International Journal of Engineering Trends and Technology (IJETT) – Volume 9 Number 2 - Mar 2014

A Framework for Pricing Schemes for Networks

under Complete Information
Aditya Goyal#1, Akanksha Tyagi#2, Manisha Bhende*3, Swapnil Kawade#4
#
Student & Department of Computer Engineering,
Padmashree Dr. D.Y. Patil Institute of Engineering & Technology,
Pimpri, Pune, Maharashtra, India
*
Faculty & Department of Computer Engineering,
Padmashree Dr. D.Y. Patil Institute of Engineering & Technology,
Pimpri, Pune, Maharashtra, India

Abstract— The revenue maximization problem of service provider can distinguish different groups of users, it
provider is considered and different pricing schemes to solve the announces the pricing and the admission control decisions to
above problem are implemented. The service provider can different groups of users.
choose an apt pricing scheme subjected to limited resources, if he In this paper, the optimal usage-based pricing problem in a
knows the utility function and identity of the user. The complete
resource-constrained network with one profit-maximizing
price differentiation can achieve a large revenue gain but has
high implementation complexity. The partial price service provider and multiple groups of surplus-maximizing
differentiation scheme to overcome the high implementational users is studied. In wireless communication networks,
complexity of complete price differentiation scheme is also however, the usage-based pricing scheme seems to become
studied. A polynomial- time algorithm is designed for partial increasingly popular due to the rapid growth of wireless data
price differentiation scheme that can compute the optimal partial traffic.
differentiation prices. The willingness of the users to pay is also
considered while designing price differentiation schemes.
II. RELATED WORK
Keywords— Network Pricing, Price Differentiation, Revenue
Management, Resource Allocation
The maximum revenue that can be achieved by a
monopolistic service provider under complete network
I. INTRODUCTION information has been studied in [1]. The authors proposed two
Pricing is an important concern for management of pricing schemes with incomplete information, and showed
networks. It also used for design and operation of networks. that by properly combining the two schemes there would be
Many sophisticated pricing mechanisms to extract surpluses very small revenue loss in a two-group case while maintaining
from the consumers and maximize revenue (or profits) for the the incentive compatibility.
providers have been proposed. A typical example is the A model to study the important role of time-preference in
optimal nonlinear pricing. In practice, however, it is often network pricing has been presented in [2]. In the model
observed simple pricing schemes being deployed by the presented, each user chooses his access time based on his
service providers. Typical examples include flat-fee pricing preference, the congestion level, and the price that he would
and (piecewise) linear usage-based pricing. The optimal be charged. Without pricing, the "price of anarchy" (POA) can
pricing schemes derived in economics often have a high be arbitrarily bad. The authors then derived a
implementation complexity. Besides having a higher simple pricing scheme to maximize the social welfare. From
maintenance cost, complex pricing schemes are not “customer the SP's viewpoint, the authors considered the revenue-
friendly” and discourage customers from using the services. maximizing pricing strategy and its effect on the social
The task of achieving the highest possible revenue often with welfare. The authors showed that if the SP can differentiate
complicated pricing schemes requires knowing the its prices over different users and times, the maximal revenue
information (identity and preference) of each customer, which can be achieved, as well as the maximal social welfare.
can be challenging in large scale communication networks. However, if the SP had insufficient information about the
The service provider wants to maximize its revenue by users and can only differentiate its prices over the access times,
setting the right pricing scheme to induce desirable demands then the resulting social welfare, especially when there are
from users. Since the service provider has a limited total many low-utility users, can be much less than the optimum.
resource, it must guarantee that the total demand from users is Otherwise, the difference is bounded and less significant.
no larger than what it can supply. The details of pricing The results on price-based discrimination for bandwidth
schemes depend on the information structure of the service allocation in wire-line communication networks were
provider. Under complete information, since the service presented in [3]. In general, the problem of mechanism design

