Research on Charge Pricing Model of New Energy Cloud Service

Jinsong Liu

, Yu Shan

, Caijuan Qi

, Xiangying Xie

, Zelong Zhang

, Yufei Ai

and Leyi Ge

Development Planning Department, State Grid Co., Ltd., Beijing, China

Photovoltaic Cloud Division , China Net E-commerce Co., Ltd., Beijing, China

Institute of Economics and Technology, National Network Ningxia Electric Power Co., Ltd., Ningxia, China

zelongncepu@163.com, ayf0919@163.com, geleyi@sgec.sgcc.com.cn

Keywords: New Energy Cloud Service, Pricing Model, New Energy Cloud Business Model.

Abstract: The new energy cloud aims to help achieve the national clean energy transformation and promote the

upgrading of energy transformation production and consumption. Based on China's energy resources and

energy transformation needs, as an emerging field, the new energy cloud business model still has a lot of

research space. In order to use price leverage to adjust the relationship between user demand and cloud

service supply, this paper examines its cloud service pricing model. This paper studies the contract pricing

method to make the use of idle resources in the service market more reasonable, improve the utilization

efficiency of cloud service resources, and alleviate market pressure. Finally, the purpose of rationally using

cloud service resources and optimizing market cloud service resources is achieved.

1 INTRODUCTION

Cloud computing resources are virtualized by a huge

number of physical IT resources such as CPU,

storage, and network. The amount of tasks the

system has to handle is very arduous, and the

resources are dynamic and heterogeneous. In the

cloud environment, there are a large number of

users. In addition, as cloud computing shifts from

technology to application, system-based resource

allocation can no longer meet business needs. There

is an urgent need to use the market as a carrier to

study the optimization of cloud services. Therefore,

"cloud computing" as a business model, how to

schedule resources in the cloud service market

efficiently and reasonably, and how to conduct

pricing and transaction reasonably to achieve the

global optimization of the market has become the

focus and difficulty of cloud computing market

research (

Lin, Su, Meng, Liu, Liu, 2013). Scholars such

as Christof Weinhardt proposed two important

pricing models: pay-as-you-go pricing model and

subscription pricing model. Based on that, Lamia

Youssef and other scholars proposed three forms of

pricing models, and pointed out that no matter which

pricing model, it must adopt at least one form. The

three forms are: per-unit pricing, tiered-pricing, and

subscription-based pricing (

Eason, Noble, Sneddon,

1955, Jose, Calero, Aguado, 2015, Yu, Chen, Li, 2014

In addition, Amazon had also proposed a dynamic

pricing form as pricing on-site instances. Typical

cloud service providers, such as Google, Amazon,

and Microsoft, all adopt these four forms in their

pricing mod.

2 PRICING MECHANISM FOR

CLOUD SERVICE RESOURCES

Pricing has become an important link in the

development of new energy cloud services. Only

mature pricing strategies and mechanisms can

ensure the commercial realization of the entire new

energy cloud service market. Due to the different

usage of service resources by users, in order to meet

the diverse needs of users, the service resources

provided by the market need to be expanded

indefinitely and can be obtained at any time by

users. Therefore, how to provide users with flexible,

diverse and convenient pricing mode has become the

key to survival in the current cloud service market.

There are many pricing factors involved in the

pricing of the cloud service market, and the pricing

work is relatively difficult and complicated. Internet

has become not only a simple tool for transmitting

service information, but also a network tool for

918

Liu, J., Shan, Y., Qi, C., Xie, X., Zhang, Z., Ai, Y. and Ge, L.

Research on Charge Pricing Model of New Energy Cloud Service.

DOI: 10.5220/0011358900003440

In Proceedings of the International Conference on Big Data Economy and Digital Management (BDEDM 2022), pages 918-923

ISBN: 978-989-758-593-7

 2022 by SCITEPRESS – Science and Technology Publications, Lda. All r ights reserved

transmitting computing power in a cloud

environment. Pricing is not just pricing for network

resources, but involves more pricing factors, which

brings new challenges for network resources.

