environment and the presence of security threats. The evaluation metrics include:
• Uptime: The system was tested with regard to how it performed in case of high traffic and under attack. As the test result depicted, the dynamic resource allocation approach had capacity readiness nearing one hundred percent during the time of real high malicious traffic or high traffic period.
• Energy Efficiency: The system designed show a saving of 30% of the total energy which was compared with other traditional method of energy resource allocation. The energy optimisation module ensured that resources were being used to the best of their ability and at the same time manage to minimize the power used by the system.

Thus, it was possible to see that during DoS and other malicious activities, the system creates additional security layers in the form of firewalls and intrusion detection systems without any negative impact on cloud work. On the response time and latency, it was observed that they were not in any way affected by the injected security threats.

Table 1: System Performance metric results
Metric Dynamic Resource Allocation System Traditional Cloud Management
System Availability 99.98% 98.56%
Energy Consumption 30% Reduction Baseline
Response Time (ms) 150 ms 200 ms
Latent Threat Detection Time 2.5 seconds 6 seconds
Security Breach Recovery Time 5 seconds 15 seconds
Table 1 highlights the comparisons of the factors constituting the proposed dynamic resource allocation system with the traditional systems for cloud management. These ideal characteristics include superior availability of the system during measures to contain the security breach as well as on the energy efficiency while operating and response time during the detection of the security breach.

7.2 Security Performance Evaluation
The dynamic resource allocation system feature a capability of detecting security threats and responding to them, so as to provide real time cloud resource adjustments in an attempt to foil the threats. Based on the threat model several tests were performed which include DDoS attacks, brute force attacks and data intrusion attacks. The results indicate that:
• They observed that the system was capable of identifying DDoS attacks and was quickly adjusting for resources for the purpose of sustaining the services.
• Brute force was quickly addressed, with more computational resources allocated for authenticating the inputs thus making it difficult for the breach to be an success.
• Data leakage was averted through moves to encrypt storage of the data including moving such data to other insecure places.

Figure 2. Security Threat Detection and Mitigation Effectiveness
From the above figure, it is evident that the system is able to identify threats and respond to them in real sense. There was a reduction in the time taken to identify the security threats since the system underwent some changes and enhanced its overall effectiveness in handling security threats and security incidents.

7.3 Energy Consumption Analysis
The energy optimization module was evaluated based on its effect on consumption of power in cloud data centre. The power saving was realized mainly during the low demand period whereby all unnecessary resources were turned off to minimize on power usage. This was done with an aim of identifying and comparing the various consumptions of energy arising from the various cloud workloads within the system thus:
• Save 30% on energy during off-peak to improve resource scaling and tune down on the excessive consumption of resources.
• During peak usage, it is possible to save 10-15% of power as the resources are distributed in such an organized manor and energy efficient schedules are followed.

Figure 3. Energy Consumption Comparison
The above figure tabulates and compares the energy consumption of the proposed dynamic resource allocation system and other traditional cloud management solutions. There is a cut down on the power consumption in the process provided there be no negative impacts on quality of services.

7.4 Adaptive Learning and Predictive Modeling Performance
One of them is the capability of learning where there is the learning from the previous resource allocation decisions to enhance the security occurrences. Some of the workloads include estimating the level of demand, threat levels, and power consumption in the system. The results indicate:
• In real-time environment the accuracy of the system has risen up to 20% higher after the first month of operation due to the learning process of system.
• After some months the estimation ability of the system on the workload variation enabled to cut on most of the unnecessarily opportunities and as consequence on energy demand increases 15%.

Figure 4. The adaptive learning and its interaction with predictive modeling
Figure 3 demonstrates how the system performance enhances in the aspect of the predictive capability progressively as a result of the model is conveniently predicting the needed cloud workload and security threats more effectively and hence the allocation of resources.

7.5 Scalability and Performance Under High Demand
Finally, all the Cincinnati operations were put to the final test with a simulated high load on the cloud combined with automatically genereated attacks on the system. During the following areas, the system displayed excellent and efficient scalability, optimum performance as well as optimal security:
• Surges in demand: the system was able to adjust cloud resources for increased usage, it did not affect the performance of the applications.
• More security threats: It was easy to assign security in response to threats without much compromising the usage and stability.
Table 2: Scalability and system performance through maximum demand
Metric High Demand (Dynamic Allocation) High Demand (Traditional)
System Latency (ms) 120 ms 250 ms
Security Response Time (sec) 4 seconds 12 seconds
Resource Scaling Time (sec) 3 seconds 10 seconds
Table 2 compares scalability and system performance under high-demand conditions. Due to this, the dynamic resource allocation system yields better results in a short span of time, when compared to more demand and improved security risks.

