Description:METHOD FOR EFFECTIVE RESOURCE ALLOCATION IN CLOUD COMPUTING USING DYNAMIC STACKELBERG GAME BASED MULTI-OBJECTIVE APPROACH
Field of the Invention
[0001] The disclosure relates to cloud computing resource management, and more specifically to a system and method for dynamically allocating virtual computing resources across a multi-tenant cloud environment. The disclosure addresses the optimization of resource distribution among various cloud service providers (CSPs) and subscribers in a way that maximizes efficiency, minimizes costs, and maintains fairness in a multi-cloud computing landscape.
Background
[0002] The background description includes information that may be useful in understanding the present invention. It is not an admission that any of the information provided herein is prior art or relevant to the presently claimed invention, or that any publication specifically or implicitly referenced is prior art.
[0003] In the domain of cloud computing, the allocation of resources has traditionally been a challenge, especially when balancing the diverse objectives of cost, performance, and user demand. Early systems were rigid, often leading to inefficient resource utilization. For instance, a subscriber requiring a temporary increase in computing power for data analysis would have to provision a static virtual machine (VM) with a fixed number of resources, which may not always match their dynamic needs. As cloud computing evolved, so did the methods for managing said resources.
[0004] The traditional approach to resource allocation often involved cloud service providers (CSPs) managing their own resources independently, with allocation policies that were inherently static and isolated. The traditional approach could result in sub-optimal utilization, as resources could be over-provisioned to ensure availability, leading to increased costs and inefficiency. With the introduction of virtualization technology, became possible to create VMs that could be allocated more flexibly, but the virtualization technology did not completely solve the allocation issue as the coordination between different CSPs and the varying demands of subscribers remained complex.
[0005] Early multi-tenant architectures allowed multiple users to share the same physical hardware, however, said architectures did not provide sophisticated mechanisms to balance and dynamically allocate resources across different clouds. Consequently, users were often left to manually select and manage their resources across multiple providers, a process that was both time-consuming and prone to error. Furthermore, static pricing models did not reflect the real-time value of resources, leading to economic inefficiencies.
[0006] With the advent of cloud brokerage services, there was a move towards more centralized allocation systems. Said systems aimed to optimize resource utilization by managing the distribution of resources across multiple CSPs. Nevertheless, they often lacked a multi-objective framework that could accommodate various user requirements, cloud capabilities, and dynamic market conditions.
[0007] To address said limitations, researchers began exploring market-based allocation strategies, such as auction-based models, which could dynamically price resources based on current demand. However, said models did not always account for the strategic behavior of CSPs and the potential for pricing games that could emerge as CSPs vied for subscribers. The game-theoretical models such as the Stackelberg game were applied to create hierarchical structures of competition among CSPs, which could lead to more strategic and efficient allocation of resources.
[0008] The Volume-Weighted Average Price (VWAP) method emerged as a valuation strategy to ascertain the fair value of virtual resources, which was especially useful in scenarios where CSPs provided a mix of static and dynamic VMs. The VWAP allowed for the determination of a cloud's price leadership within the market, influencing the strategic positioning of CSPs in the allocation process.
[0009] Despite said advancements, there remained a need for an integrated system that could seamlessly incorporate user requirements, dynamic resource valuation, and strategic provider behavior into a cohesive allocation model. The system would not only require the ability to handle targeted requests but also manage open requests through auctions and adapt the allocation of dynamic virtual resources in real time.
[00010] Hence, the prior art reflects a progression from static, isolated resource management towards more integrated and strategic allocation systems. The evolution of said systems illustrates a continuous search for balancing efficiency, cost, and performance in a rapidly growing and competitive cloud services market. The challenges of previous systems pave the way for approaches that can leverage both market-based and game-theoretical models to optimize multi-objective resource allocation in cloud computing environments.
Summary
[00011] The disclosure relates to cloud computing resource management, and more specifically to a system and method for dynamically allocating virtual computing resources across a multi-tenant cloud environment. The disclosure addresses the optimization of resource distribution among various cloud service providers (CSPs) and subscribers in a way that maximizes efficiency, minimizes costs, and maintains fairness in a multi-cloud computing landscape.
[00012] The following presents a simplified summary of various aspects of this disclosure in order to provide a basic understanding of such aspects. This summary is not an extensive overview of all contemplated aspects and is intended to neither identify key or critical elements nor delineate the scope of such aspects. Its purpose is to present some concepts of this disclosure in a simplified form as a prelude to the more detailed description that is presented later.
[00013] The following paragraphs provide additional support for the claims of the subject application.