ISSN: 2231-5381 http://www.ijettjournal.org Page 69

 International Journal of Engineering Trends and Technology (IJETT) – Volume 9 Number 2 - Mar 2014

for resource allocation was very complex, and the focus was Some GPRS / data UMTS access to the Internet in some
on studying simple mechanisms that show promise of countries of Europe has no flat rate pricing, following the
widespread adoption in the arena of Internet pricing. The traditional "metered mentality". Because of this, users prefer
objective was to study the revenue efficiency of single and using fixed lines (with narrow or broadband access) to
multi-class pricing schemes as compared to the maximum connect to the Internet.
possible revenue.
In particular, the focus was on flat entry fees as the IV. PROPOSED METHODOLOGY
simplest pricing rule. A lower bound for the ratio between the The partial price differentiation problem includes complete
revenue from this pricing rule and maximum revenue, which price differentiation scheme and single pricing scheme as
the author referred to as the Price of Simplicity was presented. special cases. The optimal solution to partial price
The characterizations of types of environments that lead to a differentiation problem is found out. The differentiation gain
low Price of Simplicity was done and it was shown that the and the effective market size are the two important factors
loss of revenue from using simple entry fees was small in a behind the revenue increase of price differentiation schemes.
range of environments. A network with a total amount of S limited resource is
In the communication network pricing literature, it is the considered. The resource can be in the form of rate, bandwidth,
linear pricing schemes that have been largely adopted as the power, time slot, etc. The monopolistic service provider
means of controlling network usage or generating profits for allocates the resource to a set I = {1, . . . , I} of user groups [8].
network service providers. In [4], the authors extended the Each group i ∈ I has Ni homogeneous users with the same
framework mentioned above to non-linear pricing and utility function:
investigated optimal nonlinear pricing policy design for a ui (si) = ϴi ln(1 + si)
monopolistic service provider. The problem was formulated as Where si is the allocated resource to one user in group i and ϴi
an incentive-design problem, and incentive (pricing) policies represents the willingness to pay of group i.
were obtained for a many-users regime, which enabled the It is assumed that ϴ1>ϴ2> …. >ϴI. Since the service
service provider to approach arbitrarily close to Pareto- provider has a limited total resource, it must guarantee that the
optimal solutions. total demand from users is no larger than what it can supply.
The details of pricing schemes depend on the information
III. EXISTING SYSTEM structure of the service provider.
A system model of charging, routing and flow control, The correspondence between the service provider and
where the system comprises both users with utility functions user can be described as follows:
and a network with capacity constraints has been described by
Frank Kelly[5]. Standard results from the theory of convex
optimization show that the optimization of the system may be
decomposed into subsidiary optimization problems, one for
each user and one for the network, by using price per unit
flow as a Lagrange multiplier that mediates between the
subsidiary problems.
TCP variants have recently been reverse-engineered to
show that they are implicitly solving this problem, where the
source rate vector x ≥ 0 is the only set of optimization
variables, and the routing matrix R and link capacity vector c
are both constants in [9].
maximize ∑ Us(xs)
subject to Rx≤c
Utility functions Us are often assumed to be smooth,
increasing, concave, and dependent on local rate only,
although recent investigations have removed some of these
assumptions for applications where they are invalid.
Fig. 1. General Proposed Model
Flat-fee pricing: For Internet service providers, flat The service provider declares the pricing schemes in
rate is access to the Internet at all hours and days of the year Stage 1 and users interact with their demands in Stage 2. The
(linear rate) and for all customers of the telco operator users demand to maximize their surplus by optimizing their
(universal) at a fixed and cheap tariff. Flat rate is common claim according to the pricing scheme. The service provider
in broadband access to the Internet in the USA and many maximizes his revenue by making available a right pricing
other countries. scheme to users.
It is considered that the service provider has the complete
A charge tariff is a class of linear rate where the user
information of the user. The service provider can choose from
is charged on the basis of uploads and downloads (data
transfers) and hence differs from the flat rate system.