Currently, the pricing mechanism of the cloud

service market mainly includes the following:

Amazon's EC2 pricing mechanism, and Google App

Engine's pricing mechanism. In order to meet the

needs of different users, Amazon has developed a

flexible and elastic pricing mechanism. Amazon's

Elastic Compute Cloud is the largest online retailer

on the Internet. Amazon has introduced different

billing models based on multiple service instances of

cloud service resources. In terms of use, users need

to select a virtual machine with corresponding

configuration (instance storage space, memory,

computing unit, etc.) as the running instance

according to their own needs. However, Amazon

will generally choose one or multiple running

instances at the same time to meet user needs when

deploying network programs. In the elastic cloud

EC2, users can manage each instance on the cloud

platform through the web operation interface. The

instance payment pricing method generally varies

depending on the user's usage. The instance payment

method of Elastic Cloud EC2 is hierarchical, which

is mainly based on the two aspects of instance type

and specification type. Users select the appropriate

instance type and specification type according to

their own needs. Elastic Cloud EC2 is based on the

computing power required to process user tasks and

the configuration of the selected virtual machine as

the upper limit of the parameters. The EC2 instance

types are summarized into the following three

categories. It is large-scale, small-scale and medium-

sized instance. Amazon sets prices for them

according to the above three instance methods,

namely contract pricing, on-demand pricing, and

spot real-time pricing. The pricing of data

transmission mainly considers two aspects: the

amount of outgoing and incoming data. The

outgoing data is free of charge within the first 1GB

flow rate per month, and the subsequent service flow

is charged according to the user's actual usage. The

more the price, the cheaper it is. The incoming data

is priced at a fixed price. There is no charge for data

transmission in the internal network, and the price is

based on the amount of data in different regional

networks. Data storage uses hard drives as the

pricing factor, and adopts a comprehensive pricing

strategy for duration and usage. Google App Engine

implements a flexible and changeable service

pricing mechanism. Each service resource is

measured and charged according to two parts: a

fixed quota and a charging quota, and allows daily

users to purchase resources in advance according to

their own wishes. The charging quota is the

maximum resource usage required by the user daily.

The fixed quota is the maximum amount of service

resources that each user can obtain on the market set

by Google itself.

3 NEW ENERGY CLOUD

SERVICE CHARGE PRICING

MODEL

In order to formulate a pricing model for the new

energy cloud service market. There are three

aspects: First, determine the contract and on-demand

pricing strategy for new energy cloud services in the

electricity market under the condition of maximizing

user utility. Second, based on contracts and on-

demand pricing to determine the market's optimal

pricing strategy according to the maximization of

market revenue. Third, according to the actual use of

the market, there will be a large amount of idle

resources during certain periods of use by users. At

this time, it is necessary to study and formulate

pricing strategies for spot instances to make

reasonable use of these idle resources. Pricing at this

time is also seen as an incentive mechanism to adjust

users' market service usage behavior, so as to

achieve market equilibrium and efficient use of the

overall market. The cloud service provider develops

and provides available cloud services continuously.

On the one hand, the resources that constitute the

service requires cost expenditure; on the other hand,

because the provision of resources or services is in

order to profit, the provider will charge a certain fee

to users for resources or services. Here, the resource

is the basic module that constitutes the service,

which is controlled and managed by the provider.

Providers combine different resources into different

performance levels of services in a variety of ways

and provide them to customers. For customers, they

use services, resources are invisible to customers,

but resources are a key element for measuring

service performance and the level of service it can

reach. Therefore, customers sometimes have to pay a

certain fee in order to continue to rent and obtain the

qualification to use the service. Moreover, the level

and performance of the services obtained are related

to the level of fees paid. The higher the cost, the

higher the performance and level of the service. The

cost paid by the customer depends on the price of

Research on Charge Pricing Model of New Energy Cloud Service

919

the service provided by the provider. The following

are several cloud service charging pricing models:

3.1 Pay-As-You-Go Pricing Model

Pay-as-you-go is that customers pay for each unit of

resources they use at a stable price, sometimes called

pay-for-use. In other words, pay-as-you-go is

charged based on the amount of resources actually

used by the customer (such as the number of

instances, amount of data, or usage time, etc.). The

actual amount of resources used can be the total

amount of use after a period of time (such as one

month, one day, and one hour). During this time

period, the price of the resource is stable.

3.2 Contract Pricing Model

The contract price model is that the customer signs a

contract (in the form of prepayment) to order a

certain service or combination of services at a stable

price for a long period of time. By paying an

advance payment to order a service, some papers

refer to it as an on-demand advance payment. The

two pricing models also reflect the characteristics of

cloud computing services that can be expanded on

demand and flexibly customized. Therefore, the

pricing strategy of cloud computing service

providers for cloud services is based on these two

pricing models.

3.3 Pricing Per Unit

Pricing per unit is the basic form of the pay-as-you-

go pricing model, which is usually applied to data

transmission services or services that use memory.