8. Discussion
In view of the conclusion and result of the dynamic resource allocation system, we can certify that the system is capable to solve two significant issues at present cloud environment, which are cloud security and energy consumption. Perhaps, the most exciting element of this system is that it combines real-time security monitoring with adaptive resource scaling since it proposes a solution that will guard cloud resources from potential threats as well as save energy. This is a step chicken for cloud service providers as the move enables the creation of a secure environment for the custody of data as well as making it possible to contain costs when dealing with large data volumes in the risk-prone IT environment.

An important factor that is closely related to the case is the energy efficiency of the system that was improved by 30%. Cloud data centers are currently a significant consumer of electricity, and it is crucial to strive to attain resource efficiency to cut down on cost and make a minimal impact to the environment for cloud computing. This way, the resources, being in a position to be allocated depending on workload and the security demands are effectively used while little power is wasted. Such energy savings are critical in low demand because traditional processes allocate extra resources in anticipation of the demand surge, which is not only unnecessary but also wasteful.

Another factor is the capability of the system to address the security threats online in an adequate manner. Three major risks found in attacking cloud systems, which if not well managed, have a tendency to produce devastating breakthroughs are DDoS attacks, unauthorized access, and resource exhaustion attacks. The system was characterized by a quick identification of these threats and the subsequent identification of resources to counter them. The flexibility to give a swift response to security threats without impairing cloud services is one of the crucial differences compared with the older-type systems that introduced set security procedures that might address a menace just after it unfurls. First, the system offers potential for immediate and end users learning and predictive modeling that can help prevent any changes in the cloud resources which means that the cloud is more effective as it can adapt to problem change.

These factors are considered to be important factors relevant to the success of the system, especially the components of the learning and the predictive modeling. It was however evident that the system kept updating it database and improving on the decisions taken in the utilization of resources. The use of ML also made it possible to predict works loads and security risks in order to improve on resource allocation as well as security and energy utilization. This capability is the key in allowing ongoing adjustments to the system because workloads, and threats in the cloud computing are dynamic.

Nevertheless, there were some issues that need to be addressed: However several challenges are still realized at the simulation and controlled conditions of the system. Indeed, the quality of the data can have a significant amount of influence over the performance of the system since most parameters calculated for expectations and estimations depend on the appropriate data. Because of it, the prediction could be imbalance, noisy or outdated that makes the system unproductive. The system also relies on acquiring high quality single-cell data or traffic logs that may not be available in unstructured data processing environment. By this way, it can be pointed that data integrity and its consistency are one of the crucial factors for system performance.

The current approach has another disadvantage: it requires much computational effort to be applied. The principles of its work are clear: providing high levels of security and energy conservation while taking considerable amounts of computing power for real-time threat identification, learning, and modeling. This makes the system particularly appropriate for organizations that offer massive cloud services with considerable infrastructure but it may not be so suitable for the organizations which may have a small or limited resources. There is also the need to carry out further enhancement of the machine learning algorithms and models for better scalability of the system and lower cost of monitoring in real time and the adaptive allocation of resources.

Further, although the presented system was shown to enhance the security and efficiency, the results have to be verified in different forms of clouds and types of applications. It should be tested with more varieties of cloud environments, such as private cloud, hybrid cloud, multi-cloud, and so on and more specific workloads. This will aid in generalization, ensuring that the system has the capability to perform according to the cloud computing setting.

Here are development opportunities for the next steps of research:

• Multi Modal Data Integration: Incorporating the multiple type of data sources, for instance Genomic data, Immunohistochemistry images, and others could help to improve the accuracies of predictions and resource management in the system.
• Possible further research for this problem should be directed to study better optimization algorithms to be employed for dynamic resource management, more especially in situations of high demand or security breaches.
• Edge Computing Integration: With edges computing, integrating this system with the different edge computing platforms can enhance real-time resource provisioning nearest to the data source hence improving the system’s response time.
• Feedback Loops: There is also an additional feedback loop mechanism that adds an ability to automate the security model improvement upon its operations and the results based on its performance related to energy utilization.

Thus, the dynamic resource allocation system described in this patent proves to be a viable solution for CSPs where security and energy efficacy are favoured: simple yet effective. This means that because of Machine, the system is capable of cloud management in a way that is more effective than conventional cloud in terms of providing quick duty to resource allocation based on threats and workload. With the development of cloud computing, it remains the future of this cloud model to foster better solutions in the management of cloud resources from secure and sustainable aspects.

9. Conclusion
In conclusion, the dynamic resource allocation system introduced in this patent constitutes a new development in the allocation of cloud resources by solving two major problems that is security and energy issues in cloud computing. It is by the aid of adaptive learning coupled with predictive modeling that the system is able to recommend suitable cloud environments to be provisioned based on the real-time data as well as security threats that may exist at the time of usage in order to achieve optimal resource utilization, optimum security measures, and optimized energy consumption.