ISSN: 2231-5381 http://www.ijettjournal.org Page 70

 International Journal of Engineering Trends and Technology (IJETT) – Volume 9 Number 2 - Mar 2014

complete price differentiation scheme, the single pricing differentiation scheme. The clusters are defined. Each cluster
scheme, and the partial price differentiation scheme. is a set of groups which have been charged same price. The
partial price differentiation problem is solved in three levels.
A. Complete Price Differentiation
The service provider knows the utility and the identity of
each user; it is possible to maximize the revenue by charging a
different price to each group of users.

Fig. 4 Partial Price Differentiation

Algorithm:
Fig. 2 Complete Price Differentiation Step 1: Pricing and resource allocation in each cluster: For a
fix partition α and a cluster resource allocation, focus on
Algorithm: pricing and resource allocation problem within each cluster.
Step 1: Solve users’ maximization problem which leads to Step 2: Resource allocation among clusters: For a fix partition
unique optimal demand. α allocate resources among clusters.
Step 2: In stage 1, the service provider maximizes his revenue Step 3: Cluster Partition: Cluster partition problem is solved.
by choosing a price Pi and number of users ‘n’ for each group
I, subject to total resource ‘S’.
Step 3: Perform Complete Price Differentiation scheme by V. MATHEMATICAL MODEL
charging each group by different price.
A. Complete Price Differentiation
B. Single Pricing This method can be carried out in two stages as given
In this scheme, the service provider charges a single price below:
to all groups of users. This scheme may suffer a considerable 1) User’s Surplus Maximization Problem:
revenue loss. If a user in group i has been admitted into the network and
offered a linear price pi , then it solves the following surplus
maximization problem [8],
Maximize ui (si) – pi si
si ≥0

2) Service Provider’s Pricing and Admission Control

Problem:
CP: maximize ∑ni pi si
p≥0,s≥0,n i∈ G
subject to si = ( ϴi / pi - 1 ) +, i ∈ G,
Fig. 3 Single Pricing
ni ∈ {0, . . . , Ni } , i ∈ G,
∑ni si ≤ S
Algorithm:
i∈G
Step 1: Single pricing follows the same approach as complete
pricing scheme. A solution is obtained which shares a similar B. Single Pricing
structure as complete price differentiation. This problem can be formulated as given below [8]:
Step 2: There exists an optimal solution of this scheme that SP: maximize p ∑n i si
satisfies following conditions: p≥0, n i∈G
 All users are admitted: n*I = Ni for all i ∈I[8]. subject to si = ( ϴi / pi - 1 ) +, i ∈ G,
 There exists a price p* and a group index threshold ni ∈ {0, . . . , Ni } , i ∈ G,
Ksp ≤ I such that only the top Ksp groups of users ∑ni si ≤ S
receive positive resource allocations [8]. i∈G

C. Partial Price Differentiation Here the service provider charges a single price p to all groups
of users.
The service provider offers only a few pricing plans for the
entire users population; it is termed as the partial price

ISSN: 2231-5381 http://www.ijettjournal.org Page 71

 International Journal of Engineering Trends and Technology (IJETT) – Volume 9 Number 2 - Mar 2014

C. Partial Price Differentiation <<extend>>

Set Pricing Parameters

The Partial Price differentiation (PP) problem is formulated as Set Features

follows [8].
Admin
PP: maximize ∑ni pi si
ni, pi, si,pj , ai j i∈G
subject to si = ( ϴi / pi - 1 ) +, ∀i ∈ G, Customer Registration

ni ∈ {0, . . . , Ni } , ∀i ∈ G, <<extend>> Get User Ty pe

∑ni si ≤ S
i∈G Customer
Login

<<extend>>
The service provider can choose the price charged to each System

System
group. Usage Charge for Normal Users

VI. ANALYSIS MODEL <<extend>>

The system consists of two modules as shown in Fig. 5. Usage + additional features Charge for Partial Premium Users