[00014] In the realm of cloud computing, efficient and strategic allocation of resources is essential to meet the growing demands of users and to maintain a competitive edge. The multi-objective resource allocation system described here is a solution that addresses the need by orchestrating a collaborative resource pool across a consortium of cloud service providers. At the core of the system is an allocation unit, infused with an intelligent policy engine capable of processing a multitude of subscriber requests, each with the unique requirements. The unit serves as the linchpin that harnesses resources from the combined offerings of the participating cloud providers, aligning them with user needs through a refined understanding of request attributes and allocation policies.
[00015] The system transcends traditional resource allocation models by incorporating an interface within the allocation unit that differentiates between targeted and open requests. Targeted requests, which pinpoint a specific cloud service provider, are seamlessly executed as per the subscriber’s preferences. In contrast, open requests are propelled into a competitive auction arena, where unassigned static virtual machines are leveraged, promoting a healthy competitive environment and ensuring resource optimization.
[00016] Underpinning the joint resource pool is a sophisticated management system that interlaces static virtual machines, perpetually ready for deployment, with dynamic counterparts that come into play on an as-needed basis. The valuation of said resources is meticulously calculated using the VWAP method, ensuring equitable market valuations and fostering transparency within the marketplace.
[00017] Advancing the system's ingenuity, a scheduler within the allocation unit employs the principles of the Stackelberg competition model, intelligently determining the pecking order of cloud service providers based on resource valuations. The strategic manoeuvring ensures a balanced and efficient allocation process that aligns with the dynamic nature of market forces and demand fluctuations.
[00018] Concerned with open requests, the auction mechanism is a game-changer, enabling the system to dynamically provision dynamic virtual machines based on real-time auction outcomes and pending request volumes. The feature ensures that even as market conditions and user demand shift, the resource allocation system remains agile and responsive, dynamically adjusting resource distribution and maintaining service continuity across the cloud ecosystem.
[00019] Thus, the multi-objective resource allocation system stands as a hallmark of research, redefining how resources are allocated in the cloud to meet diverse user needs, while promoting fairness, efficiency, and strategic resource utilization among cloud service providers.
[00020] The method for allocating resources in a multi-objective resource allocation system represents a significant leap forward in cloud computing, ensuring optimal resource distribution across a diverse subscriber base. The method begins by processing a plethora of resource requests from subscribers, which are then methodically categorized. Targeted requests pinpoint a specific cloud service provider based on predefined user preferences or contractual obligations, while open requests remain agnostic, leaving room for strategic allocation from any provider within the pool.
[00021] The allocation of said requests is not arbitrary, but is conducted within the framework of a meticulously crafted joint resource pool that amalgamates the capabilities of numerous cloud service providers. The guiding principle behind allocation is a set of predefined policies, carefully designed to strike a balance between the intricate demands of the subscribers and the nuanced attributes of the providers. The harmonization ensures that resource distribution is not only fair but also aligns with the overarching objectives of both service consumers and providers.
[00022] A standout feature of the method is the application of the Volume Weighted Average Price (VWAP) method. By utilizing VWAP, the system can dynamically assess the current market value of virtual machines, ensuring that the allocation of resources is in step with market conditions. The valuation plays a pivotal role in sequencing the engagement of cloud service providers, as valuation helps establish a clear hierarchy of price leadership, which in turn influences the distribution of resources within the pool.
[00023] The method further distinguishes per se by honoring the terms of targeted requests. Said requests are allocated to predetermined cloud service providers at fixed rates, as dictated by existing contracts, and remain immune to the ebb and flow of market price changes. The method provides a level of predictability and stability for both subscribers and providers.
[00024] Additionally, the method incorporates an auction-based mechanism for open requests. The mechanism allows non-allocated static virtual machines to enter a bidding process, ensuring that resources are utilized efficiently and that market dynamics are leveraged to benefit the system. Following the auction, the system calculates Stackelberg output volumes, which is a sophisticated approach to dynamically provision virtual machines based on the interplay between the auction outcomes and the unserved demands.
[00025] Lastly, upon the completion of each allocation cycle, the method entails a rebalancing of prices. The rebalancing of prices is achieved by applying advanced technical analysis techniques as specified by the allocation policy, ensuring that prices reflect the latest market conditions and system objectives. The inclusion of the step is pivotal in maintaining equilibrium in the marketplace, thereby ensuring that the resource allocation remains viable, competitive, and fair over time.