Security: This is the system’s capability to detect and prevent all forms of unwanted activities in real time as well as reducing energy consumption in real time as being one of the most important achievements made in the field of cloud computing. Compared to the traditional and rigid value systems, the dynamic value system enables the CSPs to be more adaptable to changes internally and externally and hence increase the chances of working efficiently and more sustainability in cloud operations. Moreover, the employment of the machine learning algorithm makes the system more flexible in directing resources since the algorithm improves the existing strategy when it processes more data.

Although having good results in terms of performance it is still requires additional improvements in the specific problems related to the qualities of the data, computational burden and the ways of the efficient scaling of the system over the different types of clouds and many – sided usage scenarios. The progressive advancement of the cloud infrastructure and the enhancement of the threats and energy consumption requirements make the utilization of such an initiative of enhancing security and sustainability for cloud computing inevitable.

It is important to note that integrating the dynamic resource allocation system described in this patent will revitalize the management of cloud services to meet the users’ demands and address their concerns under new conditions and threat models for cloud services. If it is refined and validated in other futuristic cloud settings and in permutations for other industries, then this system would likely form the basis of the next era of cloud computing where users are enjoying relevant, optimized, secure and energy-wise solutions for their needs as much as the service providers are getting an efficient and cargo-carrying infrastructure for their solution offerings.

In conclusion, this invention presents almost a holistic approach that cloud service provides can leverage to enhance the security and performance of their cloud services while at the same time making them energy efficient for the future sustainability of cloud computing services.
, Claims:10. Claims
1. An approach to adaptive provisioning of resources for clouds to ensure secure and energy-efficient operation of the system.
• A data collection module capable of synchronously monitoring, in a cloud environment, system performance, workload, power usage, and security threat;
• A security detection module that is responsible for scanning and searching for any related threats and activities that are malicious in the system, including DDoS, unauthorized access, and system vulnerability.
• A resource allocation module that will make automated decisions regarding distribution of resources in the cloud owing to issues of security threats and work load demand so as to provide the best resource output and energy efficiency.
• In this case, an energy optimization module that will serve to adjust the usage and distribution of resources so as to lower energy utilization when the demand is low, or during an attack.
• An agent to build factorial models for precise prediction of future workload scenarios, new risks, and power requirements which provides an innovative assistant for readjusting current decisions depending on the previous experiences.
2. The above mentioned resource allocation module adjusts the amount of resources required for a cloud in actual time depending on the security threats and the amount of work, thereby efficiently distributing the computational capacity, storage and networking resources.
3. The system as claimed in claim 1, where the security detection module utilizes the anomaly based detection and machine learning algorithms to detect amount and types of resource that are being consumed or the behavior that deviates or at par with malicious activity, the resource allotment is then adjusted based on the activity.
4. Cloud elastic approach that can be used to manage resources with a motive of controlling threats and power usage, which comprise the following steps:
• Real time tracking of the status and metering of activity, workload, power consumption and security status of cloud resources.
• Identifying security threats with the help of the machine learning algorithms and anomaly detection method to isolate any probable malicious actions or unauthorized access.
• Adapting or dynamically mapping the clouds resources in relation to security threats and workload loads, assigning security priority to the cloud resources during an attack and mapping cloud resources to high energy consumption during low traffic.
• Intelligently controlling the amount of resource usage necessary to attain the least power consumption while using fresh approaches to scheduling so as to scale the power efficiency as effectively as possible.
• Forecasting the resources required in the future, and using planning techniques in relation to past data, and machine learning and adaptive learning.
5. In this particular aspect, the method of claim 4 states that the system uses the concept of adaptive learning in an effort to predict further workload and security threat so as to adjust resources in the future.
6. The method as recited in claim 4, wherein the method includes the step of recalibrating the predictive model to reflect further data or dynamic cloud profiles to make certain that the system remains properly adjusted to the current conditions with regard to security as well as energy efficiency.
7. A system for dynamic management of resources in a cloud computing environment, which is defined as
• The first one is a system for dynamic resource allocation with respect to security and energy efficiency concerns specified as a claim of the invention.
• An examination module that assesses the effectiveness of the dynamic resource allocation decisions, and applies it to improve on other new resource allocation solutions in light of the operations and real-time results.
8. A process for enhancing the protection and power utilization efficiency in the cloud storage facilities:
• Real-time identification of security threats using ML algorithms; Real-time security threats anomaly detection.
• Adapting to various infringements that may occur and provide secure management focus to secure affairs while not upsetting consumption and delivery of other cloud services.
• Studying how to use energy resources in an efficient manner by providing attention towards utilization of them at particular periods which can majorly cut down electricity usage yet not affect the quality of services offered.
9. The method as specified in claim 8, which is to learn from data and adapt the security actions and energy management proactive percentage depending on the usage of cloud assets in the whole life cycle.
10. Therefore, the invention of a cloud service provider platform include the following components:
• The system of claim 1 for dynamic resource allocation;
• Cloud performance and security information, energy consumption statistics of the system and cloud resources as well as the extent of threats to cloud; A mechanism that can enable the cloud administrator to change the resources reallocation when necessary in real time manner.