• User Module
The User has to register first and only then he can use the
product.
• Service Provider Module
Service Provider selects a pricing scheme to maximize his
revenue. All Feature Charge

Fig. 6 System Analysis Model

VII. APPLICATIONS
 The Omaha based Network Pricing, Mitec, has a
variety of pricing options that target personal,
business and corporate users. They offer Web
development solutions and a variety of different
access speeds. For users seeking "personal
solutions," for example, Mitec offers them an option
of a flat-fee account with unlimited access for $19.95
(flat-fees also help in gaining market share), or a
tiered account with a flat-fee of $9.95 for the first 20
hours and a $1 additional charge for each hour
Fig. 5 System Architecture
thereafter. For the family on the Internet, Mitec offers
unlimited access for $24.95 with five separate email
accounts.

 AT&T shifted to usage-based wireless data plans in

June 2010 [6]. AT&T announced two new wireless
data plans based on the amount of data subscribers
use. The change spelled the end of unlimited wireless
data use for new customers and likely higher charges
for existing customers who use more than 2GB of
data per month for activities such as watching videos
and online gaming.
 Verizon followed up with similar plans (like AT&T)
in July 2011. Similar usage-based pricing plans have
been adopted by major Chinese wireless service
providers including China Mobile and China
UniCom.

ISSN: 2231-5381 http://www.ijettjournal.org Page 72

 International Journal of Engineering Trends and Technology (IJETT) – Volume 9 Number 2 - Mar 2014

[2] L. Jiang, S. Parekh, and J. Walrand, “Time-Dependent Network Pricing

VIII. CONCLUSION and Bandwidth Trading,” in Proceedings of IEEE Network Operations
The revenue-maximizing problem for a monopoly service and Management Symposium Workshops, 2008.
[3] S. Shakkottai, R. Srikant, A. Ozdaglar, and D. Acemoglu, “The Price
provider under complete network information is studied. The of Simplicity,” IEEE Journal on Selected Areas in Communications,
focus is to investigate the trade-off between the total revenue vol. 26, no. 7, pp. 1269–1276, 2008.
and the implementational complexity (measured in the [4] H. Shen and T. Basar, “Optimal nonlinear pricing for a monopolistic
number of pricing choices available for users) under complete network service provider with complete and incomplete information,”
IEEE Journal on Selected Areas in Communications, vol. 25, no. 6,
information. The partial price differentiation is the most pp.1216–1223, 2007.
general one among the three pricing differentiation schemes [5] F. Kelly, “Charging and rate control for elastic traffic,” European
that have been proposed (i.e., complete, single, and partial), Transactions on Telecommunications, vol. 8, no. 1, pp. 33–37, 1997.
and includes the other two as special cases. An algorithm that [6] “AT&T shifts to usage-based wireless data plan.” available on AT&T
official website.
computes the optimal partial pricing scheme in polynomial [7] A. Daoud, T. Alpcan, S. Agarwal, and M. Alanyali, “A stackelberg
time, and numerically quantizes the trade-off between game for pricing uplink power in wide-band cognitive radio networks,”
implementational complexity and total revenue has been in Proceedings of IEEE CDC, 2008.
designed. [8] Shuqin Li, Jianwei Huang, “Price Differentiation for Communication
Networks”, in IEEE TRANASCTIONS ON NETWORKING, 2013.
[9] M. Chiang, S. Low, A. Calderbank, and J. Doyle, “Layering as
REFERENCES optimization decomposition: A mathematical theory of network
[1] S. Li, J. Huang, and S-YR Li, “Revenue maximization for architectures,” Proceedings of the IEEE, vol. 95, pp. 255–312, 2007.
communication networks with usage-based pricing,” in Proceedings of
IEEEGLOBECOM, 2009.

ISSN: 2231-5381 http://www.ijettjournal.org Page 73