Brief Description of the Drawings
[00026] The features and advantages of the present disclosure would be more clearly understood from the following description taken in conjunction with the accompanying drawings in which:
[00027] FIG. 1 pictorially portrays an architectural paradigm of a multi-objective resource allocation system, according to some embodiments of the present disclosure.
[00028] FIG. 2 figuratively illustrates an exemplary schematic flow diagram of a method for allocating resources in a multi-objective resource allocation system, according to some embodiments of the present disclosure.
[00029] FIG. 3 represents a block diagram of the multitenant resource allocation logic for pool of cloud service provider.
[00030] FIG. 4 represents a block diagram of the multi-objective resource allocation process.
Detailed Description
[00031] In the following detailed description of the invention, reference is made to the accompanying drawings that form a part hereof, and in which is shown, by way of illustration, specific embodiments in which the invention may be practiced. In the drawings, like numerals describe substantially similar components throughout the several views. These embodiments are described in sufficient detail to claim those skilled in the art to practice the invention. Other embodiments may be utilized and structural, logical, and electrical changes may be made without departing from the scope of the present invention. The following detailed description is, therefore, not to be taken in a limiting sense, and the scope of the present invention is defined only by the appended claims and equivalents thereof.
[00032] Pursuant to the "Detailed Description" section herein, whenever an element is explicitly associated with a specific numeral for the first time, such association shall be deemed consistent and applicable throughout the entirety of the "Detailed Description" section, unless otherwise expressly stated or contradicted by the context.
[00033] Disclosed herein a multi-objective resource allocation system 100 is a sophisticated technological framework designed to efficiently distribute computing resources from a pool of cloud service providers to subscribers. According to a pictorial depiction of FIG. 1, illustrating an architectural setup of the system 100 that can comprise functional elements, yet not limited to a pool of cloud service providers 102, an allocation unit equipped with a policy engine 104, and a joint resource pool 106. A person ordinarily skilled in art would prefer those elements or components of the system 100, to be functionally or operationally coupled with each other, in accordance with the embodiments of present disclosure. The disclosure explores the resource allocation system 100 in-depth, examining the key features, functionalities, and the role the system plays in the modern digital landscape.
[00034] In yet another embodiment, the primary purpose of the multi-objective resource allocation system is to optimize the allocation of computing resources based on the evolving demands of subscribers and the attributes of cloud service providers. The allocation system relies on a policy engine that receives and processes resource requests from subscribers, allowing the allocation system to make allocation decisions that align with the broader goals and requirements of the system. Thus, the system acts as a dynamic intermediary between subscribers and cloud service providers, orchestrating the distribution of resources in a manner that maximizes efficiency and effectiveness.
[00035] One of the distinguishing features of the resource allocation system is the ability to categorize resource requests into two distinct types, targeted requests and open requests. Targeted requests are directed towards a specific cloud service provider, whereas open requests are not tied to any particular provider. The allocation unit, which forms the core of the system, plays a pivotal role in managing said requests. The allocation unit ensures that targeted requests are fulfilled by the designated cloud provider, while open requests follow a different path, involving a competitive bidding process.
[00036] Pool of cloud service providers can be akin to a virtual marketplace where various cloud service providers offer their computing resources. Said providers may differ in terms of infrastructure, pricing, geographic locations, and service offerings. For instance, Amazon Web Services (AWS), Microsoft Azure, and Google Cloud Platform (GCP) could be part of the pool.
[00037] The allocation unit incorporates a policy engine that receives and processes resource requests from subscribers. The policy engine is not a static set of rules but a dynamic system that adapts and evolves based on real-time data and changing requirements. For instance, when a subscriber requests additional server resources to handle increased web traffic during a promotional campaign, the policy engine evaluates the request based on factors like cost, availability, and performance requirements.
[00038] In yet another embodiment, the joint resource pool represents the collective computing resources provided by the cloud service providers in the system. The pool is not static, but is dynamically managed to ensure optimal resource availability. The pool includes a mix of static virtual machines and dynamic virtual machines. Static virtual machines are pre-provisioned and continuously available. Static virtual machines serve as the foundation of the resource pool, providing stability and reliability. Dynamic virtual machines, on the other hand, are provisioned on-demand based on the current allocation requirements. The flexibility ensures that resources are allocated efficiently as demand fluctuates. For example, a subscriber may request static virtual machines for their core workloads and dynamic virtual machines to handle sudden spikes in user activity.
[00039] To determine the value of virtual machines within the joint resource pool, the system employs the Volume Weighted Average Price (VWAP) method. VWAP takes into account both the quantity and price of resources, providing a fair market value assessment. For instance, if a cloud service provider offers a higher number of virtual machines at a lower price, their VWAP may be more favorable, attracting more demand from subscribers.