The provider determines a fixed price per unit of

resource or service in advance, and then charges the

total service fee based on the total amount used by

the customer by multiplying the price per unit by the

total amount used. Cloud service providers such as

Google, Amazon, and Microsoft charge USD 0.01

per thousand per month for PUT and POST, and

charge USD 0.01 per ten thousand per month for

GET and HEAD. This is the price per unit, and the

units are every thousand times per month and every

hundred times per month. Another example is GAE

(Google App Engine) which provides APIs with a

free quota per application (/app/day) per day, and

pricing for parts exceeding the free quota.

Table 1: Table1 Pricing of APIs.

APIs Free Quota/

app/da

Price beyond Free

Quota(US)

Datastore

API

50k freeread/

write/small

$0.10/100k write ops

$0.07/100k read ops

$0.01/100k small ops

Blobstore

API

5G $0.13/G/month

Email API 100 recipients $0.01/100recipients

Gae APIs takes the form of pricing per unit, and

the unit of free quota is per application per day (/

APP / day). In other words, the free quota for daily

reading, writing or small operations of the datastore

API is 50K; The daily free quota of blobstore API is

5g; The daily free quota of email API is 100

recipients. The units that exceed the free quota are

different. For example, the datastoreapi charges $0.1

per 100k write operation, $0.07 per 100k read

operation and $0.01 per 100k small operation. Its

unit has become x operation per 100k (x refers to

read, write and small), which is different from the

free quota unit (every day). As for the charge per

"write" operation, it is higher than that per "read".

This is generally because "write" consumes more

storage devices than "read", and "write" is more

difficult to achieve than "read", which takes a long

time and occupies more resources. Therefore, the

cost of "write" operation is usually higher than that

of "read" operation. When the blobstore API exceeds

5g per day, it will charge us $0.13 per g per month,

and the unit will become per g per month (/ g /

month). It will be charged in natural months after the

total amount is accumulated. After the email API

receives more than 100 recipients per day, it will

charge us $0.01 for every 100 recipients, and its unit

will become every 100 recipients (/ 100recipients). It

is the same unit of measurement for the same

resource / service, but the free quota is different

from the unit of measurement after exceeding the

free quota, which shows that the pricing per unit is a

relatively flexible and simple pricing form. It allows

users to flexibly customize the size of resources and

the pricing method of resources according to specific

application requirements, which is suitable for

customers who need to frequently adjust and

calculate the scale of resources. Customers don't

have to think too much about the troublesome it

resource purchase plan, and can break up the large

cost of one-time purchase of resources into the

scattered cost of multiple purchases.

BDEDM 2022 - The International Conference on Big Data Economy and Digital Management

920

3.4 Tiered Pricing

Tiered pricing refers to the provision of cloud

resources to customers in the form of several

different levels, and different prices are set for

different levels of resources. The level of resource

mapping to the customer is the level of service.

Service Level (Service Level) is a set of expected

and implicit quality of service. The resources of each

level provide the same and fixed computing power

or storage capacity to customers in the service level.

The higher the level, the higher the quality of

services such as the quantity and performance of

resources, and the different prices. Provider ’ s

tiered pricing is a combination of resource tiered and

per-unit pricing. Classification is to classify the

performance or total amount of resources or

services, and the price per unit is the price of

resources or services within each class. Pay-as-you-

go pricing models also often use this pricing form.

The storage services, computing instances, and data

transmissions in Google, Amazon, and Microsoft

cloud services all adopt hierarchical pricing to

provide customers with different levels of services

and charge different fees based on the total amount

of resources used by customers. For example, the

computing instances provided by GCE (Google

Compute Engine) have four levels: small, medium,

large, and super large. The quality of service

achieved by each level is gradually improved, that

is, the configured virtual kernel, memory, and hard

disk gradually expand per unit.

The resources configured by GCE for the small

instance level are 1 virtual kernel, 3.75gb memory

and 420gb hard disk. This level is equivalent to

2.75x gceu. The price of small instances is US

$0.145 per hour. If calculated in units per gceu per

hour, one small instance is equivalent to 2.75 times

of gceu, and the price per gceu per hour is US

$0.053. Therefore, the cost of using small instances

per hour is US $0.053 * 2.75 = US $0.14575.