[00040] In yet another embodiment, the allocation unit is equipped with a scheduler that applies a Stackelberg competition model among the cloud service providers. The Stackelberg model establishes a hierarchy based on the valuation of resources, allowing the system to optimize the order of resource allocation. In practical terms, consider a scenario where multiple subscribers have open requests for dynamic virtual machines. The scheduler uses the Stackelberg model to prioritize which cloud provider should fulfill said requests first, taking into account factors like VWAP and resource availability.
[00041] Open requests, which are not tied to a specific cloud provider, are processed through an auction mechanism. The mechanism allows subscribers to bid on non-allocated static virtual machines. For example, if two subscribers have open requests for additional computing resources, they can submit bids to access said resources. The subscriber willing to pay a higher price may secure the resources they need.
[00042] Based on the outcomes of the auction and the unserved requests, the system recalculates and adjusts the Stackelberg output volumes for provisioning dynamic virtual machines from the respective cloud providers. The dynamic adjustment ensures that the allocation process remains responsive to changing market conditions. For instance, if the auction results in a surge in demand for a particular cloud provider’s resource, the system can allocate more dynamic virtual machines from that provider to meet the increased demand.
[00043] In yet another embodiment, the multi-objective resource allocation system described above offers several significant benefits and has wide-ranging implications for both subscribers and cloud service providers. Subscribers can access computing resources at competitive prices, especially through the auction mechanism for open requests. The cost efficiency can significantly reduce operational expenses for businesses.
[00044] Subscribers can scale their resource allocation up or down as needed, with the ability to request both static and dynamic virtual machines. The flexibility is crucial for adapting to changing workloads and demands. The use of VWAP ensures fair market value assessments, preventing subscribers from overpaying for resources. The fairness fosters trust in the system.
[00045] Cloud providers benefit from participating in a dynamic marketplace that considers the value of their resources. Cloud providers can attract more subscribers by offering competitive pricing and maintaining a favorable VWAP. The Stackelberg competition model and scheduler help cloud providers optimize their resource allocation, ensuring that resources are allocated efficiently.
[00046] Cloud providers have the opportunity to generate additional revenue through the auction mechanism for open requests. Higher demand for their resources can lead to increased revenue streams. The multi-objective resource allocation system has profound implications for the broader digital landscape. The system's ability to dynamically allocate resources based on demand and value contributes to efficient resource utilization in cloud computing, reducing wastage and maximizing productivity.
[00047] In yet another embodiment, the competitive bidding process and Stackelberg competition model introduce market dynamics into the cloud computing space, promoting healthy competition among cloud service providers. Subscribers can easily scale their operations up or down, fostering business agility and adaptability in an increasingly fast-paced digital environment. The system relies on real-time data and algorithms to make allocation decisions. The data-driven approach is a hallmark of modern digital ecosystems, where insights are derived from vast amounts of data to drive decision-making.
[00048] By using VWAP and a competitive auction process, the system ensures fairness and transparency in resource allocation, mitigating the risk of monopolistic practices. The multi-objective resource allocation system described here represents an approach to managing computing resources in the cloud. The system leverages dynamic allocation, valuation methods, competitive mechanisms, and data-driven decision-making to optimize the allocation of resources. The system's ability to balance the interests of subscribers and cloud service providers while promoting efficiency and fairness makes the system a valuable tool in the ever-evolving digital landscape.
[00049] Proposed method outlines a comprehensive approach for efficiently allocating resources in a multi-objective resource allocation system. The method is particularly relevant in cloud computing environments where subscriber’s resource demands can vary significantly, and cloud service providers offer diverse capabilities.
[00050] Referring to a figurative illustration of FIG. 2, represents a flow diagram of the method 200 that can comprise steps of, yet not restricted to, (at step 202) receiving a plurality of resource requests from subscribers, (at step 204) categorizing the requests into targeted requests that specify a preferred cloud service provider and open requests that are provider-agnostic, and (at step 206) allocating resources from a joint resource pool formed by the collective capabilities of multiple cloud service providers. Said steps of the method 200 can be performed or executed, collectively or selectively, randomly or sequentially or in a combination thereof, in accordance with the embodiments of current disclosure.
[00051] In yet another embodiment, the method begins by receiving a multitude of resource requests from subscribers. Said requests can encompass a wide range of computing resources, such as virtual machines, storage, or network bandwidth. Subscribers may have varying needs, and the system must effectively allocate resources to meet said requirements.
[00052] Once the resource requests are received, they are categorized into two distinct types, targeted requests and open requests. Targeted requests specify a preferred cloud service provider to fulfill the request, while open requests do not specify a provider preference and are considered provider-agnostic.
[00053] In yet another embodiment, the heart of the method involves allocating resources from a joint resource pool. The pool comprises the collective capabilities and resources of multiple cloud service providers participating in the system. The allocation process considers both the categorization of the requests and predetermined allocation policies. Allocation policies are predefined rules and guidelines that dictate how resources should be distributed. Said policies are carefully crafted to take into account the specific requirements of subscribers and the attributes of cloud service providers. The goal is to optimize resource allocation to achieve a balance between subscriber satisfaction and provider efficiency.
[00054] Virtual machines within the joint resource pool are valued using the Volume Weighted Average Price (VWAP) method. VWAP determines the current market value of said resources by considering both the quantity and price of available virtual machines. The valuation process ensures that resources are allocated fairly and transparently. The sequence in which cloud service providers are engaged in resource allocation is determined based on the valuation of their resources. Providers with more favorable VWAP values are prioritized, effectively establishing price leadership within the provider pool. The providers offering more competitive pricing or better resource availability have an advantage in resource allocation.
[00055] Targeted requests, which specify a particular cloud service provider, are assigned to their predetermined providers at fixed rates. Said rates are typically established through contractual agreements between subscribers and providers and remain unchanged, even in the

face of market fluctuations. Said rates ensures stability and predictability for subscribers who have specific provider preferences.
[00056] Open requests, which do not specify a particular provider, are subject to an auction-based mechanism. In the process, non-allocated static virtual machines are made available to subscribers through a bidding system. Subscribers can bid for said resources, and the highest bidder secures the resources they need. The competitive approach allows for market-driven resource allocation.
[00057] Following the auction process and considering the outcomes and unserved requests, the method calculates Stackelberg output volumes. Said volumes determine how dynamic virtual machines are provisioned to meet the remaining demands. The Stackelberg model, with the hierarchical provider prioritization, is used to optimize the allocation.
[00058] At the end of each allocation cycle, a price rebalancing step is performed. The rebalancing aims to adjust resource prices to align with the allocation policy's objectives. Technical analysis techniques prescribed by the allocation policy are applied to achieve the rebalancing. Essentially, the step ensures that resource prices remain in line with the desired balance between subscriber requirements and provider attributes.
[00059] Referring to one or more preceding embodiments, the method for allocating resources in a multi-objective resource allocation system is a sophisticated and data-driven approach that aims to optimize the distribution of computing resources among subscribers and cloud service providers. The method 200 leverages categorization, valuation, contractual agreements, auctions, and price rebalancing to create a dynamic and efficient resource allocation ecosystem that balances the interests of all stakeholders involved.
[00060] Referring to one or more preceding embodiments, the disclosure pertains to the field of distributed computing systems, wherein a multitude of users or subscribers require access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, services) that can be rapidly provisioned and released with minimal management effort or service provider interaction.
[00061] The disclosure focuses on the development and implementation of an allocation unit that serves as an intermediary between subscriber’s resource requests and the collective resources of multiple CSPs. The allocation unit is equipped with a logic mechanism to categorize and fulfil requests based on various attributes such as required processing power, memory, storage, and associated costs, as well as subscriber-defined criteria like target cloud preference.
[00062] Additionally, the disclosure encompasses the establishment of a market-based allocation system that includes auction mechanisms for open requests and Stackelberg game-based provisioning for dynamic resource allocation. The multi-objective system ensures that resources are allocated not only based on subscriber needs but also in accordance with economic models that promote efficient and cost-effective usage of cloud resources.
[00063] Thus, the system finds applicability in industries and sectors where cloud computing services are utilized, including but not limited to data analytics, web and application hosting, backup and recovery, content delivery and media distribution, and software development and testing. The technical field further involves the application of economic and game theory principles to the allocation and pricing of virtual resources in cloud computing environments.
[00064] Referring to one or more preceding embodiments, the objectives of the proposed the multi objective resource allocation system may include establishment of a dependable and effective method for assigning resources. The challenge of distributing resources among various users for a collective of cloud service suppliers, can be addressed. A resource distribution strategy that adheres to Service Level Agreements (SLAs) and Quality of Service (QoS) standards, can be implemented. Further, management of pricing in a manner that prevents resources from experiencing trepidation about being competitively bid upon, through the use of financial analysis methods, can also be achieved.