Generally speaking, the measurement of instance

computing power is nothing more than the number

of cores or the general measurement of CPU,

memory, hard disk and other industries. However,

sometimes cloud service providers will specially

formulate their own measurement units to measure

instance computing power for internal use, such as

Google's gceu and Amazon's ECU (ec2compute

unit). The customer can measure the computing

power of the instance according to the general

measurement of the industry, or the computing

power of the instance according to the supplier's

own measurement unit. When the provider provides

its own measurement unit to the customer, the

provider may price its own measurement unit per

unit, such as US $0.053 per gceu per hour. Of

course, the two pricing units can be converted to

each other, and the final calculated rate must be

consistent (one small instance can be converted to

2.75 times of gceu; the rate for small instances is

0.145 per hour, and the rate after gceu is converted

to small instances is US $0.14575. The error

between the two is 0.00075, and the rate is basically

the same.) therefore, in terms of classification, GCE

is divided into four levels: small, medium, large and

super large. In terms of measurement units, GCE has

two measurement units. One is the standard

measurement unit commonly used in the industry,

namely, the number of cores, memory, hard disk,

etc; The other is a self-defined unit of measure,

gceu. In terms of pricing units, GCE also divides

into two kinds of units of measurement. One is the

unit of measurement combining grade and service

time (US $0.145 per hour for small instances, US

$0.29 per hour for medium instances, US $0.58 per

hour for large instances and US $1.16 per hour for

super large instances). This is a typical hierarchical

pricing method combining resource grading and per

unit pricing; The other is the measurement unit ($/

gceu / hour) combining gceu and usage time, which

is the basic pricing method per unit and has nothing

to do with the classification of resources.

3.5 Scheduled Pricing

Scheduled pricing is a relatively preferential price

(relative preferential refers to the price per unit

pricing and tiered pricing) established by the

provider for customers to book a certain

consumption level of service. Of course, there is a

prerequisite for the relative discount, that is, the

customer needs to pay a deposit (or reservation) in

advance, and the provider can specifically reserve

the lease of resources for the customer. Since

reservation pricing does not accurately measure the

actual use of resources and services by customers,

On the one hand, providers may suffer losses due to

underestimation of the actual usage of resources

used by customers. On the other hand, the provider

may overestimate the actual use of resources,

making the resources underutilized, resulting in

inefficient use of resources。However, because the

provider shows the customer a relatively favorable

price model for the reservation form, it has become

the most widely used reservation pricing model that

attracts customers. At the same time, reservation

pricing combines the strategy of per-unit pricing and

Research on Charge Pricing Model of New Energy Cloud Service

921

hierarchical pricing, dividing resources and services

into different levels, determining the prepayment of

each level, and tiered pricing. Cloud computing is

considered to be a business model for the delivery of

new IT resources. Due to it is a new business service

that its price largely determines the extent to which

the service attracts customers, and thus determines

the provider's competitive advantage. In order to

compete for the huge benefits brought by cloud

computing, various cloud computing providers have

launched a "price war". Providers constantly adjust

the price level of services according to market

reflections, competitors ’ actions, their own

expectations, etc., constantly launch new service

models and new pricing methods to meet customer

service needs to a greater extent. Keep existing

customers and attract new customers to compete

with competitors.

4 OPTIMAL PRICING METHOD

FOR NEW ENERGY CLOUD

SERVICE MARKET-

CONTRACT PRICING

METHOD

Due to the particularity of new energy cloud

services, which provide services for specific objects

in specific markets, this paper studies the use of

contract pricing methods. The contract pricing

strategy refers to charging a fixed service fee to the

user during the service time period specified by the

user. The fee charged by the market has nothing to

do with the user's service resource usage during the

service time period. It is a contract that has been

signed with the market in advance. Such as monthly

pricing method is suitable for the early establishment

of the cloud market, under a small number of service

users, the cloud service providers in order to expand

the number of users and the rapid subsequent

occupation as a pricing market used. The monthly

pricing method is that the service provider provides

a fixed amount of cloud services for users, and a

fixed total fee is deducted once a month, and each

individual service is no longer separately deducted.

In the pricing method of the cloud service market,

contract pricing method is very common. It has the

following advantages: For cloud service providers,

they can take advantage of the obvious lower service

price advantage of the initial cloud service market

and the relatively low initial construction cost of the

cloud service market to provide users with a simple

and affordable way to obtain information services.