[00065] Referring to one or more preceding embodiments, the system can ensure the dynamic provision of Virtual Machines (VMs) utilizing Stackelberg competition models, alongside horizontal and vertical scaling, enabling flexible VM access for users. The system can maximize the income for cloud service providers while maintaining a balance in the allocation of resources. The system can establish a flexible pricing model for cloud resources that relies on a weighted volume average approach for a more accurate valuation of VMs. Additionally, the system can maintain fair competition among cloud service providers, ensuring each has an equal opportunity for growth. The system can decrease the overall burdens on the system in terms of both administrative and financial aspects.
[00066] According to a pictorial illustration portrayed in FIG. 3, represents a block diagram of the multitenant resource allocation logic for pool of cloud service provider. Referring to the FIG. 3, multiple users sharing multiple resources across multiple clouds, indicating various individuals or organizations that are utilizing cloud services. Said users are looking to access multiple resources which are distributed across different cloud environments.
[00067] In the middle of the diagram, two interlinked elements labeled "Allocation Logic" and "Allocation Policy" are depicted with a gear icon, suggesting the decision-making or the processing happens. Allocation Logic refers to the algorithms or methods used to decide how resources are assigned to users. Allocation Policy amy refer to the set of rules or guidelines that govern the allocation logic, ensuring that resources are distributed according to specific priorities, limits, or user requirements.
[00068] Cloud icons are depicted, representing different cloud services or platforms. The arrows indicate that the allocation logic and policies are applied across said multiple clouds, showing the interaction between the user demands and the cloud resources. The icons resembling servers or storage units, labelled as "collective resources of multiple clouds", indicates the pooled resources from various cloud providers that are available for allocation to users. The arrows coming from the clouds down to said resources signify the direction of resource assignment based on the allocation logic and policy.
[00069] The overall concept is that of a cloud broker or a management layer that takes into account multiple users with different demands and allocates resources from a collective pool of multiple clouds. The system aims to efficiently utilize cloud resources, maintaining balance and fairness among multiple tenants, and possibly also optimizing for cost, performance, or other specified criteria.
[00070] Multitenant resource allocation logic for pool of cloud service provider emphasizes a model for allocating resources in a multi-tenant environment, which is a common scenario in cloud computing where multiple users or customers share the same infrastructure or platform.
[00071] According to a pictorial illustration portrayed in FIG. 4, represents a block diagram of the multi-objective resource allocation process. The abstract representation of a resource allocation system used in cloud computing, specifically focusing on how multiple objectives are considered when assigning resources to users. On the left side of FIG. 4, there are several icons representing users. Said users could be individuals or entities that are requesting resources from a cloud service provider.
[00072] Further, in FIG. 4 represented a data structure or an object that holds the requirements of the users. Each Object requirement has at least two attributes, user id (the unique identifier of the user) and req (the list of requirements that the user has). The "Allocation Logic" is likely the core component where the system determines how resources should be allocated based on user requirements and the available resources.
[00073] The label "Allocation," which has attributes such as userid, vmid (virtual machine identifier), and allprice (allocation price). The allocation suggests that the allocation logic takes into account the user's ID, selects an appropriate virtual machine, and calculates the price for allocation. There are also methods or functions like allocate () which probably assigns resources based on the requirements and available virtual machines, and map requirement () which might be a function to map user requirements to specific resources. The data structure or object that seems to represent the virtual machines available for allocation. The data structure contains attributes such as vmid, vmtype (type of the virtual machine), cloud id (identifier for the cloud where the machine is hosted), char (which might stand for characteristics of the virtual machine), and price.
[00074] On the right side, there are two sets of icons labeled "Logical Resources" and "Physical Resources." Logical resources could refer to virtualized resources such as virtual machines, virtual networks, or storage, which are abstractions over the physical hardware. Physical resources are the actual hardware like servers, storage units, and network equipment. The icons seem to be color-coded and organized into columns, possibly to represent different types or tiers of resources.
[00075] Arrows show the interactions between users, their requirements, the allocation logic, and the virtual machines. Users present their requirements, which are processed by the allocation logic to determine the best virtual machine fit based on the objectives, which could include cost, performance, availability, and the like. Thus, the FIG. 4 outlines the multi-objective resource allocation process as a sequence of interactions between users, their requirements, and the resources provided by the cloud service.
[00076] The system must consider various factors and constraints to efficiently allocate resources that satisfy user requirements while optimizing for multiple objectives, which could include cost efficiency, performance, and resource utilization. FIG. 4 can be useful in explaining a cloud computing architecture, a resource management algorithm, or an IT infrastructure planning guide, where multi-objective optimization is an important consideration.
[00077] Example embodiments herein have been described above with reference to block diagrams and flowchart illustrations of methods and apparatuses. It will be understood that each block of the block diagrams and flowchart illustrations, and combinations of blocks in the block diagrams and flowchart illustrations, respectively, can be implemented by various means including hardware, software, firmware, and a combination thereof. For example, in one embodiment, each block of the block diagrams and flowchart illustrations, and combinations of blocks in the block diagrams and flowchart illustrations can be implemented by computer program instructions. These computer program instructions may be loaded onto a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions which execute on the computer or other programmable data processing apparatus create means for implementing the functions specified in the flowchart block or blocks.
[00078] Throughout the present disclosure, the term ‘processing means’ or ‘microprocessor’ or ‘processor’ or ‘processors’ includes, but is not limited to, a general purpose processor (such as, for example, a complex instruction set computing (CISC) microprocessor, a reduced instruction set computing (RISC) microprocessor, a very long instruction word (VLIW) microprocessor, a microprocessor implementing other types of instruction sets, or a microprocessor implementing a combination of types of instruction sets) or a specialized processor (such as, for example, an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a digital signal processor (DSP), or a network processor).
[00079] The term “non-transitory storage device” or “storage” or “memory,” as used herein relates to a random access memory, read only memory and variants thereof, in which a computer can store data or software for any duration.
[00080] Operations in accordance with a variety of aspects of the disclosure is described above would not have to be performed in the precise order described. Rather, various steps can be handled in reverse order or simultaneously or not at all.
[00081] While several implementations have been described and illustrated herein, a variety of other means and/or structures for performing the function and/or obtaining the results and/or one or more of the advantages described herein may be utilized, and each of such variations and/or modifications is deemed to be within the scope of the implementations described herein. More generally, all parameters, dimensions, materials, and configurations described herein are meant to be exemplary and that the actual parameters, dimensions, materials, and/or configurations will depend upon the specific application or applications for which the teachings is/are used. Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific implementations described herein. It is, therefore, to be understood that the foregoing implementations are presented by way of example only and that, within the scope of the appended claims and equivalents thereto, implementations may be practiced otherwise than as specifically described and claimed. Implementations of the present disclosure are directed to each individual feature, system, article, material, kit, and/or method described herein. In addition, any combination of two or more such features, systems, articles, materials, kits, and/or methods, if such features, systems, articles, materials, kits, and/or methods are not mutually inconsistent, is included within the scope of the present disclosure.

Claims
I/We Claim:
1. A multi-objective resource allocation system, comprising: a pool of cloud service providers; an allocation unit configured with a policy engine to receive and process resource requests from subscribers; and a joint resource pool formed by resources of the cloud service providers, wherein the allocation unit is further configured to allocate resources from the joint resource pool based on request attributes and allocation policies derived from the interaction of subscriber requirements and cloud service provider attributes.
2. The system of claim 1, wherein the allocation unit includes an interface for categorizing resource requests into targeted requests, which are directed to a specific cloud service provider, and open requests, which are not specific to any cloud provider, and is further configured to manage said requests such that targeted requests are fulfilled by the specified cloud provider and open requests are processed through a competitive bidding process.
3. The system of claim 1, wherein the joint resource pool is dynamically managed and includes a mix of static virtual machines, which are pre-provisioned and continuously available, and dynamic virtual machines, which are provisioned on-demand based on the current allocation requirements, with the valuation of said virtual machines being determined by the Volume Weighted Average Price (VWAP) method to maintain fair market value assessments.
4. The system of claim 3, wherein the allocation unit is further configured with a scheduler that applies a Stackelberg competition model to establish a hierarchy among the cloud service providers, based on the valuation of resources, to optimize the order of resource allocation and ensure efficient market dynamics.
5. The system of claim 4, wherein the allocation unit is adapted to facilitate an auction mechanism for open requests that allows bidding on non-allocated static virtual machines, and subsequently, based on auction outcomes and unserved requests, calculate and adjust Stackelberg output volumes for provisioning dynamic virtual machines from the respective cloud providers.
6. A method for allocating resources in a multi-objective resource allocation system, the method involving: receiving a plurality of resource requests from subscribers; categorizing the requests into targeted requests that specify a preferred cloud service provider and open requests that are provider-agnostic; and allocating resources from a joint resource pool formed by the collective capabilities of multiple cloud service providers, with allocations being performed in accordance with both the categorization of the requests and predetermined allocation policies that consider both subscriber requirements and provider attributes.