For users, a fixed service can generally be obtained

at a relatively suitable price, and the expenditure is

relatively fixed, which can stimulate the user's

service usage to a certain extent. Under the

contractual pricing strategy, maximizing the user’s

personal utility will drive the user to maximize the

use of cloud service resources in the market. When

all users in the market adopt this kind of contract

pricing strategy, it will lead to collective irrational

market congestion. Then the quality of cloud service

resources provided to users will decrease, and the

service time will increase. Therefore, it is necessary

to use a price mechanism to reduce the user's usage

of cloud service resources, so that the user's usage of

service resources is less than its corresponding

service resource utilization rate.

Contract pricing method is the most widely used

pricing model in the cloud environment for resolving

service negotiations with users. It is a model based

on a contract mechanism and is often used to

manage the exchange of goods and services in the

service market. It can help find an appropriate

service resource to complete a given user demand

task. Service resource consumers (users) publish

their budget and time required for task execution

through the cloud broker. If the service resource

provider is satisfied with the conditions issued by

the cloud broker, it will sign an agreement with the

cloud broker. The advantage of this model is that if

the selected service resource provider (cloud service

provider) cannot provide a satisfactory service

result, the cloud service resource consumer can

choose other capable cloud service resource

providers.

5 CONCLUSIONS

This paper studies the contract pricing model in

terms of pricing issues in the cloud service market.

This model does not require cloud brokers to

negotiate with buyers and sellers but directly signs

purchase contracts, which greatly simplifies the

process of resource invocation and improves the

service efficiency of service resources. However, in

the actual market billing, there are also on-demand,

spot instances, optimal pricing, and cost pricing

issues during crowded periods that need to be

considered. These issues should combine organically

to regulate market supply and demand effectively, so

as to improve the market environment.

BDEDM 2022 - The International Conference on Big Data Economy and Digital Management

922

ACKNOWLEDGEMENTS

This Paper was supported by the science and

technology project of State Grid Corporation of

China “Research on Typical Application Scenarios,

Business Model and Operation Simulation

Assessment Technique of New Energy Cloud”

(SGNXJY00GHJS2000018).

REFERENCES

Chuang Lin, Wen-Bo Su, Kun Meng, Qu Liu, Wei-Dong

Liu. Cloud computing security: architecture,

mechanism and modeling [J]. Chinese Journal of

Computers, 2013, 9(6):1765-1784.

Carlos Antunes, Ricardo Vardasca. Performance of Jails

versus Virtualization for Cloud Computing

Solutions[J]. Procedia Technology, 2014, 5(16): 649-

658.

Farrukh Shahzad. State-of-the-art Survey on Cloud

Computing Security Challenges, Approaches and

Solutions[J]. Procedia Computer Science, 2014,

10(37):357-362.

G. Eason, B. Noble, and I. N. Sneddon, “On certain

integrals of Lipschitz-Hankel type involving products

of Bessel functions,” Phil. Trans. Roy. Soc. London,

vol. A247, pp.529–551, April1955.

Jose M. Alcaraz Calero, Juan Gutiérrez Aguado.

Comparative analysis of architectures for monitoring

cloud computing infrastructures[J]. Future Generation

Computer Systems, 2015, 2(47): 16-30.

Miao Wang, GuiLing Wang, Jie Tian, et.al. An Accurate

and Multi-faceted Reputation Scheme for Cloud

Computing[J]. Procedia Computer Science, 2014,10

(34):466-473.

Mark Stieninger, Dietmar Nedbal, Werner Wetzlinger,

et.al. Impacts on the Organizational Adoption of

Cloud Computing: A Reconceptualization of

Influencing Factors[J]. Procedia Technology, 2014,

(16), 85-93.

M. Conti, S. Giordano, M. May, A. Passarella. From

opportunistic networks to opportunistic computing[J].

IEEE Commun. Mag., 2010,48 (9):126–139.

Qiang Yu, Ling Chen, Bin Li. Ant colony optimization

applied to web service compositions in cloud

computing[J]. Computers & Electrical Engineering,

2014,8(41):18-27.

S.Abolfazli, Z. Sanaei, E. Ahmed. Cloud-Based

Augmentation for Mobile Devices: Motivation,

Taxonomies, and Open Challenges[J].

Communications Surveys & Tutorials, IEEE, 2014, 16

(1): 337–368.

Wang Yi-Jie, Sun Wei-Dong, Zhou Song, Pei Xiao-Qiang,

Li Xiao-Yong. Key Technologies of Distributed

Storage for Cloud Computing[J]. Journal of Software.

2012:23(4): 962-986.

Research on Charge Pricing Model of New Energy Cloud Service

923