7. The method of claim 6, wherein valuing virtual machines includes applying the VWAP method to determine a current market value for the resources, and wherein determining the sequence of cloud service provider engagement in resource allocation is based on said valuation, thereby establishing price leadership within the provider pool.
8. The method of claim 6, including assigning targeted requests to predetermined cloud service providers at fixed rates, whereby said rates are established by contractual agreements and remain unadjusted for market fluctuations.
9. The method of claim 8, which includes implementing an auction-based mechanism for open requests, where non-allocated static virtual machines are made available to the highest bidder, followed by a calculation of Stackelberg output volumes for the dynamic provisioning of virtual machines to satisfy remaining demands.
10. The method of claim 9, further involving the step of price rebalancing at the end of each allocation cycle, wherein the rebalancing is achieved through the application of technical analysis techniques prescribed by the allocation policy.

METHOD FOR EFFECTIVE RESOURCE ALLOCATION IN CLOUD COMPUTING USING DYNAMIC STACKELBERG GAME BASED MULTI-OBJECTIVE APPROACH
Abstract
The disclosure relates to a multi-objective resource allocation system for cloud computing environments, designed to optimally allocate computational resources from a collective pool shared by multiple cloud service providers. The system includes an advanced allocation unit equipped with a sophisticated policy engine that processes incoming resource requests from subscribers by analyzing their specific attributes. The allocation unit judiciously assigns resources from the aggregated pool, making decisions based on a rich matrix of subscriber demands and the unique characteristics of each cloud service provider's offerings. By intelligently mediating between user requirements and provider attributes, the system facilitates a dynamic and strategic distribution of resources, ensuring efficient utilization and optimal service delivery within the multifaceted ecosystem of cloud services.
Fig. 1 , Claims:Claims
I/We Claim:
1. A multi-objective resource allocation system, comprising: a pool of cloud service providers; an allocation unit configured with a policy engine to receive and process resource requests from subscribers; and a joint resource pool formed by resources of the cloud service providers, wherein the allocation unit is further configured to allocate resources from the joint resource pool based on request attributes and allocation policies derived from the interaction of subscriber requirements and cloud service provider attributes.
2. The system of claim 1, wherein the allocation unit includes an interface for categorizing resource requests into targeted requests, which are directed to a specific cloud service provider, and open requests, which are not specific to any cloud provider, and is further configured to manage said requests such that targeted requests are fulfilled by the specified cloud provider and open requests are processed through a competitive bidding process.
3. The system of claim 1, wherein the joint resource pool is dynamically managed and includes a mix of static virtual machines, which are pre-provisioned and continuously available, and dynamic virtual machines, which are provisioned on-demand based on the current allocation requirements, with the valuation of said virtual machines being determined by the Volume Weighted Average Price (VWAP) method to maintain fair market value assessments.
4. The system of claim 3, wherein the allocation unit is further configured with a scheduler that applies a Stackelberg competition model to establish a hierarchy among the cloud service providers, based on the valuation of resources, to optimize the order of resource allocation and ensure efficient market dynamics.
5. The system of claim 4, wherein the allocation unit is adapted to facilitate an auction mechanism for open requests that allows bidding on non-allocated static virtual machines, and subsequently, based on auction outcomes and unserved requests, calculate and adjust Stackelberg output volumes for provisioning dynamic virtual machines from the respective cloud providers.
6. A method for allocating resources in a multi-objective resource allocation system, the method involving: receiving a plurality of resource requests from subscribers; categorizing the requests into targeted requests that specify a preferred cloud service provider and open requests that are provider-agnostic; and allocating resources from a joint resource pool formed by the collective capabilities of multiple cloud service providers, with allocations being performed in accordance with both the categorization of the requests and predetermined allocation policies that consider both subscriber requirements and provider attributes.
7. The method of claim 6, wherein valuing virtual machines includes applying the VWAP method to determine a current market value for the resources, and wherein determining the sequence of cloud service provider engagement in resource allocation is based on said valuation, thereby establishing price leadership within the provider pool.
8. The method of claim 6, including assigning targeted requests to predetermined cloud service providers at fixed rates, whereby said rates are established by contractual agreements and remain unadjusted for market fluctuations.
9. The method of claim 8, which includes implementing an auction-based mechanism for open requests, where non-allocated static virtual machines are made available to the highest bidder, followed by a calculation of Stackelberg output volumes for the dynamic provisioning of virtual machines to satisfy remaining demands.
10. The method of claim 9, further involving the step of price rebalancing at the end of each allocation cycle, wherein the rebalancing is achieved through the application of technical analysis techniques prescribed by the allocation policy.